April 6, 2010 
< | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | >
April 6, 2010 [00:00:25] <tsoome> yea, well, but that means you use still gcc, not versioned name[00:00:40] <tsoome> just an guess tho, it can work just fine[00:00:46] <Stric> well, you can pick any of them[00:00:48] <alanc> stephen hahn has proposed a verexec solution for Solaris/OpenSolaris to pick between multiple installed versions to be the default when using the name without a version[00:00:56] <Stric> never bumped into any issues with that so far[00:00:58] <crichardso> what is the best way to see why a zfs vol is being marked as inuse/busy?[00:01:06] <alanc> but it's just a proposal, not yet code[00:01:32] <tsoome> snapshots/clones?[00:01:37] <crichardso> i keep trying to do some zfs destroys and i am getting dataset is busy but i dont have any clones[00:01:52] <Stric> nfs/iscsi/lofi?[00:01:56] <tsoome> ah, with snaps the error message should be different[00:02:06] <crichardso> here hold on[00:03:11] <crichardso> http://pastebin.com/wJTfnbHw[00:03:27] <crichardso> I am using comstar but neither should be used at the moment[00:03:38] <crichardso> espicaly the snap[00:04:38] *** datadigger has joined #opensolaris[00:05:57] <tsoome> you sure you dont have any links up still to it?:D[00:06:12] <datadigger> alanc: I got Twinview working fine now (nvidia). My mistake was in my innocence I previously selected 'separate X-session'. Thanks for the heads-up![00:06:27] <crichardso> tsoome: not the snap[00:07:21] <crichardso> the snap is of a vol that is active but you should be able to delete the snaps with out any issue right[00:07:44] <tsoome> well, i meant are you sure there are no iscsi links up?[00:07:50] <crichardso> i almost think it might be a hung zfs send or something[00:08:40] <tsoome> but then you should have snapshots[00:15:52] <crichardso> i mean the dataset is busy for the destroy on a snap[00:18:20] *** Andrew3 has left #opensolaris[00:23:48] <Stric> seems like someone hasn't understood this dns thing.. :P[00:24:20] *** lattera has joined #opensolaris[00:24:41] *** Reepicheep has quit IRC[00:24:59] *** TomJ has quit IRC[00:25:25] *** TomJ has joined #opensolaris[00:28:44] *** moepmoep has quit IRC[00:29:14] *** andersenep_ has joined #opensolaris[00:29:25] *** TomJ has quit IRC[00:29:51] *** TomJ has joined #opensolaris[00:30:15] *** piwi has quit IRC[00:31:10] <datadigger> crichardso: Maybe some indexer/tracker is recursing the .zfs directory?[00:31:35] *** chilicuil has quit IRC[00:32:44] *** andersenep has quit IRC[00:32:50] *** andersenep_ is now known as andersenep[00:33:02] *** moepmoep has joined #opensolaris[00:34:06] *** comay has quit IRC[00:34:14] *** radsy has joined #opensolaris[00:34:26] <crichardso> datadigger: how would i check that[00:34:31] <richlowe> fuser, probably[00:34:46] *** TomJ- has joined #opensolaris[00:34:47] <richlowe> but by default, .zfs doesn't exist until you cd into it, so unless something already knew it existed, it wouldn't walk down into it[00:34:56] *** TomJ has quit IRC[00:35:08] *** TomJ- has quit IRC[00:36:06] <datadigger> richlowe: .zfs can be set to visible, i guess a process would find and walk into it then.[00:36:50] <datadigger> crichardso: ^^^ That's for you, not for richlowe ;[00:37:43] <crichardso> these are vols tho[00:37:45] *** yarihm has joined #opensolaris[00:38:01] <datadigger> hm[00:39:01] *** psychicist has quit IRC[00:44:51] <Netwolf> any word on 2010.03?[00:44:52] *** relling has quit IRC[00:45:01] *** ewdafa has quit IRC[00:45:31] <JeremyK> Netwolf: topic[00:45:36] *** relling has joined #opensolaris[00:45:51] <Netwolf> JeremyK: ah[00:46:27] <alanc> Yes, the word is "Late"[00:47:01] <JeremyK> as the topic says quite clearly ;)[00:50:38] *** Mdx4 has quit IRC[00:50:40] *** relling has quit IRC[00:51:01] *** andersenep has quit IRC[00:51:42] *** relling has joined #opensolaris[00:52:37] *** pw_thirdfloor has joined #opensolaris[00:53:32] *** pw_thirdfloor has quit IRC[00:55:41] *** relling has quit IRC[00:55:58] *** yarihm has quit IRC[00:56:13] <CIA-21> Krishna Elango <Krishna.Elango at Sun dot COM>: 6895400 SDIO Error Handling Support for RF Platform[00:56:14] *** relling has joined #opensolaris[00:59:15] *** TomJ has joined #opensolaris[01:02:49] *** Beket has quit IRC[01:06:38] *** DesiJat has left #opensolaris[01:07:47] *** kwinz3 has quit IRC[01:08:18] *** Jenkens_ has quit IRC[01:08:19] *** kwinz3 has joined #opensolaris[01:17:55] *** |Inthewings| has quit IRC[01:18:04] *** niq has quit IRC[01:23:44] *** g4lt-mordant has quit IRC[01:26:07] *** radsy has quit IRC[01:26:13] *** Nofer has quit IRC[01:28:32] *** g4lt-mordant has joined #opensolaris[01:30:25] *** Netwolf has quit IRC[01:31:46] *** andersenep has joined #opensolaris[01:34:18] *** datafirm has quit IRC[01:35:59] *** Netwolf has joined #opensolaris[01:38:37] *** aszeszo has quit IRC[01:45:49] *** g4lt-mordant has quit IRC[01:47:14] *** pizdec has quit IRC[01:48:51] *** paddy has quit IRC[01:51:25] <duckinator> hm[01:52:02] <duckinator> alanc: btw, this worked: CC=`which gcc-4.3.2` make .....except now i have to download GMP/MPFR for the cross compiler to work properly :([01:52:05] *** radsy has joined #opensolaris[01:52:23] * duckinator wanders off to figure out what he needs to do to compile those for a gcc build that outputs i386-elf[01:54:19] *** pizdec has joined #opensolaris[01:56:25] <CIA-21> Srikanth Suravajhala <srikanth.suravajhala at oracle dot com>: 6895299 No SAS_phy_stat kstat counters are created after moving cable from one SIM to another[01:56:27] <CIA-21> Jesse Butler <Jesse.Butler at Sun dot COM>: 6889153 Dubious check in pmcs_unattach, 6936793 VTS diskmediatest causes system panic in pmcs_lock_phy, 6855987 pmcs: need to revisit cv_timed_wait() and cv_wait() use, 6937422 deadlock panic during storage poweron[01:57:35] <bball> is it sane to install BFU bits from an older release ?[01:57:57] <bball> for example if I have onnv_134 running and I downloaded BFU bits from a tree based on onnv_100[01:58:18] <bball> is it sane to install those on my onnv_134 system ?[02:02:16] <bball> well.. since nobody from the channel replied I hold the channel responsible if doing this bricks my system ;-)[02:03:50] <jamesd2> bball, should be safe since sun likes backward compatibility and the libraries should be close enough[02:04:12] <alanc> umm, bfu'ing that far backwards is likely to break anything in the other 80% of the OS that depends on new features in the bits you're bfu'ing[02:05:17] <bball> but that could be true just bfu-ing a single release backwards right ?[02:05:44] <bball> I understand I greatly increase my probability of breaking stuff[02:05:58] <bball> since my gap is bigger[02:06:50] *** cybernd has quit IRC[02:08:00] <alanc> a single release is less likely to hit problems, since each consolidation usually builds their packages for build N on either OS build N-1 or N-2, since you can't build on build N until you've built it...[02:08:17] <alanc> but yes, the farther back you go, the more likely to cross flag days[02:08:41] *** dalibor has quit IRC[02:08:47] <duckinator> what's osol's thing for managing networking called?[02:09:11] *** melbogia has quit IRC[02:09:26] <alanc> NWAM >[02:09:28] <alanc> ?[02:09:46] <duckinator> yea, that :P thanks[02:10:14] *** hunter has quit IRC[02:12:05] <duckinator> On NWAM, is there a way (aside from patching it ;P) to get the icon to show the signal strength? (like what NetworkManager does) it shows it on the list of networks, but not the status icon itself[02:12:36] <devians[w]> heya, after a scrub i have this : c9t18d0 ONLINE 0 0 1 43K repaired is this something i should be worried about or can i just clear it[02:13:19] *** melbogia has joined #opensolaris[02:14:18] <johannes> 1 one wrong checksum shouldn#t be a problem ... more a reason to praise zfs[02:14:54] <johannes> as there might be somany harmless reasons[02:14:56] <Andys^> devians: its a small problem. if it keeps happening on the same disk you might want to replace it[02:15:07] <Andys^> (it could also be a cabling problem)[02:15:43] <devians[w]> fair enough. ill keep an eye on the disk and see how it goes.[02:17:58] *** ttys0 has quit IRC[02:19:24] <devians[w]> has anyone else implemented this? http://blogs.sun.com/constantin/entry/new_opensolaris_zfs_auto_scrub[02:19:41] <devians[w]> i have it running on my osol box, not exactly sure if its working however.[02:21:39] <jdoe> what's wrong with a cron job?[02:21:42] * jdoe shrugs.[02:23:32] <devians[w]> nothing really, it just seems nicer to run it as a service[02:23:43] *** doug_ndndn has joined #opensolaris[02:24:05] <devians[w]> ie im looking at running my virtualbox vms as services[02:24:33] * duckinator needs to get a server :([02:26:30] *** ttys0 has joined #opensolaris[02:26:43] <devians[w]> you should duckinator[02:26:48] <devians[w]> storage 3.20T 3.47T[02:26:59] <duckinator> wow, nice..i can't afford servers like that :([02:27:12] <duckinator> my "server" is an os-less 700MHz Pentium 3 with 512MB RAM and a 60GB hard drive[02:27:16] <devians[w]> thats 2 raidz's. 10TB, something like 9TiB, about 7.2TiB after redundancy[02:27:44] <ttys0> duckinator: you could replace that with an iPad[02:27:49] <devians[w]> this is an i7 930, 6gb ram and a very nice hba sas setup.[02:27:51] <duckinator> ttys0: i could replace that with a cellphone[02:28:00] <ttys0> lol[02:28:06] <duckinator> the sad part is i'm not joking :([02:28:18] *** paddy has joined #opensolaris[02:28:46] <ttys0> duckinator: yeah, my phone is within that ballpark ... same RAM, faster proc, half the storage[02:28:52] <duckinator> after i get a new hard drive, i'm gonna put opensolaris on here (it's only ony my laptop right now) :)[02:29:13] <duckinator> this was a 3.06GHz Pentium 4 w/ 768MB RAM and a 160GB IDE hard drive, i got 2 gigs now..much nicer[02:30:32] *** Triskelios has joined #opensolaris[02:31:24] *** doug_ndndn has quit IRC[02:35:05] *** ttys0_ has joined #opensolaris[02:35:45] *** achester has joined #opensolaris[02:37:03] *** ttys0 has quit IRC[02:37:08] *** ttys0_ is now known as ttys0[02:41:09] *** wdp_ has joined #opensolaris[02:43:10] *** yippi has quit IRC[02:43:59] *** yippi has joined #opensolaris[02:44:48] *** wdp has quit IRC[02:45:22] *** doug_ndndn has joined #opensolaris[02:46:20] *** RoyK^ has quit IRC[02:47:06] *** yippi has quit IRC[02:47:36] *** yippi has joined #opensolaris[02:52:10] *** achester has quit IRC[02:53:47] *** MGSsancho has joined #opensolaris[02:54:46] *** bacon000 has joined #opensolaris[02:54:53] <MGSsancho> services are slow today ~_~[02:57:21] *** bubbva has quit IRC[02:59:37] *** Odin- has quit IRC[03:01:59] *** ttys0 has quit IRC[03:02:32] <jamesd2> MGSsancho, tip them better next time[03:02:52] <MGSsancho> I plugged in a hdd on a sata port, is formrt supposed to pick it up right away or is ther another way to tell the system i added an internal device?[03:03:58] <MGSsancho> jamesd2: i offered services chips and salsa but it dint respond =([03:04:55] *** nitrile has joined #opensolaris[03:05:12] *** ahmed-tux has joined #opensolaris[03:05:17] <ahmed-tux> hi[03:05:43] <ahmed-tux> there's no prob if i install opensolarisin my external hdd[03:06:00] <MGSsancho> hai =)[03:06:08] <MGSsancho> it will be sloww ~~~[03:06:22] <ahmed-tux> MGSsancho: ogenki desu ka[03:06:37] *** tosh has quit IRC[03:07:12] <MGSsancho> genki desu ~[03:07:25] <ahmed-tux> MGSsancho: judo wo shimasuka[03:07:54] *** nitrile_ has quit IRC[03:08:50] <MGSsancho> yeah thats the limit of my japanesse.. lol[03:08:58] <MGSsancho> other than stuff ive learned from h-games[03:09:42] *** nachox has joined #opensolaris[03:11:12] <MGSsancho> mr nacho~, maybe you can jelp me out, how do I know where a sata drive is I just plugged onto the mobo? format disapoints =([03:13:26] *** skeeziks has quit IRC[03:14:37] *** fleyta has joined #opensolaris[03:14:53] <fleyta> jmcp: Are you available?[03:15:55] <nachox> MGSsancho, what does cfgadm say?[03:17:42] *** lcheung has joined #opensolaris[03:17:56] <fleyta> ircfind -name \* -a -org \( \*.sun.com -o \*.oracle.com \) -a -attribute on_sponsor[03:18:29] <MGSsancho> http://pastie.org/904818[03:18:43] <fleyta> Is Shawn Walker here?[03:19:12] <MGSsancho> yeah go ahead and LOL when you see how i misconfig my zpool arbalest :D[03:19:26] <Andys^> MGSsancho: I assume you plugged a drive into sata1/0.1 ? in which case, try: cfgadm -c configure sata1/0.1[03:19:31] <Andys^> & cfgadm -c connect sata1/0.1[03:19:47] <Andys^> oh, maybe it was sata0/5...[03:20:02] *** falconn1 has joined #opensolaris[03:20:02] <Andys^> sata0/5 says connected & unconfigured, so needs a cfgadm -c configure run on it[03:20:30] <fleyta> Hell, I forgot that this channel is empty and no one can hear me.[03:20:54] <nachox> that would be correct[03:21:06] <MGSsancho> fleyta: yup lol[03:21:19] *** bacon000 has quit IRC[03:21:40] <MGSsancho> Andys^: thanks, ill just make a new pool and move everything over >.<[03:22:13] <Triskelios> I don't recall if I've ever seen Shawn on IRC[03:22:19] <alanc> fleyta: Shawn doesn't usually come to this channel[03:22:33] <alanc> Triskelios: I have, quite a lot, just not here[03:22:58] <fleyta> alanc: Which channel do I have to join?[03:23:48] <Triskelios> #pkg5 maybe?[03:23:57] <alanc> during the daytime, he's usually in #pkg5 with the other IPS developers, but he's a couple timezones east, so it's late in the evening for him now[03:24:14] <alanc> (not as late as it is for you of course)[03:25:39] <Triskelios> did someone express interest in replacing intrd? I've made it my fun project for the week...[03:25:43] <fleyta> alanc: This is the only time I can rest in front of the laptop. On daytime our little lady wants http://www.youtube.com/user/MuppetsStudio[03:26:32] <alanc> Triskelios: I think jbk or sstallion_ had talked about it[03:26:43] <alanc> fleyta: I can understand[03:28:07] <alanc> Triskelios: ah, sstallion_ for sure, given http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6898208[03:28:54] <alanc> apparently he nuked the linked webrev from cr.os.o though[03:28:56] <devians[w]> heya, im fiddling with a service manifest of mine, and i initially made a typo in the path to the method script. i've fixed the typo in the manifest file but it's still referring to the old one when i check the log to see why its hitting maintenance mode. I've run svcadm refresh but that doesnt seem to fix it?[03:29:14] <Triskelios> ah.. yeah, thought it was him[03:29:41] *** falconn1 has quit IRC[03:30:02] <alanc> maybe if we say sstallion_work a few more times it will beep enough to wake him up 8-)[03:30:39] <ottom> devians[w]: did you 'svccfg import' your updated manifest?[03:31:28] *** chendy has joined #opensolaris[03:31:49] *** Macer has quit IRC[03:33:35] <devians[w]> ottom, ah, no i hadnt. doing so now[03:34:23] <nachox> alanc, kind of like candyman[03:35:05] <devians[w]> im using a slightly modified version of sstallions virtualbox headless. now in the log for my vm its not adding anything when i run enable, but the vm isnt started... hrm[03:35:15] <devians[w]> sorry, the service log[03:35:23] <MGSsancho> candyman candyman candyman[03:37:30] <devians[w]> ah, because it was in maintenance.[03:37:52] <devians[w]> ottom, importing the manifest again doesnt seem to have helped.[03:39:01] *** nikolam has quit IRC[03:39:12] *** fleyta has quit IRC[03:39:50] *** rlw has joined #opensolaris[03:40:13] <ottom> do a 'svccfg export' to check that it accepted the new manifest, then 'svcadm refresh <yourservice>' to put it into effect. Then if the service is still in maintenance do 'svcadm clear <yourservice>' and I'd probably finish with a 'scvcadm restart <yourservice>' just to be usre[03:40:30] <ottom> s/usre/sure/[03:40:46] *** _madmax_ has quit IRC[03:41:10] *** Shoggoth has joined #opensolaris[03:41:28] <Shoggoth> morning all![03:42:55] <devians[w]> ottom, if i do a svccfg -s myservice listprop the start/exec prop is still the same one from the old manifest.[03:43:18] *** rlw has quit IRC[03:43:45] <devians[w]> if i export it, its still the old one[03:46:10] <devians[w]> just deleted and reimported, seemed to help...[03:48:46] <MGSsancho> sun needs to sell sata cables[03:48:49] <MGSsancho> cheap ass cables are lose and causing problems -.-[03:50:06] <Triskelios> devians[w]: a refresh will always fixed that...[03:51:06] <devians[w]> Triskelios, didnt seem to, strangely enough[03:51:38] <devians[w]> I think i have it sorted now though.[03:52:08] <devians[w]> heres the next fun question. how does solaris handle services and which to enable on boot?[03:53:06] *** nikolam has joined #opensolaris[03:54:43] *** JDuke128 has joined #opensolaris[03:55:03] <JDuke128> hi , opensolaris arm port is good on performance ? and it has something like directfb ?[03:56:07] *** ganbold_ has joined #opensolaris[03:56:12] <Triskelios> JDuke128: unclear, and probably not. it's at a fairly early stage[03:56:34] *** lattera has quit IRC[03:56:40] <JDuke128> opensolaris has directfb like ?[03:56:55] <JDuke128> i need something native drawing directly on framebuffer[03:57:11] <alanc> opensolaris has X[03:57:30] <nachox> MGSsancho, sun is no more[03:57:31] <alanc> there are SPARC framebuffer drivers, but none for x86[03:57:41] <alanc> and no idea at all if ARM has any, or even boots[03:58:33] <alanc> but even on SPARC, X is the only supported interface to drawing on the screen[03:58:56] *** ganbold has quit IRC[03:59:24] <MGSsancho> nachox: oracle can lick my nuts. i demand sun back![03:59:34] <MGSsancho> ;_; oh well. maybe i need to join a 12 step program[03:59:38] <nachox> watch the language please[04:00:02] <alanc> are those nuts salted and roasted?[04:00:23] <alanc> or were you not talking about the ones that grow on trees ?[04:00:40] <MGSsancho> *I am displeased with Oracle's handling of Solaris Licencing and general lack of love tword Opensolaris[04:01:05] <MGSsancho> is that batter nachos?[04:02:14] *** freakazoid0223 has quit IRC[04:02:38] <nachox> yes, that is better, but useless. there is noone in this channel that can change that[04:02:52] <Andys^> i dont know what oracle's stance is yet... its the Radio Silence that is killing us[04:02:54] *** mroconnor has joined #opensolaris[04:03:35] <MGSsancho> only think i heard is solaris is free for 90 days[04:04:19] <MGSsancho> and some stuff will never make it to opensolaris and stay on solaris. i get the vibe that osol will be abondened but that is my personal opinion[04:04:52] <JDuke128> opensolaris and solaris are different things[04:04:55] <Shoggoth> what's the state-of-play wrt to text-based install?[04:05:15] <JDuke128> opensolaris and solaris never dies easy[04:05:41] <JDuke128> and opensolaris is becoming more popular every day[04:05:50] <JDuke128> 1 day it will be alternative against linux[04:05:50] <alanc> MGSsancho: Sun's policy for the entire existence of the project was that some stuff would never go from Solaris into OpenSolaris - originally for licensing issues, then later for added features like the fishworks management/dtrace UI's[04:05:53] <nachox> MGSsancho, but that second part was true even when sun was in charge. you'll never see the fishworks gui in opensolaris for example[04:06:08] <Triskelios> Shoggoth: there's a long list of bugs but it's mostly usable[04:06:21] *** kwinz3_ has joined #opensolaris[04:06:21] <Shoggoth> I've forgotten how to start it[04:06:23] <Shoggoth> lol[04:06:23] <JDuke128> if u say , solaris die , look sparc 128 gflob cpu that only runs on solaris[04:06:29] <MGSsancho> it used to be osol would be the place to dev the latest and greatest stuff. make it avaliable for people to help break and look for bugs. allow for the community to request features etc... now i have no idea[04:06:53] <JDuke128> opensolaris arm port is on wrong way only[04:06:56] <JDuke128> it needs some fix[04:08:25] <nachox> MGSsancho, oracle has a dury to their share holders, they need to make money from their IP. i am not against them keeping the value add behing a lock[04:08:29] <JDuke128> well is someone have any idea ? is there anything that 2 computer runs 1 OS ?[04:08:29] <MGSsancho> great just great. zpool import is taking longer than 8 mins weeee[04:08:37] <MGSsancho> alanc: nachox: thanks for the clearing up =)[04:09:20] <wrapster> hi[04:09:43] <wrapster> can anyone explain how IPS works on the latest onnv-builds in creating pkgs?[04:09:55] <Triskelios> nachox: if this hurts adoption, on the long term they won't have any customers left[04:10:02] <alanc> MGSsancho: it's still the development branch - no one has told us to stop putting new features into the nevada trees for testing before they get backported to S10 (with a few exceptions that don't apply any more, like Live Upgrade changes for S10 that aren't in Nevada since the IPS conversion)[04:10:15] <wrapster> I've been trying to ready through the docs but nothing conclusive[04:10:37] <richlowe> ask more specific questions.[04:10:38] *** kwinz3 has quit IRC[04:11:16] <devians[w]> am i right in saying that while you cant remove a vdev from a pool (ie a raidz) you could replace a vdev?[04:11:34] <nachox> Triskelios, i dont think that'll be a problem. it hasnt been a problem with fishworks[04:11:55] <wrapster> richlowe: talking to me?[04:12:16] <MGSsancho> nachox, alanc, I think this clears things up then, for production go for sol10 and pay for support, osol is a workstation OS that is for most dev. that means that osol's inards with be a great indicitator for what sol11 will eventully look like (yes i know the name in unknown and no one has even mentioned a sol11 lol)[04:12:24] <Andys^> devians[w]: you can replace the individual components in a vdev, all of them, if you like[04:12:52] <devians[w]> is there a way to do that without putting the raidz in a degraded state?[04:13:44] <Andys^> no[04:13:45] <Andys^> oh[04:13:46] <Andys^> yes[04:13:54] <alanc> wrapster: you run nightly, it creates an IPS repo with packages in - as richlowe said, what specifically don't you understand from the published docs?[04:13:59] <Andys^> you can replace a device by adding the new one and asking zfs to replace the old one with the new one[04:14:14] <Andys^> zpool replace [-f] pool device [new_device][04:14:24] <devians[w]> Andys^, but that degrades the raidz while it resilvers...[04:14:29] <Andys^> no[04:14:34] <Andys^> if you have a spare disk slot[04:14:37] <devians[w]> really?[04:14:40] <Andys^> you can add a new disk without removing the old disk[04:14:46] <devians[w]> mmmm.[04:14:49] <Andys^> its optional. hence the square brackets: [new_device][04:14:55] <devians[w]> i need to add a hot spare to this system.[04:15:49] <devians[w]> i wonder when/if we'll ever get the ability to remove vdevs from a pool. 'whups, that raidz is no longer needed. zpool evacuate'[04:16:19] <Andys^> maybe one day[04:16:29] <devians[w]> i spose theres not any more, with the oracle thing, but was there a roadmap for zfs features?[04:16:55] <Andys^> i think crypto is the next big feature[04:17:34] <wrapster> alanc: Earlier I was familiar with the svr4 packaging method , so Im used to seeing all the pkgs present in the pkgdefs dir.. Now its gone..on the flip side i see that each pkg now has a manifest file which is used to build the pkgs.. however where do all the necessary data required come from? Secondly earlier the presence of prototype_com files acted as the source telling where each binary/dir/files needs to be placed ,who is doing that job now?[04:17:46] <devians[w]> that would be kind of cool i spose. i think more people are interested in better tools for device management / migration.[04:18:34] <Andys^> devians[w]: there's a feature yet to be implemented which is rather complicated, which involves rewriting data. it would allow transformations like defragmenting, changing from raidz1 to raidz2, removing vdevs, that sort of thing.[04:18:38] <alanc> wrapster: the package manifests (the *.mf files) are the replacements for prototype_com[04:18:54] <wrapster> alanc: i understand that a local repo is created.. and that is used to build pkgs.. but what does the local repo contain... is it only the binaries(as in proto dir) or does it contain anything else.[04:19:09] <devians[w]> Andys^, ah yes, was there ever a status on that more than 'hey we should probably do this sometime'[04:19:15] <alanc> the repo contains the binaries and all the metadata/catalogs/etc. needed to install them[04:19:26] <wrapster> alanc: ok so how does IPS know what are the files provided by each pkg? if that be the case and where to put it in the installed machine?[04:19:42] <Andys^> devians: not afaik[04:19:51] <alanc> wrapster: from the package manifests[04:20:00] <wrapster> alanc: hmm..[04:20:31] <alanc> I can say "the package manifests provide the list of files to install for the package and where to install them" if saying it three times helps[04:20:43] <[lewellyn]> alanc: no, it just makes it true[04:21:57] <Triskelios> sup [lewellyn][04:22:33] <[lewellyn]> nammuch. busy busy busy. just popped onto the computer again to see if my jucr specs were validated/voted[04:22:37] <alanc> for a concrete example, usr/src/pkg/manifests/system-library.mf has a line that says "file path=lib/libc.so.1 reboot-needed=true" that tells IPS for that package to install the libc file in /lib/libc.so.1, and that a reboot is needed when you do so[04:23:48] <alanc> if you're making the mistake of looking in any SUNW*.mf, that would explain the confusion, as all but 3 or 4 SUNW* packages are now just pointers to the new package names, and don't install any files[04:23:54] <[lewellyn]> and of course they're not...[04:23:59] <wrapster> alanc: thanks a ton for explaining.. i was just about to ask that.[04:23:59] * [lewellyn] wanders back off for now[04:24:48] <wrapster> alanc: yes your second point is indeed valid.. thats exactly what i was looking at.[04:25:18] <alanc> http://hub.opensolaris.org/bin/view/Project+indiana/Renamed+Packages+in+Build+133[04:25:54] <wrapster> alanc: so to sum it all up.. all the SUNW*.mf file points to one or the other system-*.mf file which actually tells the IPS to install a particular binary to a specific location.[04:25:58] <wrapster> alanc: am I right.[04:26:52] *** Mousk4r has quit IRC[04:26:53] <alanc> to another *.mf - most aren't named system-*.mf[04:27:08] <alanc> otherwise, yes[04:27:27] <[lewellyn]> Triskelios: btw, are you able to vote on jucr? ;)[04:27:33] <[lewellyn]> gxmessage needs another +1[04:27:54] <[lewellyn]> else, i get to file a bug that solaris doesn't have xmessage, and alanc will kill me :)[04:28:09] <alanc> didn't you already file that one?[04:28:20] <Triskelios> no, I'm not even properly affiliated with sw-porters[04:28:25] <[lewellyn]> i don't think so, since i decided that i'd try to get gxmessage into jucr at some point[04:28:40] <wrapster> alanc: ok got it.. so if i need to find out which is the actual file that does the placing of the binaries/files/dirs/ I need to first look at the SUNW<pkg>.mf look at the FMRI and follow it up in the manifest dir to find the details, right?[04:29:00] <[lewellyn]> gxmessage (in pending) installs links as xmessage and gmessage (gmessage is for debian-compat)[04:29:23] <alanc> only if you have the old SUNW* name memorized - they're obsolete now[04:29:23] * [lewellyn] looks to see if there's an xmessage bug[04:29:48] <wrapster> alanc: ok[04:30:04] <alanc> if you're running build 133 or later, pkg search shows the new name, and grep *.mf would find the package with the file in[04:30:33] <richlowe> in ON, by convention, the manifest file name is the package name with every / replaced with a -[04:31:14] <[lewellyn]> alanc: i'm not seeing any bugs about solaris not having xmessage :)[04:31:15] *** doug_ndndn has quit IRC[04:31:28] <alanc> oh, then maybe I'm just remembering you harassing me about it[04:31:32] <wrapster> ok thanks guys.. im running 134 will start a nightly and come back if i more queries...[04:31:34] <wrapster> thanks a lot[04:31:49] *** Disorganized_ has joined #opensolaris[04:31:51] <[lewellyn]> probably. harrassment training is effective :)[04:32:07] *** Disorganized_ has left #opensolaris[04:32:28] *** sivanov has joined #opensolaris[04:32:37] <alanc> heh, a search of the internal bug db turned up 1265131 xmessage gets occasional "sequence lost" messages, fails to appear against Solaris 2.6[04:32:55] <alanc> but that was about Xlib breaking it[04:33:19] <richlowe> but, but, *xmessage*?[04:33:20] <oninoshiko> [lewellyn]: got a link, I'll try it now[04:33:29] <[lewellyn]> oninoshiko: to gxmessage?[04:33:31] <richlowe> you should use zenity, then use it as an excuse to yell at Gman__[04:33:52] <oninoshiko> [lewellyn]: yes[04:34:08] <[lewellyn]> http://jucr.opensolaris.org/review/packages/7491/ http://jucr.opensolaris.org/pending/p5i/0/gxmessage at 2 dot 12.4,5.11-0.111:20100330T090500Z.p5i[04:34:23] <[lewellyn]> richlowe: xmessage is cross-platform[04:34:29] <alanc> does Gman__ even use zenity any more? I thought he just wrote python with pygtk...[04:35:00] <[lewellyn]> note that i went through the trouble to make it compatible for people moving from debian, even ;)[04:35:25] <[lewellyn]> (the trouble being a symlink, but i had to hunt for whether any linux distros renamed it!)[04:35:29] <alanc> gxmessage would be easier to integrate anyway - just get JDS to pull the .spec file into their build[04:35:41] <oninoshiko> i wouldn't even know to look for that[04:35:45] <[lewellyn]> alanc: that is kind of my next step, once it ends up in contrib :)[04:35:56] <richlowe> alanc: He never does that when yelling at him would be appropriate, though.[04:35:58] <alanc> plus, less horrible accessibility review, because Xaw is not accessible[04:36:17] <[lewellyn]> and far less jarring when it pops up than xmessage :P[04:37:44] <[lewellyn]> also, didn't /. say a few months ago that you guys don't care about accessibility anymore? ;)[04:38:23] <richlowe> Think this is the first I've actually looked at jucr[04:38:25] *** achester has joined #opensolaris[04:38:34] <[lewellyn]> richlowe: it needs love :([04:39:14] <[lewellyn]> note that it eveen has simple flaws like someeone choosing GMT instead of UTC, so now it's in BST[04:39:23] <alanc> [lewellyn]: like anyone believes slashdot[04:39:40] <richlowe> *real* men have the offset to the UK memorized![04:39:46] <[lewellyn]> alanc: yeah, and as if el reg is eveer read ;)[04:40:08] <oninoshiko> i read elReg all the time... I need a laugh.[04:40:09] <[lewellyn]> richlowe: the flaw, of course, isn't the offset. it's that builds have summer time ;)[04:40:46] <alanc> Oracle cares about selling products to customers. Quite a few governments (US, EU and a bunch more) require accessibility for their purchases, and for large corporations (government contractors, businesses covered by laws like the ADA, etc.)[04:41:01] <alanc> therefore, Oracle cares about accessibility[04:41:18] <[lewellyn]> yeah. *i* know this[04:41:55] <[lewellyn]> anyhow... &[04:42:31] <oninoshiko> you mean Oracle isn't some evil cult hell-bent on knocking over people in wheel-chairs?[04:42:35] <[lewellyn]> oninoshiko: feel free to leave a comment. and if you can validate, cc65 still needs validation ;) http://jucr.opensolaris.org/review/packages/7495/[04:42:47] <richlowe> alanc: Quite a clear statement from someone without the figleaf of a disclaimer[04:42:51] <alanc> oninoshiko: only if the people are using the wheelchairs to cart away larry's money[04:42:52] <richlowe> alanc: don't have it in the kill-ring anymore? :)[04:42:56] <[lewellyn]> richlowe: it's common sense[04:42:59] <oninoshiko> [lewellyn]: it looks good to me[04:43:17] *** MGSsancho has quit IRC[04:43:31] <alanc> richlowe: no, it's been too long since I pasted it - besides, Oracle says as much on it's web page[04:43:45] *** skullone has quit IRC[04:44:21] <[lewellyn]> oninoshiko: comment and/or vote. people won't know it's good if there's no feedback on it :)[04:44:45] <alanc> I also know exactly what story [lewellyn] is referring to, and the underlying circumstances, but it would be inappropriate for me to comment in public on the specific personnel & project funding issues.[04:44:57] <oninoshiko> [lewellyn]: im getting to the voteing give me a min[04:45:11] <[lewellyn]> oninoshiko: thanks :D[04:45:18] *** doug_ndndn has joined #opensolaris[04:45:31] <[lewellyn]> alanc: i wasn't trying to put you on the spot or get you to say things you shouldn't :)[04:45:44] <[lewellyn]> just poking fun at the lack of common sense in places like /.[04:45:51] <alanc> I could point at the recent followup from the original author of the open letter: http://blog.grain-of-salt.com/index.php?itemid=395 which has gotten far less press, since it's not as juicy a story[04:46:22] <alanc> "While Oracle has yet to officially acknowledge my Open Letter, it seems they are doing the more important thing, namely starting to respond to the concerns raised therein: They sent Li Yuan and Ke Wang, two engineers from Oracle/Sun Beijing, to the hackfest."[04:46:36] *** Funk-E has joined #opensolaris[04:47:06] <oninoshiko> alanc: larry doesn't scare me, now Richard Brason, that guy is a walk james bond super-villain cliche[04:47:44] <alanc> Branson has space planes even - much harder to escape than larry's sailboat or yacht[04:48:00] *** freakazoid0223 has joined #opensolaris[04:48:10] <[lewellyn]> larry knows good business. most of the current conspiracy theories don't take into account that they'd be bad business :P[04:48:15] <oninoshiko> indeed <.< >.>[04:48:33] <oninoshiko> [lewellyn]: that's why I mock them.[04:48:37] <Andys^> [lewellyn]: armchair unix enthusiasts theories about what makes good or bad business can easily be wrong, though[04:49:14] <[lewellyn]> Andys^: killing the products you pitched to investors, and killing your ability to sell to governments? those are both big no-nos[04:50:11] <[lewellyn]> you at least wait till there's a 10Q with those product names before killing them ;)[04:51:29] <alanc> or you confuse everyone by renaming the products first, then killing them....[04:51:54] <Andys^> [lewellyn]: right but its silly to expect everything to stay the way it was, there must be change coming[04:51:57] <alanc> "Sun Solaris"? no, we never had that product, only "Oracle Solaris"[04:52:25] <[lewellyn]> Andys^: change, yes. but the degree of change suggested by current conspiracy theories? HAH[04:52:45] <[lewellyn]> i'm sure you've heard that oracle wants to kill solaris and sparc[04:53:23] <[lewellyn]> you don't say repeatedly that you're hiring more people to work on products you're about to kill, as a rule :P[04:53:28] *** brandon has joined #opensolaris[04:53:52] <richlowe> you could do that with naming sleight of hand, too. :)[04:54:14] <[lewellyn]> there's not been enough for that to be a likely outcome either ;)[04:54:31] <richlowe> [lewellyn]: See, they're already destroying the fine traditions![04:54:52] <richlowe> soon enough, customers will know the current name of the products they want to buy![04:54:52] <[lewellyn]> no no, they're keeping the best ones![04:54:55] <richlowe> Then where will everyone be?[04:54:57] <[lewellyn]> it's still impossible to buy things![04:55:00] <taemun> can you have more than one zpool replace running on one vdev at a time?[04:55:39] <[lewellyn]> taemun: yes. but... you don't want enough to lose redundancy[04:55:59] <taemun> well if you had another full controller, you could replace all eight drives with eight new drives[04:56:04] <oninoshiko> taemun: yes, provided you have enough redudency to still be able to actually build the data[04:56:05] <CIA-21> Mark Shellenbaum <Mark.Shellenbaum at Sun dot COM>: 6939941 problem with moving files in zfs[04:56:07] <taemun> (for changing drive size purposes)[04:56:07] <[lewellyn]> you don't want to do that :P[04:56:12] <taemun> :P[04:56:14] <taemun> ok[04:56:32] <[lewellyn]> do a drive at a time. :)[04:56:44] <richlowe> You want to leave enough redundancy not only for the rebuild, but for any possible *further* failure.[04:56:54] <[lewellyn]> two if it's raidz2 and you feel ballsy. three for raidz3 and you're feeling really ballsy[04:57:08] <taemun> err[04:57:21] <taemun> zpool replace doesn't need to shoot the parity drives[04:57:23] <taemun> you can do it hot[04:57:26] <taemun> if you have enough sata ports[04:57:37] <[lewellyn]> i certainly wouldn't want to be the one explaining why the raidz3 died ;)[04:57:59] <[lewellyn]> taemun: i've not needed to do it yet *knocks on wood* but i was under the impression that it takes the drive offline to replace it[04:58:05] <Andys^> hmmm[04:58:05] <taemun> nope[04:58:10] <taemun> I did a hot replace the other day[04:58:13] <taemun> just of a single drive[04:58:17] <[lewellyn]> anyhow, gf is home! back later![04:58:21] <taemun> hf[04:58:27] <Andys^> it appears some of them SHA-3 candidates can make use of the new AES-NI instruction set in Intel CPUs. that'd be nice for ZFS SHA speed down the track[04:59:40] <oninoshiko> [lewellyn]: it doesn't have to, but most people have used all the ports on their card, so most don't have it as an option (although I haven't tested to see what happens in the case you are talking about if a drive/block fails, so I'm only 90% or so sure)[04:59:49] <oninoshiko> Andys^: nice.[05:00:51] <oninoshiko> someone made the /topic less like something I would expect out of elReg's comment section.[05:01:24] *** gm152 has quit IRC[05:01:42] *** Kraln has left #opensolaris[05:03:13] <brandon> I'm visiting the old -lsocket -lnsl linking problems once again, and somehow once again I'm stumped - the libs seem to be there: http://pastebin.com/xSQnXQqp[05:06:44] <richlowe> You know you don't actually have -lsocket on that gcc commandline?[05:08:59] <Gman__> alanc, richlowe: no longer develop on zenity, and it wasn't much use for some of the things i've wanted to do recently :)[05:09:25] <oninoshiko> [lewellyn]: I'd be happier if the upstream maintainer included his email address. he's go it on his page, so it's not exactly a secret.[05:09:28] <Gman__> there's only so many RFEs for 'add yes/no dialog' you can take...[05:09:58] <oninoshiko> upstream maintainer field[05:10:02] <brandon> richlowe, ah sorry - i removed it for testing. When it is there, it cannot be found[05:10:07] <oninoshiko> that's the only complaint I have[05:11:44] <alanc> if libsocket is missing, your opensolaris system should be unbootable and unusable[05:11:54] *** swift has joined #opensolaris[05:12:39] <brandon> alanc, i see - well it is there, but somehow the linker is not finding it in that command[05:13:38] <alanc> then I can only guess that your local build of gcc is broken[05:13:53] <alanc> or if you're using a local gnu ld with it, perhaps it[05:15:01] <brandon> alanc - i suppose that is a fair assessment, anything jumping out at you from my gcc -v output? I am using /usr/ccs/bin/ld[05:18:05] <Triskelios> is pkg-na-2 down? doesn't seem to be reachable anymore[05:20:11] <brandon> alanc, I'm not sure why LD_LIBRARY_PATH seems to fix things, I would have thought the -L/usr/lib/64 -L/lib/64 would be enough[05:21:23] <Triskelios> brandon: LD_LIBRARY_PATH normally is reserved for the run-time linker only...[05:21:51] <Triskelios> brandon: gcc -v should show how it's invoking ld[05:23:08] *** Calibellus has quit IRC[05:25:34] <nachox> Gman__, are you coding these days?[05:25:52] <Gman__> nachox: not much, but thinking i should a little more[05:26:09] * Gman__ plans on spending a couple of hours doing the 'secure coding' training this arvo....[05:26:44] <brandon> Triskelios, actually, it doesn't fix the issue - false alarm...[05:27:59] <nachox> Gman__, so you're still doing marketing for oracle then? :P[05:28:35] <Gman__> nachox: product management, supposedly[05:28:43] <Gman__> so technically in engineering....[05:30:46] *** Gman__ is now known as Gman[05:30:47] *** ChanServ sets mode: +o Gman[05:31:38] <Triskelios> brandon: add -v during the linking...[05:31:52] *** Mousk4r has joined #opensolaris[05:32:12] *** Hamc has joined #opensolaris[05:32:23] *** gothos has quit IRC[05:32:27] <Triskelios> brandon: ld has a special case for /usr/lib/%{arch64} afaik... so -L shouldn't even be needed[05:34:45] *** CodeWar has joined #opensolaris[05:35:38] *** Netwolf has quit IRC[05:36:00] *** gothos has joined #opensolaris[05:36:01] <nachox> anyway, i need some sleep, night all[05:39:10] *** swift has quit IRC[05:39:22] <brandon> Triskelios, very nice, I didn't know that -v could be used that way, I posted my output here ... still looking it over: http://pastebin.com/2H1cmF3m[05:39:47] *** nachox has quit IRC[05:40:26] <CodeWar> is there a way to rdp into my opensolaris desktop[05:42:28] <lblume> You can enable desktop sharing.[05:43:14] <CodeWar> lblume, you mean System -> Desktop sharing ... that is terribly slow[05:43:40] <lblume> Is it? I never gave it more than a cursory glance.[05:44:24] <CodeWar> a good toy to write about in release notes but heck not usable .. i think its vnc at the end of the day[05:45:18] <Triskelios> you can't do much better without a dedicated X server...[05:46:02] *** swift has joined #opensolaris[05:46:09] <CodeWar> on my setup I was amazed at how well the rdp built into vbox is . its fast does autoscaling alas I need to run osol in vbox for that[05:46:26] *** Mousk4r has quit IRC[05:46:42] <lblume> Really? Every time I tried rdp top connect to a vbox osol, it was a real pain.[05:46:55] <lblume> IT worked fine with XP though.[05:47:30] <CodeWar> hmm I must have spoken too early then on my osol host I have 3 guests win7/ubuntu/centos all of which work very well with rdp[05:48:19] <brandon> anyone have a preferred (recent) gcc config for amd64? think I'll build it again and give it a try[05:50:21] <brandon> ah, seems gcc's site has some configs for solaris recorded[05:53:34] <Triskelios> 4.3 won't do?[05:54:24] <brandon> Triskelios, I'll give it a go - will have to wait a few minutes and see how it goes[05:54:56] <lblume> What's wrong with the gcc available in the repo?[05:57:42] *** Macer has joined #opensolaris[05:58:10] <Macer> ugh. had a power outage here[05:58:18] <Macer> remind me to get a small generator :)[05:59:00] <devians[w]> CodeWar, vbox is quite nice for rdp[05:59:24] <devians[w]> i just finished setting my headless install as a service in osol. vnice[06:01:34] *** Mousk4r has joined #opensolaris[06:01:52] <brandon> lblume, nothing in particular - It has been so long since I built gcc 4.4 I can't even remember why I did[06:02:27] <CodeWar> for whatever reason I coudlnt event install osol 2009.06 in vbox on osol 2009.06 host. the installation was crazy freaking slow .. picked up /dev instead that works fine but sucks over rdp[06:03:23] <lblume> Hmmm, I reinstalled 2009.06 last week in vbox, worked ok. I had slightly tweaked the configuration though.[06:04:46] <CodeWar> but Solaris10 takes home the prize... just to get the desktop up it takes 5 frigging minutes and after that no clue what it does but consumes 80% of 6 of my cores . DOING NOTHING![06:06:56] <lblume> Interesting. Last time I installed S10, on an S10 host, it was also workable.[06:07:47] *** devians has quit IRC[06:10:01] <alanc> certainly didn't seem to take that long to bring up the desktop on the s10u9_03 build I installed in vbox a couple weeks ago[06:10:20] <alanc> (running on a opensolaris 134 host)[06:10:34] <CodeWar> vbox 3.1.6?[06:11:19] <sivanov> when solaris ?/10 ?[06:11:40] <sivanov> perhaps somebody know[06:11:56] <alanc> Solaris 10 came out in 2005[06:12:33] <alanc> if that's not what you're asking, try asking more clearly, since what you said didn't make sense, so I had to guess[06:12:34] <sivanov> i mean next to 10 05/09[06:12:43] *** doug_ndndn has quit IRC[06:13:12] <sivanov> update that is[06:13:12] <CodeWar> I might have found out something .. looks like the OSOL 134 guest desktop experience over rdp imrpoves a heck lot if I disable 3d acceleration tab in vbox (whatever that does) and reduced gfx memory available to 64MB[06:13:32] *** Norrin has joined #opensolaris[06:13:35] <alanc> the release after Solaris 10 5/09 is Solaris 10 10/09[06:13:53] <sivanov> oh[06:13:59] <sivanov> an in this year?[06:14:37] <alanc> the release after Solaris 10 10/09 is not yet announced, and Oracle generally does not disclose product schedules of future product releases[06:15:24] <sivanov> hmm it looks like 10/09 has zpool v15[06:15:59] <CodeWar> well keyboard doesnt work now .. darn anyways this whole vbox business is so I can rdp into my osol session ..[06:16:03] *** devians has joined #opensolaris[06:16:19] <CodeWar> anybody of a way to enable rdp w/o virtualization .. if theres a paid product I m willing to try it too[06:16:31] <lblume> CodeWar: That makes sense, as RDP isn't tightly integrated in the OS as it is for Windows, the more video memory you have, the more it has to scan, I'd wagger.[06:17:08] <Andys^> CodeWar: X11 protocol - you may have heard of it?[06:17:26] <sivanov> and my zpool is v19 under 127[06:17:27] <CodeWar> Andys^, oh no dont take me there .. no X11 no VNC call me biased if you want[06:17:35] <Andys^> you're biased[06:17:42] <Andys^> why would you try to use RDP over X to a unix machine?![06:17:42] <Andys^> :)[06:17:43] <CodeWar> make it biased bigot[06:17:50] <lblume> CodeWar: He was joking. Nobody seriously does X11 remote sessions nowadays.[06:17:55] <Andys^> ?!?!?[06:18:09] <Andys^> it works, and it works efficiently..[06:18:12] <Andys^> unlike VNC[06:18:22] <CodeWar> how lucky you are[06:18:29] <Andys^> my mind's boggling that you want RDP instead of X?[06:18:29] <lblume> X11. Efficient. Sure, try it over ADSL.[06:18:31] <Andys^> please explain[06:18:42] <Andys^> lblume: ssh -C(ompression) flag, you may have heard of it[06:18:53] <lblume> I have. And it still sucks.[06:19:06] <Andys^> running, say, an xterm works basically as good as using a plain ssh session[06:19:23] <lblume> Andys^: He's talking about a FULL SESSION, not a single app.[06:19:28] <Andys^> running firefox with a graphics/flash intensive site, not so good[06:19:31] <sivanov> thereis xrdp package on sourceforge[06:19:31] <alanc> when I was younger we did X11 over SLIP on 56k modems and we liked it![06:19:58] <alanc> packets climbed uphill both ways, barefoot, and it was still better than CLI-only![06:20:20] * alanc goes off to find his metamucil[06:22:29] * [lewellyn] hands alanc an lbxproxy[06:22:29] <sivanov> damn, this osol 127 dies every time i run iozone over 8gb zvol[06:23:37] *** Hamc has quit IRC[06:25:25] *** lcheung has left #opensolaris[06:25:30] *** lcheung has joined #opensolaris[06:25:37] <CodeWar> woohoo my vbox osol problems are behind me .. disable graphics acceleration provide the sucker 10mb vmem and (cant believe I was doing this) disable compiz[06:25:39] <CodeWar> works like a charm[06:26:11] *** kris__ has joined #opensolaris[06:26:22] <alanc> compiz was probably coming on by default because it saw the graphics acceleration & VRAM[06:26:38] <Andys^> heheh[06:26:45] <Andys^> compiz... first thing i disable on installing any OS ;)[06:26:53] <CodeWar> probably I had no intention of enabling it :-)[06:27:13] <Andys^> presumably its only enabled by default to get good reviews from 'normal' people..[06:27:28] <CodeWar> now why normal poeple would use OSOL beats me[06:27:45] *** kris__ has quit IRC[06:28:30] *** doug_ndndn has joined #opensolaris[06:28:59] <devians[w]> CodeWar, im fairly normal i think. i use osol :)[06:29:08] <devians[w]> admittedly mainly because zfs is sexy[06:30:26] <Triskelios> I *like* compiz, and I put a lot of effort into it[06:30:54] <sivanov> i didnt give up yet the idea of building nice osol FC SAN using COMSTAR[06:32:26] <Andys^> coming soon: compiz for cars. when you press the brakes, the screen dims and sparkles appear and it animates the wheels slowing down, before actually slowing you down[06:33:04] <CodeWar> actually they have an elastic thing where your window bounces like crazy .. thats hilarious .. cant figure who would use that[06:33:07] *** devians[w] has quit IRC[06:33:20] *** devians[w] has joined #opensolaris[06:33:50] *** devians[w] has joined #opensolaris[06:36:15] <sivanov> http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=250604701891&ssPageName=STRK:MEWAX:IT[06:36:22] <sivanov> guys got it cheap[06:36:49] <sivanov> it looks like a nice zil drive[06:36:52] <[lewellyn]> CodeWar: actually, compiz seems to be very popular with those with accessibility needs. the elastic thing was popular with a low-vision guy i was talking to a few weeks ago[06:37:29] <CodeWar> [lewellyn], if thats true dont I feel like an idiot[06:38:42] <Andys^> sivanov: cool... although, you could have probably gotten two new intel x25-E for almost the same price[06:39:45] <sivanov> it has more bandwidth though[06:40:22] <Andys^> two x25-e in raid0 :)[06:40:51] <sivanov> but i heard that one has like 100k 4k iops[06:41:37] *** icarus901 has quit IRC[06:42:05] *** CodeWar has quit IRC[06:42:26] <[lewellyn]> CodeWar: it's true that i've talked to "disabled" folk who like compiz. i'm extrapolating from that, as to its general popularity.[06:42:47] <[lewellyn]> but i can certainly see how dimming background windows would help people with vision problems, for example[06:42:59] <[lewellyn]> remember that a year ago, i laughed at compiz too ;)[06:43:25] <Shoggoth> does anyone know of a page showing benchmarks of lofi aes vs blowfish >[06:43:26] <Shoggoth> ?[06:47:21] <sivanov> http://tinyurl.com/y9rct54 those 3nity guys have tested x25-e in raid 0 and raid 5[06:47:44] <sivanov> and with write cache turned off on the drives the performance is not very stellar[06:48:05] <Andys^> sivanov: yeah, i just noticed the ioextreme has really high random write IOPS[06:48:47] *** Funk-E has left #opensolaris[06:49:19] <sivanov> there is one nuance though: if its cache is power fault tolerant[06:49:35] *** devians[w] has quit IRC[06:49:47] *** devians[w] has joined #opensolaris[06:49:54] *** devians[w] has quit IRC[06:50:07] *** devians[w] has joined #opensolaris[06:50:18] <Andys^> sivanov: i dont believe it is[06:51:10] <Andys^> fusionio don't seem to make any SATA or SAS drives[06:51:42] <Andys^> :([06:51:59] <Andys^> so that means it will only ever be able to put into a single server and not shared storage[06:52:30] *** skullone has joined #opensolaris[06:52:44] <sivanov> yes, it limits its use very much[06:53:04] *** Mousk4r has quit IRC[06:53:41] <Andys^> STEC charge even more than fusionio for SAS SSDs[06:54:58] <sivanov> i want RAM bbu board with sata interface[06:55:07] <Andys^> Acard makes such a thing[06:55:18] <Andys^> it has high small block IOPS but the throughput isn't as good as flash SSD[06:55:29] *** ahmed-tux has quit IRC[06:55:46] *** radsy has quit IRC[06:57:21] *** MaxVandervelde has joined #opensolaris[07:01:39] <MaxVandervelde> Hey, I just installed openGrok and now i'm getting the following error when trying to sstart tomcat5: "/usr/bin/rebuild-jar-repository: error: Failed to set JAVACMD"[07:01:41] *** ahmed-tux has joined #opensolaris[07:01:52] <MaxVandervelde> anyone seen this before?[07:07:25] *** lattera has joined #opensolaris[07:09:00] *** ruomad has quit IRC[07:09:16] <[lewellyn]> MaxVandervelde: i'd suggest asking in #opengrok, but it's kind of empty :P[07:09:21] *** setheus has quit IRC[07:09:24] <[lewellyn]> they have a mailing list though :)[07:10:35] <MaxVandervelde> yeah, I looked in there and figured I had a better shot in here :P[07:11:27] <Triskelios> MaxVandervelde: if java is correctly installed it's probably a bug in the script, google has plenty of results...[07:12:06] <Triskelios> maybe try running it with sh -x to get more info[07:12:56] *** derchris has quit IRC[07:13:01] *** derchris has joined #opensolaris[07:13:58] <MaxVandervelde> google doesn't give me much, and I've tried what it does[07:14:46] <Triskelios> you could just use glassfish instead of tomcat[07:18:20] <MaxVandervelde> I didn't get this far with tomcat to start over :P[07:24:59] *** MaxVandervelde has quit IRC[07:27:19] *** tsoome has quit IRC[07:27:49] *** vinayak has joined #opensolaris[07:30:35] *** savitr has joined #opensolaris[07:31:18] *** ivo_ has joined #opensolaris[07:31:49] *** swift__ has joined #opensolaris[07:33:14] *** swift has quit IRC[07:33:25] *** swift__ is now known as swift[07:34:40] *** Mousk4r has joined #opensolaris[07:34:55] *** turtle has quit IRC[07:38:44] *** winstonw has quit IRC[07:39:06] *** CodeWar has joined #opensolaris[07:39:20] *** winstonw has joined #opensolaris[07:44:11] *** xapiens has quit IRC[07:46:37] *** savitr has quit IRC[07:47:25] *** devians[w] has quit IRC[07:47:39] *** devians[w] has joined #opensolaris[07:51:05] *** relling has quit IRC[07:52:38] *** ThothCrimson has joined #opensolaris[07:54:02] *** ivo_ has quit IRC[07:57:25] *** Zauberpo1y has joined #opensolaris[07:58:51] *** winstonw has quit IRC[07:59:23] *** Zauberpony has quit IRC[08:02:10] <system5> hey, is alanc still around or did he go to sleep already?[08:02:21] * alanc snores softly[08:02:27] <system5> dang[08:02:32] <alanc> what's up?[08:02:53] <system5> I just sent you an e-mail, if you could forward it to service-request, it would be much appreciated[08:03:20] <system5> we just want to replace Windoze and Ubuntu boxen with either OpenSolaris or Solaris 10 and get security patches and get a quote on the cost[08:03:36] <system5> we don't intend to be bothering Sun support personnel much[08:03:52] <system5> I probably help people more than I ask for help anyway[08:04:26] <system5> I cc-ed the e-mail to my programmer buddy at USC who wants to spend the $324 on booze, let's hope he's too wasted to chime in[08:04:54] *** tsoome has joined #opensolaris[08:05:10] <system5> although he does agree that Solaris is superior to Ubuntu even though he uses Ubuntu as his main desktop / development platform[08:07:33] <system5> I wonder what it is about programming that drives developers to alcohol[08:07:40] <Andys^> hehe[08:07:56] <Andys^> i just spent about 30 mins butchering the source to this program because sem_wait is causing a segfault on my opensolaris box[08:08:04] <Andys^> thats what.[08:08:15] *** winstonw has joined #opensolaris[08:08:46] <system5> I'm a Linux sysadmin, and after a long day of providing "therapy / counseling" to customers with borken Linux machines, I need to go home and eat some comfort food[08:08:57] <system5> Linux is like an abusive relationship and I am the battered wife, I'm afraid[08:10:19] <system5> I heard even Linus Torvalds himself lost all data on his BTRFS hard drive after a nasty yum update on Faildora[08:10:28] <system5> I remember hearing something about it on LKML[08:11:29] <lcheung> system5: I hope Linus did a push before the crash. :D[08:11:34] <system5> if even Linus can't be safe from destructive updates on Linux, then who can?[08:11:43] <system5> what hope do we mere mortals have[08:11:52] <system5> hmmm, maybe I'll dig up the URL for that on lkml[08:11:57] *** p3rror has quit IRC[08:13:18] <Andys^> yeah well[08:13:27] <Andys^> btfs is not production-ready. at least he's eating his own dogfood[08:13:41] <system5> true[08:14:00] <system5> the battle of Oracle Unbreakable BTRFS vs Oracle Unbreakable ZFS[08:14:04] <system5> such an epic battle[08:14:09] <system5> kind of like playing checkers against yourself[08:14:22] <Andys^> i suspect they will turn out to target different purposes[08:14:23] <system5> I wonder who will win?[08:14:41] <Andys^> like, oracle makes Oracle SQL server, but now also has mysql and berkely DB in its folds[08:15:05] <Andys^> ZFS will stay for the enterprise features like RAIDZ3[08:15:07] <system5> there was some smart chick who used to work at Sun, forget her name, she has short blonde hair and lives in Silicon Valley but now she does BTRFS for the Linux kernel[08:15:17] *** tsoome1 has joined #opensolaris[08:15:19] <Andys^> btrfs will support end user stuff like RAID5 online expansion[08:15:29] <system5> she was talking smack on her blog about BTRFS having all these planned features that are better and blah blah blah[08:15:34] *** tsoome has quit IRC[08:15:35] *** tsoome1 is now known as tsoome[08:15:36] <madwizard> Coffee[08:15:37] <system5> BTRFS still misses the point IMO[08:15:42] <lattera> I thought btrfs development was dead[08:15:42] *** hsp has joined #opensolaris[08:15:48] <Andys^> and i bet btrfs will come with a defrag tool too. that stuff is important to linuxites[08:16:03] <system5> you still need different layers, like BTRFS doesn't kill LVM[08:16:15] <system5> and LVM sucks therefore BTRFS will still suck since it still involves LVM[08:16:17] <alanc> system5: forwarded, though you might find service people take your e-mail more seriously in the future if the real name field doesn't look so much like spam[08:16:17] <Andys^> i thought btrfs was aiming to be a volume manager too?[08:16:28] <system5> oh, I"ll send from a real e-mail then[08:16:31] *** doug_ndndn has quit IRC[08:16:44] <system5> well, my other ones were sent from my real gmail that had my real name in it[08:17:08] <system5> I don't think BTRFS simplifies the equation by getting rid of all the layers like ZFS does[08:17:37] <alanc> okay, was just saying - I sent it on anyway, but I think the people who answer it are all in the US, so probably won't see it until tomorrow morning[08:17:43] <system5> Sorry Alan, for a long time I was paranoid about things like spokeo.com that post false information about people on the internet[08:18:13] <system5> I'm trying to keep my real name internet profile as low as possible so that it's easy to get security clearance jobs and other goodies in the future[08:18:27] <alanc> clearly a foreign concept to me[08:18:40] *** hajma has joined #opensolaris[08:18:46] <system5> I don't use or visit myspace or facebook[08:19:08] <alanc> my mother commented once that she tried googling coopersmith and found pages and pages of my work mailing list posts[08:19:13] <system5> but yeah, I keep a very low profile, but OpenSolaris was so good, I had to get involved even though I wasn't really sure if I wanted to be involved since I make a living off of Red Hat[08:19:42] <system5> wouldn't it stink if I went in to my next Red Hat sysadmin job and all that came up when my boss googled me was hundreds of pages of my real opinions about how much I hate Red Hat?[08:19:57] <Andys^> hey guys[08:20:00] <system5> I'm not an SCSA or anything yet, so Red Hat jobs are all I get right now[08:20:00] <Andys^> humour me for a minute[08:20:03] <system5> ok[08:20:09] <Andys^> why is size_t 32 bit on this opensolaris box, but 64 bit on freebsd-amd64 box[08:20:15] * system5 puts on his humor cap[08:20:31] <system5> Andys^ what is your output for:[08:20:34] <alanc> because you're compiling a 32-bit app on opensolaris and 64-bit on freebsd?[08:20:34] <system5> isainfo -kv[08:20:40] <Andys^> alanc: i'm just trying to use gcc[08:20:41] <madwizard> system5: I *have* SCSA and still all I get is CentOS jobs. :)[08:20:47] <Andys^> # isainfo -kv[08:20:47] <Andys^> 64-bit amd64 kernel modules[08:20:50] *** ahmed-tux has quit IRC[08:21:02] <system5> not bad for a Red Hat guy eh?[08:21:03] <Andys^> main() { printf("%d\n", sizeof(size_t)); } <-- prints "4"[08:21:18] <alanc> gcc defaults to -m32 on Solaris, but -m64 on most other OS'es, since they didn't do the 32-bit compability as much originally[08:21:24] <Andys^> oh[08:21:29] <system5> try using Sun Studio instead of gcc maybe[08:21:33] <alanc> add a -m64 to your gcc command on Solaris[08:21:38] <system5> Sunpro / cc[08:21:46] <system5> or what Alan said[08:21:53] <alanc> Sun Studio has the same default - need to use -m64 there too to get a 64-bit app[08:22:00] <system5> interesting[08:22:01] <Andys^> alanc: thanks.. that fixed it.. i had no idea i was compiling all this stuff in -m32 default[08:22:32] <alanc> this bites OpenSolaris everytime Phoronix does a benchmark, since they only use compiler defaults[08:22:43] <system5> true[08:23:06] <Andys^> oh this is going to cause me a huge headache. i can't use a 64 bit library in a 32 bit ruby...[08:23:09] <system5> did those Phoronix guys every figure out how to compile 64 bit on OSol? Or are they still benchmarketing 32-bit OpenSolaris vs 64-bit Linux?[08:23:24] <Andys^> system5: bahaha[08:23:30] <system5> benchmarketing..... IBM style[08:23:56] <system5> we give benchmarks for everything except our most expensive mainframes, which don't need to be benchmarked because they are "magic"[08:24:25] <alanc> they know how, they just poke fun in their presentations about lame vendors who don't set their compilers to optimize for benchmarks by default[08:24:29] <system5> see why I need to be Anonymous? what if I got a job at IBM and this stuff came back to kick me in the arse[08:25:18] *** ahmed-tux has joined #opensolaris[08:25:24] <system5> well, gcc isn't a vendor, is it? It's GNU[08:25:48] <alanc> being reasonable online is the other alternative - I've had plenty of people let me know that I've got a spot at their company if Sun/Oracle laid me off because they know who I am from my online postings[08:26:08] <alanc> gcc on Solaris is configured for that default by Sun[08:26:10] <system5> once we get a native OpenSolaris port to the mainframe we can do a real benchmark vis a vis an UltraSparc T5440 or something[08:26:22] <system5> native as in, not running inside zVM[08:27:31] <system5> btw, is it true that the OpenSolaris on System Z guys can't get libC or something that they need?[08:27:39] <system5> I heard that on the rumor mill[08:27:59] <system5> that it was working for a while, but now they can't get the latest libc or some other key binary only component[08:28:08] <system5> wonder if it's true[08:29:24] *** p3rror has joined #opensolaris[08:29:26] <[lewellyn]> libc is binary-only now?[08:29:49] <alanc> some of the i18n routines always have been[08:30:10] <[lewellyn]> well, yeah. but not libc itself[08:30:30] *** doug_ndndn has joined #opensolaris[08:31:11] <alanc> apparently the Sun engineer that was compiling those bits of libc for them is gone, and they've been whining to the register and mailing lists that no one else is doing it now[08:31:54] <[lewellyn]> i don't see how lack of i18n is a huge failing for them yet, what with the status of the project i last saw :P[08:32:23] *** phimic has joined #opensolaris[08:32:28] <[lewellyn]> of course, whining is a great way to get it back on people's radars[08:33:12] <alanc> and if IBM really cared, I bet they could contribute replacement code to the emancipation project from ICU or AIX or one of their many other code bases[08:33:47] <trygvis> has IBM ever shown any interest in the system/z port?[08:33:55] <[lewellyn]> yeah. i keep failing to see how oracle should be solely responsible for a system/z port[08:34:02] <[lewellyn]> trygvis: only slight[08:34:44] <[lewellyn]> they recently got an emulator license, i recall[08:34:52] <system5> I thought the main Sun engineer working on the Sirius port was Jeff Savit and he still works at Sun AFAIK[08:35:16] <system5> Jeff Savit wrote a book on IBM Mainframe VM performance which is famous and which I also can't find anywhere for some reason[08:35:18] <alanc> never heard of him[08:35:34] <trygvis> Sirius port? which platform is that?[08:35:43] <system5> Sirius port = port to IBM system Z mainframe[08:35:45] <system5> architecture[08:35:56] <system5> which usually only runs z/OS, z/VM, Linux, etc.[08:36:05] <trygvis> ah, ok[08:36:22] <system5> this is what it looks like:[08:36:23] <system5> http://en.wikipedia.org/wiki/IBM_System_z9[08:36:47] <system5> I've logged in to one before, it was really hard trying to get a TN3270 emulator that worked on OpenSolaris 2008.05 back in the day[08:37:11] <system5> this is a bigger / better pic: http://upload.wikimedia.org/wikipedia/commons/2/21/Front_Z9_2094.jpg[08:37:18] <[lewellyn]> system5: alanc has a tendency to use vague words. gone can mean anything from "assigned to another project instead" to "fell out of an airplane without a parachute"[08:37:26] *** Beket has joined #opensolaris[08:38:18] <[lewellyn]> i take his words at face value, most of the time. it's not fair to overinterpret them :)[08:38:21] <alanc> I thought they said he was gone from the company in the e-mail they sent to osol-discuss - I see a Jeff Savit in the employee database, but having never heard of him before, have no idea what his situation is, or if he was even the employee in question[08:38:30] <system5> alanc, check out this link for Jeff Savit's blog:[08:38:31] <system5> http://blogs.sun.com/jsavit/entry/opensolaris_on_system_z_finally[08:38:38] <system5> IBM is totally scared of this guy[08:38:51] <[lewellyn]> he was still in sales in phoenix a couple weeks ago...[08:38:55] <alanc> scared that he might help them sell hardware? frightening[08:39:16] <system5> because they post all this stuff about how great their mainframes are, and then he knocks them down with actual math and numbers showing that some of their claims are impossible[08:39:17] <alanc> and this conversation is already way more than I care about the system/z port[08:39:31] <system5> and he used to be a big expert on VM before he started working at Sun[08:39:40] <[lewellyn]> alanc: likewise[08:40:08] <system5> he had an epic battle where Jeff Savit would post something on his blog, and then 10 IBM dudes would respond, and he would take them all on at once in argument[08:40:18] * [lewellyn] wanders toward the warm bed with a battery-powered psone[08:40:23] <alanc> I can at least see why there are some people interested in PPC or ARM, but mainframes? how boring[08:40:36] <madwizard> Coffee[08:40:37] <alanc> and how unlikely to ever be used by anyone[08:40:42] <system5> hey, there are still more IMS transactions per day than there are web page views on the internet[08:40:54] <system5> used any time you withdraw money from an ATM[08:41:17] <alanc> sorry, I meant it was unlikely the opensolaris port would be used[08:41:34] <system5> I think it could be used to kick the butt of the LInux port[08:41:50] <[lewellyn]> the only time i've used system/z was when ibm convinced a company i was at that it was a good migration path off of s/390. i still disagree :P[08:41:54] <alanc> if you're blowing millions on a mainframe, it's not to run a OS with no vendor support that no other mainframe vendor in the world uses[08:42:05] <system5> you guys don't realize how much you scared the #$%@ out of Red Hat and SUSE with the possibility of that port though[08:42:34] <system5> mainframe is the biggest cash cow in IT right now, and Red Hat and Suse want to keep that revenue to themselves (and to IBM)[08:43:07] <[lewellyn]> but it's useless if there's no software to run on it[08:43:16] <causality> i doubt we'd ever get rid of our mainframe[08:43:18] <causality> s[08:43:44] <system5> just get NetBSD pkg-src running like Joyent did[08:43:57] <system5> and now your weird ports like Power, Itanium and System Z become useful[08:43:58] <causality> although these days i'm wondering if people use it for anything more than an ftp server[08:43:59] <[lewellyn]> i suppose you COULD argue that solaris + oracle on a system/z may make people seriously look at it. but i don't care to argue any side of that[08:44:46] <system5> same with the ARM port of OpenSOlaris[08:44:59] <system5> if only we had better pkg-src integration....[08:45:07] <[lewellyn]> the arm port has some vital flaws[08:45:19] <[lewellyn]> czfs being one of them[08:45:23] <system5> really? I don't have an ARM CPU to play with so what are they?[08:45:25] <system5> czfs?[08:45:30] <[lewellyn]> they forked zfs[08:45:34] <alanc> Czech file system?[08:45:43] <system5> I do have Hercules though, which actually, I think does build on OpenSolaris for playing with Z architecture[08:45:45] <asyd> \_o<[08:45:49] <[lewellyn]> alanc: that'd explain a lot! ;)[08:45:55] <system5> LOL![08:46:02] *** Section9 has joined #opensolaris[08:46:17] <[lewellyn]> i forget which nationalities were mostly behind the arm port[08:46:33] <Andys^> alanc: any idea how to make 64 bit the system-wide default for gcc?[08:46:44] <[lewellyn]> i tend to not notice things like that, especially since most people have a hard time figuring out mine from the hours i keep :([08:46:52] <alanc> I think you'd edit it's spec file, but I've never done it myself[08:46:55] <system5> all you sun engineers need to check out this sun blog link:[08:46:56] <system5> http://blogs.sun.com/jsavit/entry/no_there_isn_t_a[08:47:01] <[lewellyn]> isn't gcc sfw?[08:47:07] <system5> it's the most epic battle ever fought between Sun and IBM on the Internet[08:47:12] <system5> like Kursk or Stalingrad or something[08:47:17] * [lewellyn] isn't a sun engineer so he is glad he doesn't have to click that link[08:47:18] <system5> scroll down to the blog comments[08:47:53] <system5> starts with "I really don't like to use this blog for refuting IBM exaggerations about mainframes...."[08:48:29] <system5> shows the difference between IBM's marketing hype and IBM's reality[08:48:45] *** Bateau has quit IRC[08:49:32] *** raichoo has joined #opensolaris[08:49:33] *** melbogia has quit IRC[08:49:36] <system5> Here is the IBM counterattack at mainframe.typepad.com[08:49:39] <system5> http://mainframe.typepad.com/blog/2008/07/response-to-j-1.html[08:50:11] <system5> anyway, the IBM marketing guys were being real jerks in how they compared Sun hardware if you look at the fine print, and the thing is that a lot of people believe the hype[08:50:29] *** melbogia has joined #opensolaris[08:50:35] <Andys^> alanc: for the record, i had to use CFLAGS = CXXFLAGS = LDFLAGS = "-m64" ; configure --build=x86_64-pc-solaris2.11[08:50:38] <Andys^> that seems to have done the trick[08:50:53] <Andys^> now I can mmap() a 1TB file :X[08:51:31] <alanc> heh, yeah, 64-bit is kinda required for that[08:51:35] <alanc> along with a big disk[08:51:42] <Andys^> nope - its a sparse file :)[08:51:53] <Andys^> its an in-memory database which no longer fits in memory.. hehehe[08:54:53] <Andys^> sadly, now ruby on this box is useless unless every time i go to install a new native library i use the same compile flags[08:55:11] <system5> don't you wish you had a ports tree Andys^ ?[08:55:27] <system5> either that or Gentoo seems to simplify the building[08:55:45] <system5> I guess you could build your own SysVR4 packages or set up your own IPS repository to speed things along[08:55:52] <Andys^> i think that while opensolaris gets so many things right, freebsd got the 32->64 bit switch more right[08:56:00] <system5> how is that?[08:56:09] *** FireflyST has quit IRC[08:56:19] <Andys^> this problem simply doesn't happen on freebsd[08:56:27] <system5> ic[08:56:44] <system5> one day Gentoo OpenSolaris might see the light of day (again)[08:57:00] <system5> a 100% build from source OpenSolaris distro, what fun that would be[08:57:17] <system5> I'll just go on vacation or something before I start to compile OpenOffice on a slow box[08:57:26] <alanc> because FreeBSD chose to optimize for building new software and assumed you'd rebuild it all when you moved, while OpenSolaris chose to preserve compatibility with existing binaries, since you may not have all the sources to the libraries to rebuild them[08:57:45] *** |AbsyntH| has joined #opensolaris[08:57:49] <system5> well both are good[08:57:55] <alanc> different constraints, different design decisions[08:58:00] <system5> FreeBSD is kind of doomed because of lack of IHV and ISV support[08:58:04] <alanc> different target customers[08:58:23] <system5> I don't even think Nvidia makes binary blob drivers for FreeBSD[08:58:29] <system5> or that Oracle, SAP, IBM, etc. support it[08:58:40] <alanc> nvidia makes freebsd binary blobs[08:58:42] <system5> the only IHV that seems to have FreeBSD love is Intel[08:58:44] <system5> oh they do[08:58:47] <system5> awesome[08:58:56] <system5> I had problems with nvidia on openbsd back in the day[08:58:58] <system5> big problems[08:59:16] *** nikolam_ has joined #opensolaris[08:59:36] <alanc> and it is quite amusing to hear you switch so quickly from promoting OpenSolaris/zseries to dissing FreeBSD/x86 for lack of IHV & ISV's, when FreeBSD has more of those than OpenSolaris/Z ever has or will[08:59:54] <system5> I'm not dissing FreeBSD[08:59:58] *** nikolam has quit IRC[09:00:00] *** biant92 has joined #opensolaris[09:00:06] <system5> I just tried to use it as my main desktop OS for a long time before I switched to OpenSolaris[09:00:13] <system5> and it was even harder to use than OpenSolaris was[09:00:29] <system5> a lot of my work involves hooking up a console cable to a CISCO device[09:00:55] <[lewellyn]> you have cu in both places[09:01:38] <system5> and I was running FreeBSD, and there were no Keyspan USA 19HS USB to DB9 serial device drivers for FreeBSD, so I had a hard time figuring out how to console in from my FreeBSD laptop (which had no serial port) to the CISCO switch[09:01:49] <system5> eventually I gave up and installed OpenSolaris 2008.05 and now I'm here :-)[09:02:20] <system5> FreeBSD is still one of my all time favorites as a server OS, but I have trouble getting work done with it as a sole desktop OS[09:02:30] <system5> OpenSolaris allowed me to get more work done[09:02:58] *** nikolam__ has joined #opensolaris[09:03:00] <system5> any of you guys ever use FreeBSD as your desktop instead of Solaris or Linux or Windoze ?[09:03:12] <asyd> or mac[09:03:16] <system5> or Mac[09:03:18] <system5> good point[09:03:19] * blkruby uses ubuntu for desktop and arch for server[09:03:28] <system5> yeah, Arch is popular now[09:03:42] <causality> system5, i've tried in the past[09:03:48] <causality> but ultimately windows is best for me[09:03:53] <system5> how did it work for you? I had a rough time[09:04:09] <causality> system5: nothing works for me on the desktop as well as windows[09:04:28] <system5> well, I run windows in VirtualBox seamless mode on my OpenSolaris[09:04:30] <[lewellyn]> system5: this isn't the place to poll people for their freebsd usage[09:04:47] <system5> when I was running FreeBSD on my desktop, there was no VirtualBox port for FreeBSD, so I was kind of stuf[09:05:01] * blkruby wishes seamless mode was a bit faster on his 3 year old laptop[09:05:14] <system5> * I was kind of stuck[09:05:25] *** nikolam_ has quit IRC[09:05:25] <system5> ok, I was just wondering if anyone else had a similar journey[09:05:45] <system5> frustration with Linux takes you to *BSD and then difficulty with BSD on the desktop takes you to OpenSOlaris 2008.05[09:05:51] <Andys^> alanc: right. except... you can run 32-bit binaries from a 32-bit freebsd system, quite happily in a 64-bit freebsd system (in fact, the entire precompiled binary ports tree only generates 32-bit binaries)[09:06:11] <Andys^> but you're right - most freebsd users tend to compile everything from scratch[09:06:22] <system5> you can run 32-bit Solaris binaries on 64-bit Solaris[09:06:38] <system5> I do it all the time[09:06:41] <Andys^> right[09:06:54] <Andys^> but as discovered today... when you run "make" on 64bit, it makes a 32bit by default[09:07:06] <[lewellyn]> it depends on your compiler settings[09:07:12] <Andys^> system5: your story sounds like mine.. i ran freebsd as my desktop OS for many years before finally getting fed up[09:07:15] <system5> only Linux and Windows don't support this type of behavior (actually in Linux I think there was a way to "hack" it but it's not as seamless as it is on Solaris so I won't go into it here)[09:07:21] <Andys^> lewellyn: "by default"[09:07:32] <system5> thought there would be someone else who did the same journey :-)[09:07:45] <[lewellyn]> Andys^: would you prefer to build a binary which will run on any system, or one which targets what you happen to be on?[09:07:52] <Andys^> lewellyn: the one i'm on![09:07:59] <Andys^> that should be the default[09:08:06] <[lewellyn]> no, it shouldn't ;)[09:08:06] <system5> it's not FreeBSD's fault that they don't have some gigantic corporation like Apple telling people that they have to port Photoshop to FreeBSD[09:08:07] <Andys^> if you want to target a particular arch, thats your choice, with flags[09:08:20] <system5> * telling Adobe that they have to port Photoshop to FreeBSD[09:08:23] <[lewellyn]> you just contradicted yourself, Andys^[09:08:30] <Andys^> the final insanity is installing an application and you end up with both 32 and 64 bit copies installed[09:08:46] <[lewellyn]> you're referring to e.g. mysql?[09:08:53] <system5> I did notice a lot of people outside of Sun in the "community" have BSD-ish backgrounds[09:08:53] <Andys^> lewellyn: the default is to target i386, even if thats not your arch. i say the default should be to target the arch of the current machine you're logged into[09:09:08] <[lewellyn]> Andys^: you'll find lots of people who would disagree with that[09:09:23] <[lewellyn]> Andys^: hell, i get mail complaining about SFEwine's CFLAGS[09:09:46] <Andys^> lewellyn: lots of people == solaris users, because they're used to it, i guess? it was the complete opposite of what i was expecting, coming from linux/bsd[09:09:48] <causality> system5: i've been an openbsd user since about 15 years[09:10:18] *** plompi has joined #opensolaris[09:10:19] <[lewellyn]> Andys^: lots of people == people who are used to packaging something up and deploying it everywhere.[09:10:43] <Andys^> lewellyn: right... that to me, is a special case[09:10:49] <[lewellyn]> the build box may be some fancy quad-core box, but the destination may be an aging p3[09:10:59] <[lewellyn]> that's the generic case for solaris :)[09:10:59] <Andys^> thats an enterprise mindset shining through again[09:11:29] <[lewellyn]> as you said yourself: "if you want to target a particular arch, thats your choice, with flags"[09:11:55] <Andys^> this enterprisey mindset that if you're building something, its because you want to deploy it to your cluster which contains both 32- and 64-bit CPUs[09:12:14] <Andys^> as opposed to the end-user-centric mindset of freebsd/linux, where you're building something on this machine for use on this machine[09:12:38] <[lewellyn]> it certainly makes it easier to package something up that you've built and pass it on to a friend[09:12:43] <Andys^> side note: lets a have poll, who in here is actually using a 32-bit CPU? ;)[09:12:57] <Andys^> (and when was the last time Sun made a CPU that was 32 bit?)[09:13:04] <blkruby> i wish i had 64bit[09:13:07] <Andys^> hehehe[09:13:15] <Andys^> what have you got blkruby?[09:13:21] <[lewellyn]> Andys^: when was the last time sun made an x86? (which most of this channel is using)[09:13:24] <Auralis> i still have box with a 32bit cpu in use[09:13:27] <blkruby> 32 bit dell inspiron[09:13:29] <madwizard> Andys^: I've seen this "enterprisey" mindset on good Linux admins, because they don't give a flying shit about building 7 different packages, just because they have 7 different (but x86 compatible) platforms in house[09:13:47] <Andys^> even a $300 netbook is now 64-bit...[09:13:49] <madwizard> They preffer one, well build, generic package with one build process and one repository[09:14:05] *** hajma_ has joined #opensolaris[09:14:12] <madwizard> Anyway, performance gains were negligible[09:14:17] <blkruby> madwizard: and the keyword 'stable'[09:14:18] *** hajma has quit IRC[09:14:38] <Andys^> sometimes people do actually need the address space of 64 bit[09:15:04] <[lewellyn]> as i asked earlier: 00:13 @[lewellyn]: you're referring to e.g. mysql?[09:15:09] <Andys^> its nice to laugh at phoronix but what about poor schmucks like me who were using the defaults all this time without releasing... :([09:15:14] <madwizard> blkruby: I thought it doesn't have to be mentioned. :)[09:15:18] <Andys^> err, without realising[09:15:39] *** plompi has left #opensolaris[09:16:01] <[lewellyn]> Andys^: most of the time, you aren't gaining anything from building 64-bit. as mysql shows, you can actually take a performance hit in some cases, by running a 64-bit binary needlessly[09:16:05] <blkruby> madwizard: well *cough* McAfee didn't get the memo on how enterpris software should be stable :P[09:16:26] <madwizard> blkruby: Seems many didn't :)[09:16:42] * causality would need considerably more hardware if he wasn't running 64-bit[09:16:42] <Shoggoth> is there any know problems with running zfs on top of lofi encrypted disks?[09:17:08] *** twisti has joined #opensolaris[09:17:08] <blkruby> oh and the makers of ParidiseEO didn't get the memo on how an installation script should work :P[09:17:35] <madwizard> Andys^: I'd say if a poor schmuck is administrator and does not educate him/her/itself (whatever the PC of this day requires), they are really schmucks :)[09:18:13] <Andys^> madwizard: this is just one of many 'gotchas' i've run into in coming to opensolaris from linux/bsd[09:18:23] <Andys^> i run into it, i learn, i move on..[09:18:36] <brandon> Triskelios, alanc - my problem earlier in gcc was -Wl,-Bstatic being a gcc parameter since there were no static libraries for socket and nsl, i think[09:18:43] <madwizard> Andys^: and this is how it works[09:18:50] <madwizard> Andys^: No better way so far[09:18:52] <Andys^> but it always makes me think, "why does it have to be this way?"[09:19:16] *** zazenrasta has quit IRC[09:19:19] <Andys^> I learned when writing software applications, its often preferable to do what the user expects and not what is some theoretical best[09:19:55] <blkruby> Andys^: and don't forget to be nice to the system admin and make it scalable :P[09:20:08] *** ma3x_ has joined #opensolaris[09:20:13] *** q5sys has quit IRC[09:20:29] <Andys^> like when /usr/gnu didn't used to be the default, then it was changed to be the default - this was an example of being nice to incoming users :)[09:20:29] <[lewellyn]> brandon: that'd be because static libs are bad? ;)[09:20:36] <Shoggoth> Andys^ lewellyn system5 you guys always seem to have something intelligent to say... any thought?[09:20:44] <[lewellyn]> Andys^: with luck, it'll be gone again soon[09:20:51] <[lewellyn]> Shoggoth: no clue.[09:20:53] <Andys^> Shoggoth: uummm... well, performance won't be optimal, other than that, it works?[09:21:21] <brandon> i agree, i'm just trying to build this monster[09:21:22] *** hajma__ has joined #opensolaris[09:21:25] *** zazenrasta has joined #opensolaris[09:21:41] <Shoggoth> I'm currently doing essentially the same thing except instead of lofi I've get a linux dom0 with dmcrypt and a 2008.11 domU on top with ZFS[09:21:53] <Shoggoth> I figure that if I use lofi encryption the overhead should be lower[09:21:58] *** hajma_ has quit IRC[09:22:15] <Andys^> Shoggoth: yes, i'd say that would be less overhead[09:22:42] <Shoggoth> but I'm a little worried that there might be stability issues in the way the two subsystems interact[09:22:54] <Shoggoth> also... the lofi encryption doesn't look particularly strong[09:23:08] <Andys^> pretty sure i've seen someone use lofi with zfs before, for testing[09:23:15] *** ma3x__ has quit IRC[09:23:28] <Shoggoth> Andys^: zfs on top of lofi , or lofi on top of zvol's ?[09:23:34] <Andys^> zfs on top of lofi[09:23:58] <Andys^> [ding!] http://blogs.sun.com/darren/entry/opensolaris_disk_encryption_in_snv[09:25:21] <Shoggoth> Andys^: thanks for that link... very useful[09:26:24] <madwizard> Andys^: /usr/gnu was one of worst things to do to incoming users[09:26:39] <Andys^> madwizard: and why's that?[09:27:14] <sivanov> thermo rectal cryptoanalysys is ftw against disc encription http://www.modchipworld.co.uk/store/images/15w_soldering_iron.jpg[09:27:32] <sivanov> *encryptio[09:27:34] <sivanov> n[09:27:47] <madwizard> Andys^: Because it is hiding the differences. It's like you marry some nice model and at the night she takes off her blond hair, big breasts and straps on some additional layers of fat[09:28:12] <madwizard> And when you're all angry that she lied to you, all she can say is: Murdock made me do this[09:28:15] *** tsoome has quit IRC[09:28:40] *** tsoome has joined #opensolaris[09:28:43] <madwizard> And once a user gets idea that OSOL does not differ that much, they *may* want to start using ACLS on ZFS of projects[09:28:47] <madwizard> And all the hell gets loose[09:28:58] <madwizard> Because, wow, manpages say nothing about acls[09:29:00] <madwizard> How so?[09:29:09] <madwizard> Where the fuck they have gone?[09:29:51] <madwizard> If your average user can't cope with ifconfig having to have -a flag, they won't be able to find that PATH has been broken for them[09:30:11] <Andys^> madwizard: i'm not suggesting opensolaris should actively drop its good and differentiating features in order to emulate linux..but for most of these little issues...[09:30:34] <madwizard> Yes[09:30:57] <Shoggoth> lol... that's one thing I'm glad linux has (ifconfig -a) when I stopped doing sun admin and started doing hpux and linux admin at least linux was a cozy familiar beast :-P[09:30:59] <[lewellyn]> Andys^: the better solution is to fix the userland to accomodate that the world's changed in the past 25 years[09:31:02] <madwizard> It won't be fixed for you, if you want to start using OSOL chmod, ps etc.[09:31:10] <madwizard> You'll have to forget oll the gnu goodnes[09:31:19] <madwizard> And then you'll be angry[09:31:45] <madwizard> Just as I am every time I try to set acls on new OSOL machine and forget about this /usr/gnu thing[09:31:50] <Andys^> i thought that getting a larger userbase and momentum in the community may be more important than some of the particular defaults which don't make sense to a n00b[09:32:00] *** Erwann has joined #opensolaris[09:32:14] <madwizard> Andys^: It's not working anyway[09:32:18] <[lewellyn]> Andys^: again, my previous comment stands[09:32:47] <[lewellyn]> madwizard: but didn't you know? if you forget to set $PATH to something sane, you're incompetent![09:32:53] <madwizard> Shoggoth: I don't give a damn, since OS X which I had to use for 2 years didn't have it[09:33:14] <madwizard> And Windows which I have to use now doesn't have ifconfig at all[09:33:21] <madwizard> So why not learn?[09:33:24] <madwizard> And live with it[09:33:38] <Shoggoth> woah dude I wasn't trying to upset you[09:33:56] <Shoggoth> I actually agree with you[09:34:14] <[lewellyn]> madwizard: you can validate on jucr, right? :)[09:34:33] <causality> madwizard: windows has an equivalent of ifconfig[09:34:39] <[lewellyn]> ipconfig[09:34:44] <lblume> madwizard: It's ipconfig on Windows :-P[09:34:48] <Shoggoth> I was trying to show the irony inherent here as most of the linux stuff seemed to be very solaris-ish early on before it went on a 100% gnu rampage[09:34:51] <causality> [lewellyn]: the equivalent is netsh[09:34:55] <causality> not ipconfig..[09:35:03] <[lewellyn]> netsh is more like dladm :)[09:35:33] <madwizard> lblume: I know, I use it[09:35:41] <madwizard> [lewellyn]: I can[09:35:53] <lblume> madwizard: Sorry, I really assumed you knew already, I was teasing ;-)[09:36:00] <madwizard> lblume: :)[09:36:01] <madwizard> Cffee?[09:36:01] <lblume> madwizard: Want some coffee?[09:36:04] <madwizard> Yah![09:36:18] <[lewellyn]> madwizard: would you be kind enough to validate http://jucr.opensolaris.org/review/packages/7495/ so i can see how badly i screwed it up? :)[09:36:21] <Andys^> if i had to choose between (up to date, functional on latest hardware opensolaris kernel + gnu userland) vs (ailing opensolaris kernel + superior/traditional solaris userland) .. :([09:36:25] <madwizard> Shoggoth: I'm not upset because of what you said, I'm upset since inception of Indiana project[09:36:58] <lblume> The Solaris userland is hardly superior. It's stuck in the 80s for most of it.[09:37:00] <madwizard> Because it was fucking smart to break basic (for me) functionality just to get users and then say to them: Ya know, if you really want to use OSOL, you still have to *gasp* learn![09:37:05] <madwizard> lblume: I agree[09:37:06] * [lewellyn] still wonders what the actual goals of indiana were/are :P[09:37:19] <madwizard> lblume: But it has features which gnu userland lacks[09:37:27] *** jinx099 has left #opensolaris[09:37:31] <madwizard> I'd love to have osol userland brought forth in time[09:37:32] *** dalibor has joined #opensolaris[09:37:45] <lblume> Actually, is there any plan in the GNU tools to integrate support for new Solaris stuff?[09:37:51] <[lewellyn]> afaik, no[09:37:59] <[lewellyn]> not even old solaris stuff[09:38:10] <lblume> madwizard: The reason /usr/gnu is first is PATH is that they gave up on this.[09:38:13] *** ozux has quit IRC[09:38:14] <[lewellyn]> at&t seems keen on adding support, so that's promising[09:38:20] <[lewellyn]> ast++[09:38:37] <[lewellyn]> lblume: no, they didn't give up. they didn't even do the basics :P[09:39:32] <lblume> Yes, I know, I was being nive because the weather is good, make me in a good mood.[09:40:10] *** toasticles has joined #opensolaris[09:40:11] <lblume> nice, even.[09:40:11] <[lewellyn]> sorry. i'm in a bad mood, so everyone else gets to be too[09:40:21] <lblume> Fair enough.[09:40:25] <causality> you're always in a bad mood! ;)[09:40:32] *** tsoome has quit IRC[09:40:41] <madwizard> causality: And that sets the overall tone for the channel :)[09:40:49] <madwizard> jbk: Ping![09:40:52] <[lewellyn]> no, if i were always in a bad mood, i'd be abusing my @ ;)[09:40:52] *** tsoome has joined #opensolaris[09:41:19] * causality boggles at another failfox update[09:42:29] *** hemantha has quit IRC[09:42:58] *** RoyK has quit IRC[09:43:14] *** hemantha has joined #opensolaris[09:45:28] *** lattera has quit IRC[09:46:32] * robinbowes thinks /usr/gnu being the default is a bad idea too, fwiw[09:47:33] <robinbowes> And I'm a linux user/sysadmin[09:47:36] <lblume> I wonder, was there ever anybody finding it a good idea? :-D[09:47:53] <lblume> Those who do, please raise your hands so your names can be written down.[09:48:05] <toasticles> i'd like to set up a build environment for the latest ON, but have tried in the past and ran into problems. unfortunately i don't remember the specific problems i had, but to clarify before i try again from scratch: is a vanilla install of 2009.06, with updates, and the ON build tools and Sun Studio 12u1 from IPS sufficient to build the whole ON package?[09:48:06] <madwizard> Too bad Ian can't do that, can we phone him?[09:48:18] <Shoggoth> hey all I've got another left field question...[09:48:26] <robinbowes> I mean, if I'm on a machine with zfs filesystems that support ACLs, I want a toolset that can like list them and change them![09:48:35] <Shoggoth> what kind of integrity checking does zfs send/receive do?[09:48:51] <Andys^> Shoggoth: quite alot...[09:48:52] *** ozux has joined #opensolaris[09:48:59] <robinbowes> ls -lV ... Huh? Ah, crap... /usr/sun/bin/ls -lV[09:49:15] <Andys^> if its a differential send, it ensures the correct snapshot is on the receiving end, and that it hasn't been modified[09:49:35] *** nitrile has quit IRC[09:49:36] <Shoggoth> Andys^: let me put that question in context... I don't trust large file copies... so whenever I do a copy I always do a diff -rq afterwards to ensure[09:49:53] <Shoggoth> but I figure with all the end-to-end checking in zfs on disc[09:50:09] <Andys^> shoggoth: yes.. no need to do that with zfs :)[09:50:14] <Shoggoth> I'd assume that send/receive also verifys the destination data[09:50:21] <Andys^> at the block level, yes[09:50:30] *** nitrile has joined #opensolaris[09:50:42] <Shoggoth> Andys^: can you point me at the doco that talks about that... or perhaps the part of the source :)[09:51:31] <lblume> I'd have assumed that zfs send was sending the checksum information along, but I never actually checked that.[09:52:16] <Shoggoth> I guess what I'm really asking is if the on disk checksum is copied over from the source to the destination or is it recalculated... if it's the latter then any scrambling that happens en-route won't be caught[09:52:33] <Andys^> i think its the on disk one..[09:52:35] *** doug_ndndn has quit IRC[09:52:50] <Shoggoth> or better still does the checksum get sent and the destination get recalculated and compared :)[09:53:47] <Andys^> i've seen checksum errors thrown before[09:53:47] <lblume> cp is just cp. It gets data from one side, and write it on another. No checksumming. There was talk at one point about a zcp that would optimise cp on zfs (eg, not moving data between zfs on the same zpool).[09:53:48] <Shoggoth> does anyone know where this is documented?[09:53:55] *** twisti has quit IRC[09:53:59] <Andys^> during a zfs send/receive... (due to version mismatch, but still..)[09:54:21] <Shoggoth> lblume: yes that I expected... I was asking about zfs send/receive[09:54:48] <Andys^> Shoggoth: well, its dead easy to test.. :)[09:54:56] <Alasdairrr> Is it out yet?!?![09:55:21] <Shoggoth> Andys^: at the risk of sounding thick... care to illuminate?[09:55:23] <Stric> Yes, you missed it.[09:55:29] *** turtle has joined #opensolaris[09:55:29] <Shoggoth> lmao[09:55:31] *** nitrile_ has joined #opensolaris[09:55:34] <Andys^> snapshot1 ; echo 'hello' > testfile ; snapshot2 ; zfs send -i snapshot1 snapshot2 > stream ; edit stream and replace 'hello' with 'h3ll0' ; try to zfs receive it ;)[09:55:38] *** twisti has joined #opensolaris[09:55:53] *** nitrile has quit IRC[09:55:58] <Andys^> fuckit, i'll try it right now[09:55:58] <Shoggoth> hmmm... not a bad idea[09:58:20] <Andys^> cannot receive incremental stream: invalid stream (checksum mismatch)[09:58:21] <Andys^> :-)[09:58:31] <Shoggoth> fan-freakin-tastic[09:58:32] <Shoggoth> thanks![09:58:33] *** nikolam__ is now known as nikolam[09:58:45] *** ImmerDerBest is now known as nettezzaumana[09:58:58] <Shoggoth> hey iirc you're in miranda right?[09:59:00] <Andys^> yeah?[09:59:17] <Shoggoth> nothing... just seeing if my short term memory wasn't scrambled[09:59:20] <Shoggoth> :)[09:59:29] * Andys^ draws his curtains shut[09:59:34] <Shoggoth> lmao[09:59:35] <Andys^> and deadlocks his doors[10:00:22] *** FireflyST has joined #opensolaris[10:00:23] <Andys^> Shoggoth: i suspect it just uses the checksum from the disk... however i know it doesn't support compression[10:00:25] <Shoggoth> I'm wondering if we're the only sydney-siders on here[10:00:44] <Shoggoth> ... doesn't support compression[10:00:49] <Shoggoth> In what way?[10:00:55] <Andys^> which is kinda cool, since it means i can recompress using a much stronger compression if i want to[10:01:03] <Stric> zfs send stream is uncompressed[10:01:04] <Andys^> even if you have compression enabled on the pool, zfs streams are not compressed[10:01:08] <Shoggoth> ahhh... you mean the stream is decompressed[10:01:11] <Shoggoth> right[10:01:19] <Andys^> then it has to recompress it at the remote end, if you have compression enabled there[10:01:33] <Andys^> and there's no dedupe support whatsoever..[10:01:41] <Andys^> it'll simply dedupe the incoming data at the remote end, if you have dedupe turned on there[10:01:43] <Shoggoth> which probably means if you're replicating via a WAN link you want to zfs send | gzip | zfs receive[10:01:47] <Andys^> yes[10:01:54] <Andys^> "pigz -9" instead of gzip...[10:02:01] <lblume> I've just seen in the man that there is dedup.[10:02:03] <Andys^> pigz == multi-threaded gzip[10:02:12] *** vinayak has quit IRC[10:02:13] <Shoggoth> are you for real[10:02:17] <Andys^> lblume: oh? i thought it was "coming one day" ..[10:02:21] *** alanc has quit IRC[10:02:22] <Shoggoth> I was about to type rofl for the great typo[10:02:26] <lblume> Never tried. Just saw it.[10:02:34] <Andys^> URL?[10:02:44] <Andys^> oh wow[10:02:48] <Andys^> oh wait[10:02:48] *** Error404NotFound has joined #opensolaris[10:03:16] <Andys^> hrmm[10:03:17] <Andys^> right[10:03:27] <Andys^> yeah[10:03:32] <Shoggoth> well that leads to an obvious question... is the gzip inside of the kernel really pigz?[10:03:33] <Andys^> afaik, that only dedupes the stream itself[10:03:36] <lblume> It's rather easy to do on a stream. Just checksum all packets going through, if checksum already was seen, don't send.[10:04:03] <Andys^> shoggoth: effectively, yes.. it is only compressing blocks up to 128kb in size, but it can do multiple blocks in parallel across CPU cores[10:04:14] <Shoggoth> right... that's sensible[10:04:18] <lblume> Since the target does dedup too, it will automatically matches the checksum to the already received block.[10:04:25] <Andys^> lblume: yeah.. that isn't really good enough for me... it still has to send the whole changed data even if the new data is identical to something already on the disk at the other end :([10:05:08] <Stric> 'zfs send' stream isn't a two way communication[10:05:24] <lblume> Hmmm, I see, well, that would be the job of snapshots at this point.[10:05:25] <Stric> it can't do "I have a block with checksum xyz, do you want it or do you already have one of those?"[10:05:26] <Andys^> Stric: no, but in the case of a differential send, "zfs send" knows what "Zfs receive" already has at the other end[10:05:42] <Andys^> here's an example[10:05:57] <Shoggoth> ok thanks again Andys^ ... time for me to work off some of the fringe benefits of being an IT worker from around my waist :)[10:05:59] <lblume> Really? I thought it didn't know, just assumed the other end was at the correct snapshot level.[10:06:27] <Andys^> on box_A, i have a big data file called "data1". i do "cp data1 data2", take a new snapshot, then I want to send this new snapshot to box_B which already has the old snapshot (with only data1, and not data2)[10:07:01] <Andys^> if you zfs send that, its gonna have to copy the whole of data2. it will dedupe the contents of data2 with only itself... its not smart enough to know that the entire contents of data2 is already at the other end in the form of data1 in the original snapshot...[10:07:31] <Andys^> if there is no duplicated data internally in the file... then zfs send has to send the whole file :|[10:07:58] *** gebi has quit IRC[10:08:09] <lblume> *nod* agreed.[10:08:14] <Andys^> i remember someone's blog saying this could be fixed in a future version..[10:08:26] <Andys^> zfs send -D is a nice start[10:08:32] <Andys^> and will probably save some traffic[10:08:44] <Alasdairrr> Wouldn't you need two way communication for that?[10:08:54] *** milek_ has joined #opensolaris[10:09:24] <Andys^> Alasdairrr: with a differential send, the "source" snapshot exists identically at both ends[10:09:32] *** doug_ndndn has joined #opensolaris[10:09:54] *** gebi has joined #opensolaris[10:09:58] <Alasdairrr> I meant for avoiding sending duplicate blocks that are present on the destination server[10:11:13] <Andys^> Alasdairr: the same snapshot exists at both ends... so zfs send knows what the other end already has, by looking locally?[10:12:10] *** calumb has joined #opensolaris[10:12:49] *** stewart has quit IRC[10:16:27] <sickness> coffee time[10:17:21] *** FooBarWidget has joined #opensolaris[10:17:21] *** gebi has quit IRC[10:17:55] <FooBarWidget> I'm trying to install gcc, but what's the difference between gcc-432 and SUNWgcc?[10:21:54] *** boos has quit IRC[10:21:59] *** m4rku5 has joined #opensolaris[10:22:47] *** m4rku5 has quit IRC[10:29:30] *** boos has joined #opensolaris[10:34:24] <[lewellyn]> FooBarWidget: release or dev?[10:35:10] *** Dagobert has joined #opensolaris[10:36:51] <FooBarWidget> this is opensolaris 2009.06, so I guess release[10:37:15] *** dalibor has quit IRC[10:37:20] <FooBarWidget> it looks like it's just two different versions of gcc, installable in parallel, but that makes me wonder why there are two different versions and which one works best on opensolaris[10:37:52] <FooBarWidget> I recall having encountered a bunch of gcc bugs in opensolaris 2008.06 and I'm trying to avoid them this time[10:37:56] <trygvis> probably different versions of gcc[10:38:32] <[lewellyn]> i don't see gcc-432 in /release[10:39:24] <[lewellyn]> what's the output of uname -v?[10:44:34] *** dalibor has joined #opensolaris[10:46:18] <FooBarWidget> snv_111b[10:46:41] <FooBarWidget> this is just the opensolaris 2009.06 x86 live cd that I downloaded from the website[10:47:04] *** MattMan has joined #opensolaris[10:47:23] *** gebi has joined #opensolaris[10:47:32] *** Bateau has joined #opensolaris[10:48:55] <[lewellyn]> hrm[10:49:12] <[lewellyn]> ah! developer/gcc/gcc-432[10:49:24] <lblume> Yes, it was renamed later to gcc-43[10:49:36] <[lewellyn]> as you can guess, that's gcc 4.3.2[10:49:46] <[lewellyn]> http://pkg.opensolaris.org/release/info/0/SUNWgcc%403.4.3%2C5.11-0.111%3A20090508T154717Z says that SUNWgcc is 3.4.3[10:50:00] *** Section9 has quit IRC[10:50:04] <[lewellyn]> choices, choices, choices ;)[10:50:29] *** ingenthr1 has joined #opensolaris[10:50:39] <FooBarWidget> is it possible to configure opensolaris to not start X by default? I just want to port my software to solaris so I just need ssh[10:50:46] <FooBarWidget> it's eating a lot of memory this way[10:50:48] <[lewellyn]> smrt: explain disable x[10:50:49] <smrt> osol: svcadm disable gdm sxce: svcadm disable cde-login (See also: disable fancy boot)[10:51:06] <FooBarWidget> does that disable it permanently?[10:51:09] <[lewellyn]> yes[10:51:16] *** ingenthr has quit IRC[10:51:56] <FooBarWidget> is networking in opensolaris dependent on GNOME Network Manager? I'm running it in a VM with NAT[10:51:57] <lblume> SUNWgcc is now gcc-3, too :-)[10:52:05] <FooBarWidget> or will a system service automatically configure dhcp during boot?[10:54:25] <[lewellyn]> FooBarWidget: "it depends" :P[10:55:41] <FooBarWidget> I guess I'll just disable X and see where it gets me[10:56:00] <mui> any news with release?[10:57:13] * lblume checks if /topic is broken[10:57:16] <[lewellyn]> check osol-discuss? ;)[10:57:34] <mui> I guess I should subscribe there[10:57:41] <[lewellyn]> there's an archive[10:57:58] *** panos has quit IRC[11:02:39] <FooBarWidget> it works, thanks[11:05:36] <mui> yawn[11:06:58] <madwizard> [lewellyn]: You have two cc65 submitted[11:07:37] <[lewellyn]> one was a failure[11:08:08] <[lewellyn]> the stupid dropdown box changed from patch to copyright when i hit submit :([11:08:20] <[lewellyn]> then it yelled that i can't have two copyright files[11:08:25] *** master_of_master has quit IRC[11:08:30] <[lewellyn]> which it should have done on the javascript-laden page[11:08:56] <[lewellyn]> (and they need a non-javascript version so we can submit straight from the account we're sshed into...)[11:10:02] *** master_of_master has joined #opensolaris[11:18:48] *** FooBarWidget has left #opensolaris[11:19:46] *** hsp has quit IRC[11:22:37] *** mikefut has joined #opensolaris[11:23:42] <madwizard> [lewellyn]: Validated[11:23:54] <[lewellyn]> thanks! :D[11:26:25] <sickness> what's the state of rtorrent on solaris?[11:26:29] <sickness> http://whacked.net/2008/05/08/updated-rtorrent-libtorrent-packages-for-solaris/[11:26:42] *** hsp has joined #opensolaris[11:26:58] <sickness> is this the most up to date version available?[11:28:06] *** carpenike has quit IRC[11:28:49] *** Fallen_Demon has joined #opensolaris[11:29:54] <[lewellyn]> check SFE[11:35:04] *** held has quit IRC[11:35:38] *** Fullmoon has joined #opensolaris[11:36:33] *** toasticles has quit IRC[11:38:39] <devians> sickness: i would fully advise, having attempted the path you now start on, to just virtualise an os for torrenting. much, much easier.[11:42:13] *** Fullmoon has quit IRC[11:42:37] <[lewellyn]> or just use transmission :P[11:42:49] <[lewellyn]> it has a gui, a cli, a remote, and a web interface[11:43:19] <Beket> Is it faster to scrub 1 pool at a time or doesn't make any difference if I scrub'em'all ? (all disks on same controller)[11:44:44] <devians> transmission is terrible[11:45:03] <devians> its forever banned from various trackers because it always has some incredible showstopper bug.[11:45:04] <tomww> Beket: I think all at the same time is okay, but not if all pools are on the same disk. This would introduce long seek times since heads of the disks might jump between bot partitions[11:45:04] <[lewellyn]> but unlike the old versions of rtorrent, most trackers accept it ;)[11:45:19] <[lewellyn]> plus it's far less crashy than rtorrent[11:45:35] *** MGSsancho has joined #opensolaris[11:45:37] *** nettezzaumana has quit IRC[11:45:43] <devians> or, better idea, virtualise some sort of linux and install deluge. problem solved[11:46:02] <Beket> I have 5 disks, 2 pools (2 disks rpool and 3 disks my data pool). Both pools are mirrored (2-way and 3-way respectively)[11:46:16] *** hugo has joined #opensolaris[11:46:25] <[lewellyn]> devians: i have few, if any, problems with transmission. aside from it not being my preferred client[11:47:05] <MGSsancho> sata0/1 sata-port disconnected unconfigured failed bad cables?[11:47:13] <[lewellyn]> i tend to keep a couple of machines on my network running transmission. with 100+ torrents each[11:47:27] <MGSsancho> oh yeah, howdy do everyone[11:47:55] <devians> i think out of the last 6 or so transmission versions only 1 has been non buggy enough to be usable[11:47:59] <lblume> transmission tends to forget where torrents are saved and start over. And sometimes, forget torrent completely.[11:48:15] <lblume> It's usable, but certainly annoying at times.[11:48:21] <[lewellyn]> i'm using 1.92 here just fine[11:48:31] *** TomJ has quit IRC[11:48:38] <[lewellyn]> my BIGGEST complaint about transmission is that the UI is totally different on each platform[11:48:41] <devians> read its release logs, they're always fixing some bug that caused everything to die.[11:48:49] <devians> terrible program[11:49:09] <[lewellyn]> devians: the original question was regarding rtorrent. which is far more unstable than transmission's ever been.[11:49:15] <lblume> I'm using the stock osol version. Probably not the best.[11:49:16] <[lewellyn]> (note we're in #opensolaris)[11:49:28] <[lewellyn]> lblume: it's fine. just rebuild from the spec to get something newer :)[11:49:35] <devians> hey im not saying use rtorrent. it has its nice points, but deluge beats it out hands down[11:49:48] <devians> imo if its not deluge or utorrent, go home.[11:49:49] <[lewellyn]> isn't deluge in SFE?[11:49:55] <devians> nfi[11:50:09] * [lewellyn] prefers bitcomet, but it has some FUD that's followed it for years[11:50:45] <devians> wasnt bitcomet the one that was essentially banned from all private trackers cos it was terrible at reporting stats?[11:50:55] <[lewellyn]> what's funny is that one of those bits of fud is now done by utorrent, and the hypocrites in the world think it's fine for utorrent and not for bitcomet[11:51:10] <[lewellyn]> devians: which was proven untrue many times in the past few years[11:51:24] <[lewellyn]> i think it had some issues like 4 years ago[11:51:30] <[lewellyn]> but it was a crappy client back then anyhow[11:51:52] <nikolam> [lewellyn], dId you used transmission-daemon?[11:51:55] <hugo> all the documentation mentions virt-install and I even have a manpage for it, however there's no /usr/bin/virt-install .. running snv_134. how do I get virt-install ?[11:52:02] <[lewellyn]> the biggest complaint these days is that it has an out-of-band mechanism to allow public torrents to be shared between clients[11:52:05] <MGSsancho> utorrent just has prefrential treatment for other utorrent client cuz its version 2.0+ can do udp transfers <_<[11:52:17] <[lewellyn]> nikolam: i just run the gui and control it with the remote and web interface[11:52:22] *** sstallion_ has quit IRC[11:52:47] *** dalibor has quit IRC[11:52:53] <nikolam> I control it without its gui[11:52:58] <[lewellyn]> MGSsancho: i thought 2.0 brought a mechanism for peers to share torrent files with each other, too?[11:53:15] <[lewellyn]> nikolam: as do i. i use the gui so i can scroll through the list when i feel like it :)[11:53:30] <[lewellyn]> btw, utorrent runs well in wine[11:53:38] <[lewellyn]> at least last time i tried[11:53:55] <nikolam> [lewellyn], hm, would be interesting for gui to behave like web interface in same manner, hmm[11:54:11] <nikolam> e.g. when closing gui, it stays[11:54:14] <MGSsancho> only thing i looked into was the reasone version 2.0+ got banned and it was due to that udp thing. i havent read all the release notes so i would know about any new features per say lol[11:54:18] *** panos has joined #opensolaris[11:54:20] <[lewellyn]> i know someone who was working on a webmin module for transmission. wonder if he ever got it done[11:54:20] <MGSsancho> cfgadm: Hardware specific failure: Driver ioctl failed I/O error i think im fscked[11:54:21] *** doug_ndndn has quit IRC[11:54:39] <[lewellyn]> nikolam: hide it in the notification area! ;)[11:54:39] <devians> deluge has a really nice daemon, nice webui (slows down a bit in heavy usage), well done gtk ui, a console ncurses ui thats really only there for amusements sake[11:55:20] <nikolam> [lewellyn], when, to survive user logout, say, to deamon to be default[11:55:33] <[lewellyn]> nikolam: suggest it to the transmission devs. :)[11:55:40] <nikolam> ;)[11:56:09] <nikolam> think it would use more memory or not behave avarage user expects etc etc.[11:56:24] *** sstallion_ has joined #opensolaris[11:56:53] <[lewellyn]> suggest it. worst they can say is "wtf! no!" ;)[11:57:39] <MGSsancho> i think they would add in lol and bbq in there as well[11:58:57] <nikolam> well, it will definately should not be default. etc[11:59:35] <nikolam> since I learned Deluge seems to have it[12:00:33] <nikolam> Interesting thing about bit torrent is to move ratio from app to app. and that is mostly undoable, so you stick for an app for a long time or you loose your ratio[12:01:57] *** sstallion_ has quit IRC[12:03:32] *** TomJ has joined #opensolaris[12:03:45] <MGSsancho> is zpool import supposed to hang?[12:05:04] <Fallen_Demon> Only if you haven't run "rm -rf /" yet :P[12:05:09] *** sstallion_ has joined #opensolaris[12:06:32] <MGSsancho> lol[12:06:45] <MGSsancho> so yeah swapping out calbes and stuff isnt helping[12:06:59] <MGSsancho> i may have permenently lost 1.8tb of data =([12:07:32] <causality> sounds reliable[12:08:14] <MGSsancho> thats what happens when you have a raidz pool and you add new devices to the host pool and the the raidz array[12:08:24] <MGSsancho> *and not the[12:08:54] *** held has joined #opensolaris[12:09:02] <sickness> devians: yeah, I've always just done that, but I was asking for someone that asked this to me on another irc server :)[12:09:13] <causality> MGSsancho: i was under the impression that wasn't possible?[12:09:56] *** doug_ndndn has joined #opensolaris[12:10:01] <MGSsancho> causality: http://pastie.org/904852[12:10:27] <MGSsancho> look where i added the 3 new drives ~_~[12:11:18] <causality> which are the new ones?[12:11:39] <MGSsancho> last 3[12:11:42] *** TomJ has quit IRC[12:12:12] <causality> if you're not supposed to be able to expand a raidz pool, why does it let you add disks to it[12:12:16] *** CodeWar has joined #opensolaris[12:12:21] *** CodeWar has left #opensolaris[12:12:24] *** sipior has joined #opensolaris[12:12:26] *** TomJ has joined #opensolaris[12:13:24] *** tosh has joined #opensolaris[12:13:41] <Gugge> causality, it does not[12:13:46] <MGSsancho> they got added as stripe[12:13:52] *** CodeWar has joined #opensolaris[12:14:11] <causality> that doesnt look like a stripe to me[12:14:13] <MGSsancho> and i rewrote all the data thinking i have a giant raidz array...[12:14:28] <MGSsancho> now since 2 drives are fubared........... yeah[12:14:38] <Gugge> causality: but you are allowed to add new vdevs (consisting of raidz/mirror or single disk) to a pool[12:14:46] <causality> Gugge yes i am aware of that[12:15:23] * causality still says pool when he means vdev[12:15:47] * MGSsancho does as well[12:16:20] <Gugge> but im pretty sure it requires -f to add a single disk vdev to a pool with raidz vdev(s)[12:16:44] <Gugge> so that should make people think twice before they do it[12:17:34] <MGSsancho> i can assure you it didnt help[12:17:38] <MGSsancho> ;P[12:18:07] <Gugge> well .... i would consider that a user error :P[12:18:07] *** |AbsyntH| has quit IRC[12:18:33] * causality advises against crossing the streams[12:18:35] <MGSsancho> once i discovered the error, i powered off the machine, bought a 2tb drive to move stuff over then fix my mistake.. but when i went to boot up the machine... DUN DUN DUN!!!![12:18:46] *** calumb has quit IRC[12:18:46] <causality> nonsense destructive changes should be trapped and banished[12:19:08] <causality> MGSsancho: but of course you have backups, rite[12:19:26] <sivanov> ha, my friend started to work in russian military tech corpotation, rescently they passed trough series of fusions between firms and reimplement corporate network. they gonna give internet access to emploees on terminal server clients, for security[12:19:46] <sivanov> i wonder if it wise to recommend using solaris+zones to them[12:19:48] <MGSsancho> causality: oh yeah sure i do! lol. stacks of physical photo albums![12:19:53] <[lewellyn]> madwizard: btw, i remember why i didn't put the upstream's email in. i couldn't find docs as to whether that was allowed[12:20:04] <[lewellyn]> madwizard: i think i put mine in by habit ;)[12:20:26] <sivanov> and ditch MS[12:20:39] <Gugge> MGSsancho: but cant you still read all the old files?[12:20:54] <causality> sivanov: that'd depend on what your requirements are[12:21:07] <Gugge> MGSsancho: if you can get it to see the two disks that is[12:21:07] <causality> terminal services can be locked down as much as you want[12:21:15] <MGSsancho> Gugge: sice i rewrote the data, i have 4 vdevs effectivally in stripe, 2 are missing[12:21:40] <Gugge> MGSsancho: ahh, i missed the fact that you rewrote the data[12:21:48] <MGSsancho> if i can get them working im fixing this snafu asap lol[12:22:00] *** dalibor has joined #opensolaris[12:22:44] <[lewellyn]> woo! 2 tries for "PASSED" this time![12:23:03] <sivanov> i think giving a zone to each user for running web stuff is pretty secure, but how much overhead ?![12:23:31] <MGSsancho> depends how secure the system needs to be[12:24:14] <causality> sivanov: do you want to change just for the point of changing?[12:24:16] *** Odin- has joined #opensolaris[12:25:33] <sivanov> causality, they haven't decide yet, to go win2008 way or not[12:27:10] <sivanov> and i didnt work with zones, only with vmware esx&freebsd jails[12:30:55] *** xapiens has joined #opensolaris[12:31:06] <MGSsancho> might as well go with openvms or nonstop =)[12:31:47] <sivanov> does linux have lightweight virtualization like those?[12:32:30] *** chendy has quit IRC[12:32:42] <MGSsancho> xen?[12:33:34] <sivanov> xen is pretty heavy i beleave[12:33:46] *** sickness has quit IRC[12:34:28] <Rune-> short answer, no, afaik[12:34:32] <MGSsancho> you want to look into paa virtulization and not full on stuff like what vmware does[12:34:51] <MGSsancho> and you mentioned zones and jails which is more of what you want[12:35:12] <MGSsancho> but yeah Rune- said it faster than I did lol[12:37:35] <MGSsancho> so thanks to cfgadm i found out one of my controllers doesnt support hot plugging and thats where my 2 drives are unrecoverably failed on[12:37:51] *** swift_ has joined #opensolaris[12:38:58] *** swift has quit IRC[12:39:18] <MGSsancho> can anyone reccomend a pcie 8 port sata controller thats only like jbod or something thats more usefull thann this onboard stuffs?[12:39:22] *** swift_ is now known as swift[12:41:11] *** biant92 has left #opensolaris[12:41:13] *** ewdafa has joined #opensolaris[12:43:45] *** swift has quit IRC[12:44:20] <dnaumov> MGSsancho: I have been hearing good things about http://www.supermicro.com/products/accessories/addon/AOC-SASLP-MV8.cfm[12:45:31] *** Section9 has joined #opensolaris[12:45:53] <MGSsancho> thanks dude, i think ill get it. then mayne use white electrical tape and labme my cables and drives lol[12:46:54] *** swift has joined #opensolaris[12:47:17] <sivanov> MGSsancho, i had a link to lsi 3081e based cards on alibaba.com for like $50 each[12:47:21] <sivanov> just canty find it[12:47:35] *** swift_ has joined #opensolaris[12:48:14] <MGSsancho> $315 new >.<[12:48:20] *** Bateau has quit IRC[12:49:15] *** Dagobert_ has joined #opensolaris[12:50:22] *** swift__ has joined #opensolaris[12:50:52] <sivanov> http://www.alibaba.com/product-gs/291097500/LSI_SAS_3081E_R_PCI_E.html[12:51:05] <sivanov> something like this one[12:51:47] *** Dagobert has quit IRC[12:51:47] *** Dagobert_ is now known as Dagobert[12:51:52] *** swift has quit IRC[12:52:26] <MGSsancho> wow thats cheap[12:52:29] *** swift_ has quit IRC[12:52:40] *** swift has joined #opensolaris[12:52:54] <sivanov> ha, some of them chinese has adaptec hbas that cheap lol[12:53:14] <MGSsancho> now to hunt for sas to sata adapters lol[12:53:32] <jbit> what[12:53:38] <sivanov> of qlogic double-chanel FC hbas for $450 which sell for $1400 here[12:53:49] <jbit> sas is compatiblee with sata...[12:54:52] *** swift__ has quit IRC[12:55:03] <sivanov> you ned http://www.alibaba.com/product-tp/244369601/LSI_3Ware_CBL_SFF8087OCF_06M_0.html[12:55:08] <sivanov> *need[12:55:11] <sivanov> with that card[12:55:15] <MGSsancho> jbit: i needs break out cables. cant plug 8 sata drives into those wee connectors lol[12:55:23] <causality> MGSsancho: no pci-x?[12:55:43] <MGSsancho> causality: this is a desktop amd mobo lol[12:56:17] <causality> well you could put a supermicro pci-x card in and suffer some performance loss[12:56:22] <causality> it's well supported in opensolaris[12:56:35] <MGSsancho> i have no pci-x slots lol[12:56:43] <sivanov> aoc2-sat2-mv8[12:56:46] <sivanov> cheap[12:56:51] <causality> that's the one..[12:57:11] <causality> MGSsancho: you can put a pci-x card in a pci slot[12:57:16] <jbit> MGSsancho: ah +[12:57:20] <causality> as long as nothing physically fouls it on the board[12:57:22] <MGSsancho> unless i use a dremel tool and trmm a pci-x cad to fit in a pci slot. im sure that error will end up on bash[12:57:32] <sivanov> lol no[12:57:42] <sivanov> they fit nicely[12:57:44] <causality> MGSsancho: read above..[12:57:58] <MGSsancho> hmmm[12:58:04] <Stric> MGSsancho: pci-x != pci express, fyi[12:58:21] <MGSsancho> oh that i know lol[12:58:38] <causality> MGSsancho http://www.supermicro.com/products/accessories/addon/AOC-SASLP-MV8.cfm[12:58:45] <causality> not sure if thats supported[12:58:46] <jbit> MGSsancho: the 64bit extension will just hang over the 32bit pci slot and be unused, assuming there aren't any capacitors and such in the way on the motherboard[12:58:52] <Stric> But you don't know that people normally end a sentence with "." and not "lol"...[12:59:12] *** swift_ has joined #opensolaris[12:59:15] * causality wonders if there's an echo in here[12:59:16] <Alasdairrr> SAS HBAs? LSI 3081E-R[12:59:24] <jbit> #define lol .[12:59:25] <Alasdairrr> Same chipset used in the amber roads[12:59:36] <sivanov> buy hbas on ebay btw[12:59:40] <sivanov> much cheaper[12:59:46] <Stric> jbit: more like.. #define lol , I'm a retard.[12:59:46] *** Fallen_Demon has quit IRC[13:00:00] <MGSsancho> Stric: i got in enough troble in this chan for making up nicknames instead of using tab completition, i dont want beef over ensing sentances with lol and emotes now >.<[13:00:22] <Stric> MGSsancho: If you haven't noticed, this is a mostly technical channel.. not 4chan.[13:00:49] <MGSsancho> been in here for 8 months and still haven't noticed :D[13:01:19] <Stric> That might confirm the retard part then.. ;)[13:01:37] <MGSsancho> and all the user errors + blunders[13:02:03] *** swift has quit IRC[13:02:15] <MGSsancho> using rivits on my case instead of screws, barfing all over my monitors and keyboards when i was drunk once, etc. the list is endless[13:02:18] *** swift has joined #opensolaris[13:03:17] *** swift__ has joined #opensolaris[13:03:45] *** swift_ has quit IRC[13:04:40] *** Error404NotFound has quit IRC[13:04:44] *** sickness has joined #opensolaris[13:05:33] *** MGSsancho_ has joined #opensolaris[13:05:49] *** MGSsancho has quit IRC[13:05:55] *** MGSsancho_ is now known as MGSsancho[13:06:34] <sickness> re[13:07:09] *** swift has quit IRC[13:07:09] *** swift__ is now known as swift[13:08:19] *** calumb has joined #opensolaris[13:09:26] *** scottdickson has joined #opensolaris[13:10:02] <sivanov> hmm http://www.alibaba.com/product-gs/288036265/Western_Digital_Caviar_Black_2TB_SATA2.html[13:10:38] <sivanov> do they indeed want $25-30 for that drive?[13:11:45] *** niq has joined #opensolaris[13:11:49] <palowoda> The picture is a controller.[13:12:36] <palowoda> 2TB are going for around 150/160.[13:12:44] <sivanov> looks like they messed up their prices[13:13:03] <jbit> minimum order 20... 40tbytes for $600... sounds suspicious[13:13:42] <sivanov> cheaper than lto4 media :)[13:14:19] <palowoda> Funny when I was at Fry's today a 1TB drive was 68.00 and a box of 10 1.44meg floppies was 7.00. About a thousand factor for storage. :)[13:14:47] <mui> actually[13:14:51] <jbit> palowoda: but 1tb drive isn't backwards comaptible with a c64[13:14:52] <mui> there are other disks priced similary[13:15:28] <MGSsancho> i just rememebred i bought a blu ray drive today at frys[13:15:33] <Stric> palowoda: closer to factor 1M[13:15:42] <MGSsancho> oh shit just rememebre i got a blur ray drive at frys hold on[13:15:48] *** ismell has quit IRC[13:15:57] <MGSsancho> gah browser crashing[13:16:02] *** MGSsancho has quit IRC[13:16:23] *** MGSsancho has joined #opensolaris[13:16:38] <jbit> blur ray drive.... nice :)[13:16:46] <Stric> VCD drive[13:16:49] <MGSsancho> yes auto censors porn[13:17:00] <Stric> Blur 2000 Deluxe[13:17:35] <palowoda> Ouch your right Stric.[13:17:37] <MGSsancho> pixelation is a thing of the past[13:18:28] *** ozux has quit IRC[13:19:50] <palowoda> Can you use blueray drives a a general storage device with opensolaris?[13:20:22] *** kinkie has joined #opensolaris[13:20:31] *** niq has quit IRC[13:23:40] *** anibal has quit IRC[13:24:11] <MGSsancho> test[13:24:24] <palowoda> It's not working.[13:24:38] <MGSsancho> well what do you know[13:24:43] *** bourbon has joined #opensolaris[13:24:51] *** gm152 has joined #opensolaris[13:25:17] *** dunc has joined #opensolaris[13:25:24] *** niq has joined #opensolaris[13:25:47] <MGSsancho> you can unplug the host drive in a win7 system, move around cables and plug it back in with out killing the system[13:26:19] <causality> MGSsancho: they finally fixed that? :)[13:26:28] <MGSsancho> lmao[13:26:52] <MGSsancho> yeah accidently unplugged the power to the drive when i was aqttaching my blur-ray drive ;)[13:27:01] <sivanov> what about a system drive ?[13:27:16] <MGSsancho> yeah the system drive... fell off the cables[13:28:30] <sivanov> but how can you tell that the data didnt go corrupt on ntfs?[13:29:15] <MGSsancho> all my porn is already furbared on my zfs system, you think i care if i have to reinstall win7 on my gaming system? lol[13:29:49] <MGSsancho> 1.8TB T_T[13:30:02] <sivanov> i wanted to buy a lto4 library to backup my pr0n on at least monthly basis![13:30:41] <MGSsancho> dedupe works good on porn actully[13:30:50] <sivanov> it should[13:31:46] <Stric> one april 1st many years ago, some company introduced a special porn video codec.. with "boob shading" and stuff, to make compression more efficient due to "predictable content"[13:32:08] <MGSsancho> actress, accessories, positions, sounds, scripts, etc can all be referenced[13:32:43] <sivanov> seriously, i got 8+TB of movies at home and kinda do not want to lose them,[13:33:01] <MGSsancho> from the same labs that are workin on IPv7 right? have porn metadata in the headers?[13:33:14] <sivanov> especially that our dumb russian government is going to tighten the screws around file sharing[13:33:50] <MGSsancho> ZFS to store more pr0n![13:33:54] <sivanov> to lay responsibility of filtering content on ISPs[13:34:27] *** doug_ndndn has quit IRC[13:34:31] <sivanov> i hope that will not happen[13:35:00] <palowoda> Store your p0rn up on the space station.[13:35:14] <MGSsancho> only chance you have is ssl with private certs and keep your connection busy so they have to use more resources to inspect all that traffic[13:35:52] <MGSsancho> space station would be cool. entore planet burns away and all we have left of humanity is porno[13:36:03] <palowoda> Hmm I wonder what laws do apply to the space station.[13:36:19] <sivanov> second server will cost me around $1500, and lto4 autoloader goes for about $2k on ebay[13:36:56] <sivanov> i want an appliance to drop a HDD in, backup data, pull it out and drop another[13:37:14] <MGSsancho> if you get any info on that let me know :)[13:40:29] <sivanov> in this drive cache is a problem[13:41:17] <sivanov> with it turned off the backups will be damn slow, with it the device will need a button to flush it[13:44:11] *** mroconnor has quit IRC[13:50:00] *** doug_ndndn has joined #opensolaris[13:54:00] <palowoda> Wow Adobe has a beta of flash 10.1 for Solaris. http://labs.adobe.com/downloads/flashplayer10.html[13:55:07] *** gm152 has quit IRC[13:55:11] <asyd> oh nice[13:55:18] * asyd wait for 10.1 for few years now[13:57:00] <palowoda> Larry just agreed not to fly his jet so low over the Adobe building. :)[13:57:06] <madwizard> :)[13:57:41] *** miamano has joined #opensolaris[13:59:00] <Kapsel> I just realized, that our old fileservers mainboard was born in 2003. About time I got it replaced with a new machine :P[14:00:32] *** wewek has joined #opensolaris[14:01:31] *** TomJ- has joined #opensolaris[14:01:33] *** TomJ has quit IRC[14:04:32] *** Rune- has quit IRC[14:07:11] *** ahmed-tux has quit IRC[14:08:30] *** Yu\2 has joined #opensolaris[14:11:52] *** Fullmoon has joined #opensolaris[14:11:59] *** Fullmoon has quit IRC[14:12:31] *** tsoome has quit IRC[14:12:36] *** ahmed-tux has joined #opensolaris[14:12:57] *** tsoome has joined #opensolaris[14:14:26] *** Fullmoon has joined #opensolaris[14:14:48] *** hajma__ has quit IRC[14:15:18] *** gebi has quit IRC[14:17:00] *** timsf has joined #opensolaris[14:21:30] *** TomJ- has quit IRC[14:24:45] *** RoyK has joined #opensolaris[14:28:31] <RoyK> erm - after 'pkg set-publisher -O http://pkg.opensolaris.org/dev opensolaris.org' I run 'pkg image-update' - how long does this normally take to run through?[14:30:04] <effnorwood> RoyK - about an hour on our X4500s[14:31:48] *** plat- has quit IRC[14:31:55] *** dalibor has quit IRC[14:32:04] *** skullone has quit IRC[14:34:22] *** tg has quit IRC[14:36:34] <RoyK> erm - on 111b, swap -lh shows "/dev/zvol/dsk/rpool/swap 182,2 4K 10G 10G", alright, but then, swap -sh shows 'total: 212M allocated + 644M reserved = 856M used, 12G available' - how can this be?[14:36:53] <Dagobert> Swap is Disk + Ram[14:37:22] <RoyK> I have 4GB RAM on this box[14:38:09] *** ThePickle has joined #opensolaris[14:39:32] <madwizard> RoyK: swal -l does not include physical memory[14:40:16] *** tg has joined #opensolaris[14:40:40] <madwizard> RoyK: swap -s does include RAM[14:42:42] <madwizard> Coffee[14:42:51] <madwizard> Includes swap, RAM, HDD and coffeine[14:45:26] *** betchou has joined #opensolaris[14:46:12] *** ahmed-tux has quit IRC[14:47:37] *** aruntomar has joined #opensolaris[14:47:51] <effnorwood> smrt: you should add madwizard to your coffee list[14:47:59] *** luc^ has joined #opensolaris[14:48:22] *** ahmed-tux has joined #opensolaris[14:48:31] *** plat- has joined #opensolaris[14:48:38] *** kwinz3_ has quit IRC[14:49:40] *** mroconnor has joined #opensolaris[14:50:12] *** dalibor has joined #opensolaris[14:53:57] *** calumb is now known as calAFK[14:56:12] <madwizard> Khm, first decade of XXI Century comes to an end and my surname is still misspelled at jucr.opensolaris.org[14:56:38] <madwizard> I thought a simple case of UTF-8 would be handled properly, like on any other amateur site[14:56:54] *** tsoome has quit IRC[14:57:02] <madwizard> I have however noticed that big enterprises have that tendency, to ignore anything outside ASCII[14:57:56] <RoyK> damn - downloading updates takes for ever - 200kbps is the speed I get and I'm on a 100Mbps link[14:58:06] <RoyK> 200 kBps[15:00:53] <turtle> me too! my local area connection is like "connected at 100mbps" so i must have fast internet!![15:00:59] *** not-me-guv has joined #opensolaris[15:02:24] *** effnorwood has quit IRC[15:03:29] <Alasdairrr> Sure it's not 200 kiBps? ;P[15:06:20] *** dalibor_ has joined #opensolaris[15:06:43] *** robo has joined #opensolaris[15:06:50] *** |AbsyntH| has joined #opensolaris[15:08:12] *** Mick27 has joined #opensolaris[15:10:21] *** dalibor has quit IRC[15:10:22] *** dalibor_ is now known as dalibor[15:13:02] *** Sloar has joined #opensolaris[15:13:14] <RoyK> what's sure is that it's so damn slow I probably won't be finished with it before they throw me out of here, which is a pity[15:14:37] <EisNerd> moin[15:15:10] <EisNerd> is there a dvb-c adapter (with ci slot) known to run under osol?[15:16:56] <Alasdairrr> sounds like a job for opensolaris xvm and PCI device passthrough to a Linux guest :)[15:19:48] *** niq has quit IRC[15:22:09] *** ismell has joined #opensolaris[15:24:49] *** luc^ has quit IRC[15:24:52] *** niq has joined #opensolaris[15:24:55] <EisNerd> Alasdairrr: hm I where looking after a atom based system[15:25:18] <EisNerd> Alasdairrr: so xvm isn't a solution[15:27:57] <EisNerd> RoyK: get a connection with the copper gbit to workstation as weekest part[15:29:11] <RoyK> I have sufficient bandwidth, but the opensolaris packages server is SLOW[15:29:15] <Stric> RoyK: pkg-eu-2.opensolaris.org was much faster for me[15:29:39] <RoyK> i'll try that - thanks[15:29:52] <Stric> still butt slow compared to our local debian/ubuntu mirror, but much faster than pkg-na-1 (default) and pkg-na-2[15:30:19] <ThePickle> is there a tool for monitoring the arc?[15:30:36] <Stric> depends on what kind of monitoring you're asking about[15:30:45] <ThePickle> stric: mainly memory use[15:30:48] <ThePickle> (current)[15:30:57] <Stric> get arc_summary.pl (available on internets)[15:31:03] <ThePickle> thanks very much[15:34:45] *** doug_ndndn has quit IRC[15:35:50] *** Shoggoth has quit IRC[15:37:56] *** aruntomar has left #opensolaris[15:41:44] *** dajhorn has joined #opensolaris[15:43:19] *** xapiens has quit IRC[15:43:25] *** calAFK is now known as calumb[15:43:51] *** effnorwood has joined #opensolaris[15:44:06] *** xapiens has joined #opensolaris[15:45:48] *** ziyourenxiang has joined #opensolaris[15:45:56] *** bigal has joined #opensolaris[15:46:19] *** Sloar has left #opensolaris[15:48:34] *** MaxVandervelde has joined #opensolaris[15:54:01] *** bigal has quit IRC[15:54:03] *** stewart has joined #opensolaris[15:54:12] *** doug_ndndn has joined #opensolaris[15:55:59] *** Netwolf has joined #opensolaris[15:56:58] *** kwinz3 has joined #opensolaris[15:58:14] *** hunter has joined #opensolaris[16:01:10] *** hunter has quit IRC[16:02:02] *** andersenep has quit IRC[16:09:17] *** alanc has joined #opensolaris[16:09:17] *** ChanServ sets mode: +o alanc[16:10:22] <causality> alanc, hi[16:10:35] <alanc> good morning[16:12:02] <|AbsyntH|> vietnam[16:19:03] *** setheus has joined #opensolaris[16:19:09] *** timsf has quit IRC[16:20:16] *** nikolam has quit IRC[16:20:49] *** Mdx4 has joined #opensolaris[16:21:46] *** andersenep has joined #opensolaris[16:21:52] *** andersenep has left #opensolaris[16:22:05] *** andersenep has joined #opensolaris[16:24:25] *** haggle has joined #opensolaris[16:25:05] <haggle> Hi, can anyone help with a frozen zone ??[16:25:17] <haggle> I have a feeling it's locked up on an NFS mount...[16:25:26] <haggle> desperately don't want to reboot the global zone[16:27:19] *** phimic has quit IRC[16:28:37] *** RoyK has quit IRC[16:30:24] *** galt has joined #opensolaris[16:30:34] *** phimic has joined #opensolaris[16:30:47] *** phimic has quit IRC[16:34:19] *** betchou has quit IRC[16:37:04] *** Animal-X has joined #opensolaris[16:37:52] *** LouisJB has joined #opensolaris[16:37:53] *** LouisJB has quit IRC[16:38:57] *** LouisJB has joined #opensolaris[16:39:26] *** postwait has joined #opensolaris[16:39:29] *** postwait has left #opensolaris[16:44:31] *** GhostInTheShell has quit IRC[16:44:35] *** GhostInTheShell has joined #opensolaris[16:53:01] *** raichoo has quit IRC[16:55:07] *** timsf has joined #opensolaris[16:57:01] *** Fullmoon has quit IRC[16:57:19] *** tsoome has joined #opensolaris[17:04:02] *** wewek has quit IRC[17:05:43] <joshua_> wee, I am looking forward to working dedup in 2010.05, maybe I'll get some SSDs[17:06:35] <joshua_> good for storing DDTs[17:07:06] <asyd> DDT?[17:07:12] <joshua_> dedup tables[17:07:49] <effnorwood> dedup works now but be careful of zfs destroy on a dedup file system[17:08:42] <joshua_> by that, you mean, "it works as long as you don't do normal ZFS things on it, then your zpool falls over forever"[17:08:48] *** gebi has joined #opensolaris[17:08:54] <joshua_> I think 'rm' is about as bad as 'zfs destroy'[17:09:01] *** PMT has joined #opensolaris[17:09:02] <joshua_> just a little less likely to trigger the bug[17:09:30] <joshua_> so, no, I don't think dedup "works" right now[17:09:54] <PMT> joshua_: howso?[17:11:25] <joshua_> I think that is one of the bugs gating 2010.06, though[17:11:28] *** dalibor_ has joined #opensolaris[17:11:38] <joshua_> although we do not know for sure[17:11:54] <effnorwood> http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6924824[17:12:10] <effnorwood> gotta love bugs that include "bricks the system"[17:12:11] <joshua_> yes, that[17:12:34] <taemun> how is that a new bug[17:12:37] <joshua_> it is not a new bug[17:12:46] <taemun> Submit Date 9-February-2010[17:12:47] <jbit> most servers would be rather large bricks...[17:13:01] <taemun> I saw something from at least december saying the same[17:13:11] <taemun> I spent some time looking into this[17:13:14] <taemun> as I had a brick for nine days[17:13:17] <taemun> nine super fun days[17:13:19] <tsoome> most servers dont use dedupe, that would make no sense.[17:13:25] <taemun> whilst it brought the pool up[17:13:59] <PMT> taemun: that impresses me[17:14:04] <taemun> thanks[17:14:09] <taemun> I wasn't terribly impressed[17:14:15] <PMT> i can sympathize with similar problems[17:14:31] * PMT blinks[17:14:35] <taemun> I filed a request for some percentage on the import[17:14:38] <taemun> showing progress[17:14:39] <PMT> wait, so if you delete the original holder of a dedup-ped block[17:14:48] *** ziyourenxiang has quit IRC[17:14:49] *** dalibor has quit IRC[17:14:49] *** dalibor_ is now known as dalibor[17:14:51] <PMT> it can do horrible things until you manually remove the references to that block elsewhere?[17:14:52] <effnorwood> when it happened to me, i pulled in victor latushkin and he saved my butt[17:14:54] <PMT> FTN[17:14:57] <taemun> you need to go find the DDT entry and delete it[17:15:00] <PMT> jesus[17:15:04] <taemun> or that block will just stay lost in the matrix[17:15:05] <taemun> forever[17:15:08] <PMT> how does that even ship[17:15:24] <taemun> when I say "you" I mean the filesystem[17:15:24] <tsoome> this is beta software, so stop whining[17:15:25] <taemun> lol[17:15:30] <taemun> its not a MANUAL process[17:15:31] <effnorwood> the real bug is that the destroy block all other io[17:15:42] <effnorwood> blocks*[17:15:59] <taemun> why is the DDT stored as a billion randomly written files?[17:16:02] <taemun> can someone explain that to me?[17:16:07] <taemun> why not just have a binary chunk?[17:16:10] <effnorwood> if it took oh say 9 days and didn't block all the other IO - that would be better[17:16:27] <PMT> tsoome: i understand that it's beta[17:16:38] <taemun> effnorwood: nine days of sitting there grinding isn't great, but yeah if it was functional still, that would be nice[17:16:46] <effnorwood> taemun - mark maybee zone maybe?[17:16:58] <taemun> default osol config even stops you from logging in as a non-root user in ssh[17:17:02] <taemun> thats rather annoying[17:17:09] <PMT> that's a different argument, though[17:17:11] <taemun> not being able to ssh in because zfs is destroying something[17:17:30] <taemun> effnorwood: sorry?[17:17:31] <tsoome> why you need to log as root over network?:[17:17:51] <effnorwood> think about zfs destroy and then think about cookie monster eating cookies - makes me laugh every time[17:17:52] *** lattera has joined #opensolaris[17:17:52] <taemun> because thats the ONLY user account that can log in whilst the zfs destroy is blocking IO[17:18:00] <joshua_> because you don't hvae an IPMI card, evidently[17:18:07] <tsoome> root is not user account[17:18:11] <taemun> joshua_: I do, thanks[17:18:15] <taemun> tsoome: I had to make it one[17:18:18] <PMT> yes, because osol should require an ipmi card to work while zfs destroy runs.[17:18:20] <taemun> and enable root login in sshd_config[17:18:48] <effnorwood> you can change it so that root can log in[17:18:52] <effnorwood> not that i advise it[17:18:54] <taemun> I didn't say it was an ideal scenario, it was just what .... worked[17:19:05] <taemun> I have since changed root back to a role, and edited the sshd_config back[17:19:06] <effnorwood> just remove the role bit from user_attr in etc[17:19:21] <taemun> theres a rolemod command for it @ effnorwood[17:19:54] <effnorwood> taemun - remembering the rolemod command is somehow harder for me that editing that file :)[17:19:55] <taemun> I think zfs dedup should have an option for FILE based hashing, not just BLOCK based hashing[17:20:11] <taemun> haha @ effnorwood ... I just googled a lot[17:20:59] <taemun> still ... no brilliant reasoning for why the DDT is kept as a billion random blocks :P[17:21:24] <tsoome> ask from zfs-discuss:P[17:21:32] *** twisti has quit IRC[17:22:18] <joshua_> what else would it be taemun ?[17:22:23] *** twisti has joined #opensolaris[17:22:30] <taemun> a large streaming binary chunk?[17:24:28] <effnorwood> or a giant slab[17:24:33] <taemun> ^[17:25:29] <taemun> 152GB of on-disk dedup'd blocks on my /tank[17:25:30] <taemun> DDT-sha256-zap-duplicate: 310684 entries, size 578 on disk, 380 in core[17:25:30] <taemun> DDT-sha256-zap-unique: 1155817 entries, size 2438 on disk, 1783 in core[17:25:39] *** devians[w] has quit IRC[17:25:44] <taemun> so 1466501 blocks[17:25:45] <effnorwood> we could write some cool way to handle the memory allocator in the slab - we could call it say a bonwick allocator[17:25:52] *** devians[w] has joined #opensolaris[17:26:01] <taemun> which are seemingly randomly written to the pool[17:26:04] *** devians[w] has quit IRC[17:26:05] <taemun> from looking at iopattern[17:26:06] <Triskelios> effnorwood: LOL[17:26:07] <taemun> iostat[17:26:07] <taemun> etc[17:26:19] *** devians[w] has joined #opensolaris[17:27:02] <Triskelios> taemun: they're not handled differently from other attributes atm[17:27:12] *** _matze has joined #opensolaris[17:27:16] <taemun> Triskelios: maybe they should be :o[17:27:16] <_matze> re[17:27:36] <Triskelios> taemun: I agree[17:27:39] <taemun> \o/[17:27:46] <taemun> well we have two people on the side of sanity[17:27:52] <taemun> lets go storm Oracle HQ[17:31:59] <effnorwood> careful - larry has a mig[17:32:09] <effnorwood> uncle larry*[17:32:12] <taemun> lol[17:32:21] *** nikolam has joined #opensolaris[17:34:18] *** smrt has quit IRC[17:34:27] *** smrt has joined #opensolaris[17:34:32] *** Reepicheep has joined #opensolaris[17:34:33] *** sndcrb has joined #opensolaris[17:34:47] *** doug_ndndn has quit IRC[17:34:56] <Triskelios> a dedicated L2ARC for the DDT would go a long way toward fixing the problem without having to muck with the allocation semantics to reserve fixed blocks...[17:35:16] <taemun> orrrrr[17:35:19] <taemun> persistent l2arc[17:35:25] <taemun> with secondarycache=on[17:35:33] <taemun> would kinda achieve the same thing[17:35:43] <taemun> although, I guess that would mean that the DDT is still also in the main pool[17:35:51] <taemun> **with secondarycache=metadata[17:36:28] *** dajune has quit IRC[17:37:05] *** Section9_ has joined #opensolaris[17:37:24] *** miamano_ has joined #opensolaris[17:37:33] *** farsan_ has joined #opensolaris[17:37:41] *** nikolam has quit IRC[17:37:41] *** Mick27 has quit IRC[17:37:41] *** ahmed-tux has quit IRC[17:37:41] *** miamano has quit IRC[17:37:41] *** dunc has quit IRC[17:37:41] *** bourbon has quit IRC[17:37:42] *** Section9 has quit IRC[17:37:42] *** ingenthr1 has quit IRC[17:37:42] *** leito has quit IRC[17:37:42] *** richb has quit IRC[17:37:42] *** hrist has quit IRC[17:37:42] *** Killsudo has quit IRC[17:37:42] *** micols has quit IRC[17:37:42] *** jaywalk has quit IRC[17:37:42] *** x58 has quit IRC[17:37:42] *** trygvis has quit IRC[17:37:42] *** Tempt has quit IRC[17:37:42] *** farsan has quit IRC[17:37:42] *** evalenci has quit IRC[17:37:42] *** lisppaste3 has quit IRC[17:37:42] *** Cobi_ has quit IRC[17:37:43] *** iceq has quit IRC[17:37:43] *** mikaeld has quit IRC[17:37:43] *** tomocha6 has quit IRC[17:37:43] *** EisNerd has quit IRC[17:37:50] <taemun> I guess that you wouldn't need redundancy on your theoretical DDTcache, as you could always regenerate it from the source data[17:38:00] <Triskelios> yeah, it's strictly a cache[17:38:37] *** nikolam has joined #opensolaris[17:38:37] *** Mick27 has joined #opensolaris[17:38:37] *** ahmed-tux has joined #opensolaris[17:38:37] *** dunc has joined #opensolaris[17:38:37] *** bourbon has joined #opensolaris[17:38:37] *** Section9 has joined #opensolaris[17:38:37] *** ingenthr1 has joined #opensolaris[17:38:37] *** leito has joined #opensolaris[17:38:37] *** richb has joined #opensolaris[17:38:37] *** hrist has joined #opensolaris[17:38:37] *** Killsudo has joined #opensolaris[17:38:37] *** micols has joined #opensolaris[17:38:37] *** jaywalk has joined #opensolaris[17:38:37] *** x58 has joined #opensolaris[17:38:37] *** trygvis has joined #opensolaris[17:38:37] *** Tempt has joined #opensolaris[17:38:37] *** evalenci has joined #opensolaris[17:38:37] *** lisppaste3 has joined #opensolaris[17:38:37] *** Cobi_ has joined #opensolaris[17:38:37] *** iceq has joined #opensolaris[17:38:37] *** mikaeld has joined #opensolaris[17:38:37] *** tomocha6 has joined #opensolaris[17:38:37] *** EisNerd has joined #opensolaris[17:38:37] *** farmer.freenode.net sets mode: +o Tempt[17:38:38] *** ingenthr1 is now known as ingenthr[17:38:49] <tsoome> spam:D[17:39:14] <taemun> Triskelios: so what does your ddtcache do that l2arc doesn't presently?[17:39:30] <taemun> other than exclude other kinds of metadata / data from the device?[17:39:38] <Triskelios> not evict the DDT, ever?[17:39:58] *** ahmed-tux has quit IRC[17:40:24] *** Section9 has quit IRC[17:45:22] <taemun> oki[17:48:03] *** timsf has quit IRC[17:49:52] *** doug_ndndn has joined #opensolaris[17:50:10] <taemun> http://mail.opensolaris.org/pipermail/zfs-discuss/2010-April/039650.html[17:50:16] <taemun> nice html attachment there[17:50:47] *** Mdx4 has quit IRC[17:51:11] *** timsf has joined #opensolaris[17:51:30] <Triskelios> taemun: the DDT is an AVL tree[17:52:10] *** gerard13 has joined #opensolaris[17:52:20] * taemun goes to wiki AVL tree[17:54:23] *** _matze has quit IRC[17:54:25] *** macros73 has joined #opensolaris[17:54:36] *** hunter has joined #opensolaris[17:54:58] *** randw has quit IRC[17:55:11] *** Erwann has quit IRC[17:55:31] <taemun> Triskelios: does that preclude clumping elements together?[17:56:05] *** randw has joined #opensolaris[17:58:19] <Triskelios> taemun: even if entries in the same transaction are allocated sequentially, they're likely not be read sequentially since you get some churn with rebalancing the tree[17:58:20] *** comay has joined #opensolaris[17:58:40] *** ChanServ sets mode: +o comay[17:59:06] <taemun> well if the blocks allocated were small enough, say 512kb, then throwing out the elements you don't want shouldn't have that much overall loss[17:59:16] *** syamajala has joined #opensolaris[17:59:20] <taemun> surely that'd be quicker than lots of 512b random seeking[17:59:20] *** hunter has quit IRC[18:00:29] *** hunter has joined #opensolaris[18:00:33] <Triskelios> I'm not sure I follow[18:01:49] <taemun> I was suggesting that if you group 1024 elements together (would actually be more like 1940 elements, if that 270 byte number is correct), maybe when you're reading them you don't want to read all 1940 of them, in that order, but that at least SOME of the other elements inadvertently read would be useful[18:02:48] <taemun> unless you mean *really* not read sequentially, upon which maybe none of the elements are relevant[18:02:52] *** disharmony has quit IRC[18:02:55] <taemun> I'm not sure as to the structure[18:03:28] <Triskelios> it's a balanced-ish binary tree[18:03:58] <taemun> does the balancing completely randomise the chain though?[18:05:24] *** randw has quit IRC[18:05:37] *** Section9_ has quit IRC[18:07:05] *** disharmony has joined #opensolaris[18:09:08] *** Mick27 has quit IRC[18:10:50] *** niq has quit IRC[18:12:02] <Triskelios> it would favour the last inserted entry... after a while the fragmentation would get pretty high[18:12:14] <taemun> mm[18:12:45] *** rdub has joined #opensolaris[18:13:06] <taemun> for small sized DDT's you'd still be better off having the whole thing as a slab somewhere, and read in everything all in one go for your destroy[18:13:13] <taemun> (strore all the blocks next to each other)[18:13:22] <taemun> then just read the _whole_ thing, and organise in ram[18:13:27] *** not-me-guv has quit IRC[18:13:32] <taemun> ofc, this is predicated on being able to fit it in ram[18:13:40] <taemun> but I did say, for small sized DDT's[18:13:43] *** freitas has joined #opensolaris[18:15:12] *** cato_ has joined #opensolaris[18:15:42] *** cato_ is now known as Cato1[18:17:08] *** rdub has left #opensolaris[18:17:35] *** jollyd has joined #opensolaris[18:17:38] *** jollyd has left #opensolaris[18:19:29] <axisys> you guys use any reporting tool that shows the aggregated availability of all systems taht you manage?[18:19:30] <Triskelios> taemun: keeping a copy in ARC would be sufficient for reads, but writing still has to go through the normal allocator unless you have a special case...[18:19:46] *** bubbva has joined #opensolaris[18:20:16] <axisys> i like to explore some tools.. the report will be in excel format for my manager[18:20:21] <axisys> unfortunately![18:24:54] *** LouisJB has quit IRC[18:25:49] *** dalibor_ has joined #opensolaris[18:27:11] *** milek_ has quit IRC[18:28:03] *** dalibor has quit IRC[18:28:03] *** dalibor_ is now known as dalibor[18:29:53] *** lblume has quit IRC[18:31:21] *** Doc has joined #opensolaris[18:31:45] <Doc> does OpenSolaris include Oracle now rather than MySql?[18:32:29] <Triskelios> Doc: uh... no. is Oracle DB even redistributable?[18:32:32] <brandon> OpenSolaris is open source (except for a few bits i think)[18:32:59] <Doc> triskelios: do i look like a DBA?[18:33:28] <tsoome> looking more like normal troll:P[18:35:16] <Spencer_tt> you look like a MS(diss-ambiguation ) document Doc ;)[18:35:28] <Doc> tsoome: thanks. i try :)[18:35:53] <Doc> havent visited here for a while, so i though i should drop in and say hello[18:36:00] <tsoome> at serious side, s10 includes both mysql and postgres, havent checked about opensolaris[18:36:09] <madwizard> both[18:36:11] <tsoome> probably the same[18:36:33] <madwizard> both[18:36:38] <Spencer_tt> I couldn't get mysql to work in opensolaris[18:36:48] <madwizard> I could[18:37:24] <Spencer_tt> actually I'm telling a lie - I didn't do anything useful with mysql in osol[18:37:30] *** mikefut has quit IRC[18:37:42] <madwizard> I did,d. Disabled it.[18:37:52] <Spencer_tt> me two.[18:37:57] <Triskelios> hey, it's not installed by default[18:38:25] <madwizard> True[18:38:45] *** haggle has quit IRC[18:39:00] <madwizard> coffee'[18:40:05] <Spencer_tt> it would do the Sun_Oracle marketing a favour to have it on by default[18:42:34] *** tg has quit IRC[18:43:10] *** corretico has joined #opensolaris[18:43:45] *** tg has joined #opensolaris[18:45:10] *** twilling has joined #opensolaris[18:46:53] <alanc> OpenSolaris includes mysql, postgresql, and sqlite[18:47:01] *** ozux has joined #opensolaris[18:47:37] *** relling has joined #opensolaris[18:47:59] <Spencer_tt> by default?[18:51:38] *** sipior has quit IRC[18:52:31] <ant> wow, first time using XFS and w/in a week lost a xen image to some bizarre corruption[18:52:44] <ant> once you've had the zfs, you just can't go back[18:52:51] <ant> (and really, why would you?)[18:52:59] <Stric> xfs has defrag ;)[18:53:20] <ant> well, it's bitching about the superblock[18:53:20] *** gdamore has joined #opensolaris[18:53:50] <ant> simple, though annoying to fix[18:53:51] *** gdamore has joined #opensolaris[18:54:02] <ant> I'm just spoiled[18:54:40] <ant> I hope these guys will buy off on my reccomendation to use OpenSolaris as a share storage platform[18:55:32] <wereHamster> ant: you've obviously never seen zfs bitching about superblocks (or whatever they are called in zfs)...[18:55:40] <tsoome> uberblock[18:56:06] <tsoome> and no, i havent seen it as well. i think i never will, as i wont do stupid things with my hardware.[18:56:20] <wereHamster> Oh, I remember, zfs panics the kernel instead of bitching about corrupted uberblocks..[18:56:31] <wereHamster> way to go, zfs..[18:56:42] <eviljames> heheh I thought that was freebsd specific[18:56:45] <tsoome> ? you prefer wiping your disk instead?[18:57:49] <wereHamster> no, I'd prefer if zfs told me that the pool is corrupt instead of crashing the box[18:58:02] <tsoome> maybe you wanna learn to read?:P[18:58:12] <wereHamster> read what?[18:58:14] <tsoome> man zpool and search to panic[18:58:22] <ant> wereHamster: well, I've seen zfs do some scary bitching (with many Terabytes of data) but it's proven itself reliable by not losing data[18:58:26] <tsoome> failmode=wait | continue | panic[18:58:49] <tsoome> my kernel is not more important than my data.[18:58:55] <wereHamster> right. Is there a way to set that option when importing a pool?[18:58:56] <Stric> tsoome: isn't that what it should do when a pool goes beyond last redundancy..[18:59:23] <Stric> not what to do if stuff is corrupt..[18:59:37] <tsoome> you really should try man zpool before asking;)[18:59:46] <tsoome> and yes, you can[18:59:56] *** twisti has quit IRC[19:00:10] <wereHamster> oh, and failmode=wait is the default, and I have never touched that option. So I don't see why it should panic[19:00:23] <Stric> because you're not talking about the same thing[19:00:51] <tsoome> wait is for connectivity[19:01:10] <Stric> failmode isn't what to do when the import code gets a segfault[19:01:16] <Stric> it's when code is working, but the disks are not[19:01:37] *** twisti has joined #opensolaris[19:01:42] <tsoome> ah, ye seems you are quite correct Stric[19:01:42] <Stric> so no, you're not talking about the same thing.[19:02:01] <Stric> so both of you, ignore what the other one has said for the last 5 minutes[19:02:21] <tsoome> .oO use the source, luke.[19:02:22] *** fleyta has joined #opensolaris[19:02:23] <tsoome> :D[19:02:24] * ant observes that complimenting zfs seems to have set off a s**tstorm[19:02:28] <ant> *sigh*[19:02:33] <fleyta> bubbva: Are you available?[19:02:40] <tsoome> ant not really:D[19:02:41] <ant> mini-s**tstorm[19:02:45] <ant> :)[19:02:51] <bubbva> fleyta: in a meeting right now[19:02:58] <tsoome> its another detail revealed;)[19:03:14] <ant> let's just say that I'm *really* missing zfs right now[19:03:25] <ant> (and osol)[19:03:36] <fleyta> bubbva: When you're done please consider http://mail.opensolaris.org/pipermail/request-sponsor/2010-April/002671.html[19:03:43] <tsoome> tbh, unless the usage pattern really will hate zfs, i wont replace it with anything.[19:03:53] <wereHamster> ant: it wasn't you compliment to zfs, but your bitching about other filesystems. Each filesystem has its share of bug, and that's no different with zfs[19:03:59] <ant> going from osol to gentoo is a little hairy[19:04:08] <tsoome> checksum and self healing is way too nice to be given up[19:04:18] <ant> wereHamster: correct, I definitely agree[19:04:53] <wereHamster> ant: also, why do you use xfs when there's btrfs?[19:05:20] <ant> well, this is a new environment that I'm walking into, new job[19:05:22] <fleyta> wereHamster: XFS is mature, btrfs is not.[19:05:33] <tsoome> if you wanna loose your data, why just feed disk into shredder:P[19:05:39] <ant> btrfs is, at best, sketchy[19:05:44] <tsoome> :D[19:05:48] <wereHamster> I've never had btrfs lose my data, which I can't say about zfs..[19:05:59] *** ganbold_ has quit IRC[19:06:02] <tsoome> how did you lost it btw?[19:06:17] * ant peers down slippery slope, and quickly turns back[19:06:26] <fleyta> tsoome: XFS works just fine and may be a good alternative to UFS or XFS on small systems.[19:06:46] * alanc wonders how many people do "pkg install xfs" on OpenSolaris and are disappointed to find they got a font server instead of a file system[19:07:02] <tsoome> alternative, sure. i really dont know about good:)[19:07:02] <Stric> alanc: your fault! :)[19:07:06] <wereHamster> I don't remember the details, but basically all I did was update the BE and reboot the server. Then the whole pool was gone[19:07:26] *** TomJ has joined #opensolaris[19:07:26] <bubbva> hi fleyta, I see your request, but I'm not the appropriate person to take taht request. Also, I'm too busy with my own tasks at this moment to handle any sponsored requests. sorry.[19:07:27] <alanc> Stric: not my fault! The X Consortium used the name years before SGI...[19:07:34] <ant> lol[19:07:41] <Stric> alanc: you're the messenger.. boom.[19:08:19] <tsoome> sata/sas?[19:08:23] <turtle> wereHamster: so did you try to do anything else like plug it in to another system and import it or did you just give up at that point?[19:08:36] *** pjfloyd has joined #opensolaris[19:09:18] <ant> xfs may very well be all well and good, but on this small system, if zfs were available to me right now, I'd bet you a cargo ship full of bananas I would not be here taking up y'alls valuable time bitching about xfs[19:09:58] <Stric> mm.. bananas..[19:10:17] <ant> and I'm done with said bitching[19:12:01] *** syamajala has quit IRC[19:13:01] *** derchris has quit IRC[19:13:06] *** derchris has joined #opensolaris[19:13:46] *** |AbsyntH| has quit IRC[19:15:54] *** timsf has quit IRC[19:16:06] *** fleyta has quit IRC[19:16:14] *** hajma has joined #opensolaris[19:18:42] *** datafirm has joined #opensolaris[19:22:29] *** raichoo has joined #opensolaris[19:24:00] *** calumb has quit IRC[19:24:34] *** kevku has joined #opensolaris[19:25:28] *** MattMan has left #opensolaris[19:34:41] *** doug_ndndn has quit IRC[19:35:12] *** tg has quit IRC[19:35:41] *** cybernd has joined #opensolaris[19:36:00] <MGSsancho> some one say bannas? :)[19:36:06] *** tg has joined #opensolaris[19:36:49] <CosmicDJ> monkey[19:37:05] <MGSsancho> bannas are arsome[19:37:19] <MGSsancho> as in arson and of course not awsome lol[19:39:18] *** gswain has joined #opensolaris[19:39:47] *** gswain has left #opensolaris[19:41:05] *** LouisJB has joined #opensolaris[19:41:50] *** relling has quit IRC[19:45:51] *** tg has quit IRC[19:46:43] *** tg has joined #opensolaris[19:48:35] *** doug_ndndn has joined #opensolaris[19:49:54] <Triskelios> sstallion_work: ping?[19:54:12] *** sstallion_ has quit IRC[19:55:58] <CIA-21> Anurag S. Maskey <Anurag.Maskey at Sun dot COM>: 6938870 NWAM phase 1 does not automatically connect to hidden wifi[19:56:09] *** plat- has quit IRC[19:56:12] *** Animal-X has quit IRC[19:56:57] <nikolam> do you get X dead too, when switching VTs after some time without use?[19:57:15] <nikolam> restarting X, 131 here, not dead[19:57:35] *** plat- has joined #opensolaris[19:58:09] *** melbogia has quit IRC[19:59:27] *** held has quit IRC[20:00:34] *** fOB has joined #opensolaris[20:01:19] *** sstallion has joined #opensolaris[20:07:02] *** Mdx4 has joined #opensolaris[20:08:15] *** gavino has joined #opensolaris[20:08:23] <gavino> whats up bosses[20:09:33] *** Cato1 has quit IRC[20:09:39] *** popster has joined #opensolaris[20:09:48] * [lewellyn] looks at gavino[20:10:30] *** melbogia has joined #opensolaris[20:11:18] *** RoyK has joined #opensolaris[20:11:55] <gavino> loosk at lew[20:12:24] <[lewellyn]> i notice you're ban evading again[20:12:24] *** hunter has quit IRC[20:12:56] <popster> Does anyone know how much a used X25-M 160GB is worth US?[20:13:19] <[lewellyn]> ebay and craigslist seems to be a good guide[20:13:50] <[lewellyn]> hm. i wish i had copied my ssh keys onto this host. i can't remember any passwords :P[20:14:51] <[lewellyn]> oh hey. i bet it's because all the accounts were disabled because i log into it so rarely! :P[20:15:33] <popster> Are there anyways to check wear levels on a ssd? I guess thats what I am more worried about.[20:17:04] <sstallion_work> Triskelios: whats up ?[20:17:06] <MGSsancho> use the itnel disc utility that comes with the ssds?[20:17:16] *** relling has joined #opensolaris[20:17:55] <Triskelios> sstallion_work: how far did you get with rewriting intrd?[20:18:21] <sstallion_work> Triskelios: not as far as I wanted - basically had the daemonization written and was beginning to poll the kstats[20:18:28] <sstallion_work> why do you ask?[20:18:57] <sstallion_work> My queue has been filling up lately so I had to put it to the side[20:19:39] *** bourbon has quit IRC[20:19:54] <Triskelios> sstallion_work: I was going to do my own implementation this week - have the kstat stuff already, fiddling with how to structure the stats...[20:20:06] <sstallion_work> ahh very cool[20:20:13] <sstallion_work> go for it - if you need a reviewer I can lend a hand[20:20:26] <gavino> zfs question ok?[20:20:32] <Triskelios> there's a internal-only test suite that would be nice to also redo[20:20:34] <sstallion_work> Once I get back to it, I need to finish up efe and re before going back to intrd[20:20:55] <sstallion_work> gavino: feel free to ask; answers are optional.[20:21:12] <sstallion_work> I have a CR created already - need it?[20:21:24] *** held has joined #opensolaris[20:21:50] <sstallion_work> Triskelios: recently, a daemon() function has been integrated - I'd look into using that rather than the old fashioned daemonize() approach[20:22:01] <Triskelios> sstallion_work: yes, I am using daemon()[20:22:08] <sstallion_work> good ;)[20:22:54] <Triskelios> and yeah seeing the existing webrev would be nice[20:23:13] <gavino> if I have a box with 3 arrays and 2 internal disks[20:23:16] <sstallion_work> http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6898208[20:23:28] <sstallion_work> I don't have a webrev handy, but there is a Bug ID[20:23:33] <gavino> the disks except for 2 inernal are 320 scsi on dell arrays[20:23:43] <nikolam> Bah, I missed important appointment today because X died 2 times today after opening nbook (blank screen and pressed VT1 to get it back) And Sunbird with alarm was not working[20:23:47] <sstallion_work> heh, forgot I had some fun with the title: "Deperlification of the interrupt distribution daemon (intrd)"[20:23:50] <nikolam> I am getting sick of this[20:24:00] <nikolam> 131[20:24:12] <gavino> when I install opensol I would perhaps put the os on the 2 internal 36g drives and then zfs the rest of the 146g 320s? I want to end up with a fast nfs server[20:24:23] <sstallion_work> gavino: that would be a good start[20:24:35] <gavino> not sure how I would see the two internal vs rest fo drives in opensol setup?[20:25:06] <Triskelios> sstallion_work: thanks, alanc already linked me the Bugster CR, was just wondering about the webrev[20:25:21] <gavino> I am thinking about using freebsd zfs but I think I have to recompile thier kernel since my box has 4g of ram so pain in the ass[20:25:36] <gavino> might get fett wet with opensol[20:25:49] <sstallion_work> Triskelios: I didn't get around to generating one, and I ended up whacking that branch a couple of months ago. It wouldn't take more than an evening to get back to where I was[20:26:02] <bball> how do I add kernel command line options to the grub boot menu ?[20:26:13] <sstallion_work> the big thing I finished was cleaning up the SMF bits and added the necessary changes to ON to build and package it properly.[20:26:38] <bball> I'm editting the menu.lst file but there is a comment saying it was added by bootadm and I shouldn't edit it[20:26:39] <Triskelios> sstallion_work: don't worry about it then[20:26:42] <sstallion_work> Triskelios: np[20:27:43] *** Spencer_tt1 has joined #opensolaris[20:28:22] *** sstallion_work has quit IRC[20:28:22] *** sstallion_work has joined #opensolaris[20:30:30] *** popster has left #opensolaris[20:30:31] <sstallion_work> bball: its okay to edit[20:30:40] <sstallion_work> bball: at least for this change - don't make any structural changes[20:30:48] *** Spencer_tt has quit IRC[20:30:57] *** ChanServ sets mode: +o sstallion_work[20:31:05] <Triskelios> I wonder if it would be better for intrd to match on class "interrupts" rather than module "pci_intrs" for this, although the fact that it uses pcitool means it'll need to be updated if ever a non-PCI bus is added[20:31:07] <sstallion_work> ahh there we go[20:31:47] <sstallion_work> its been a while since I've been in there - I'm surprised its using pci_intrs[20:32:02] <sstallion_work> it may be that PCI is the only one that supports re-mapping interrupts[20:32:05] <sstallion_work> (that would be my guess)[20:32:16] *** RoyK has quit IRC[20:32:20] <sstallion_work> ISA interrupts are fixed (and SBus is no longer supported)[20:33:03] *** held has quit IRC[20:33:46] *** psychicist has joined #opensolaris[20:33:56] <sstallion_work> (INOW, I take that back, pci_intrs makes perfect sense :P)[20:36:23] <Triskelios> yeah, the pcitool interfaces are likely not going to be extensible to other buses[20:36:27] *** fleyta has joined #opensolaris[20:36:35] *** dajune has joined #opensolaris[20:37:36] *** miki has joined #opensolaris[20:41:42] <fleyta> OFED RDMA CM API TOFU![20:42:13] <fleyta> WTF? Who needs this?[20:43:10] <fleyta> Ahhhh, Lustre.[20:43:31] <MGSsancho> LustureFS![20:43:41] *** held has joined #opensolaris[20:44:46] <fleyta> New world: Opensolaris/ARM nodes fed from an IB switch.[20:46:43] *** sstallion has quit IRC[20:49:23] *** mikefut has joined #opensolaris[20:50:39] *** e1kg has quit IRC[20:52:04] *** popster has joined #opensolaris[20:52:45] *** Yu\2 has quit IRC[20:53:13] *** InTheWings has joined #opensolaris[20:53:39] <popster> I exported a pool because I had a degraded state and hotswapped a drive but I can't zpool replace the new drive in the same physical location as it doesn't detect a drive is present. How do I detect the new drive?[20:53:42] *** rcsheets`osu has joined #opensolaris[20:54:01] <rcsheets`osu> what do ON and IPS mean (in the topic)?[20:54:23] *** held has quit IRC[20:55:51] <popster> I believe that it means the latest packaged build is 134 and development is currently ensuing on build 136.[20:55:59] <CIA-21> Thomas Haynes <Thomas.Haynes at Sun dot COM>: 4799074 NFS authentication cache needs some sort of TTL[20:56:02] <CIA-21> Mark J Musante <Mark.Musante at Sun dot COM>: 6940961 back out change to media.c - breaks on disks without a label[20:56:03] <CIA-21> Mark Shellenbaum <Mark.Shellenbaum at Sun dot COM>: 6938395 zfs_zaccess Deadlock: cycle in blocking chain when running mirror_stress tests[20:56:04] <CIA-21> Pramod Gunjikar <Pramod.Gunjikar at Sun dot COM>: PSARC 2009/421 Open Fabrics User Verbs (OFUV) - primary kernel components, 6863653 Need in kernel OFED RDMA CM API[20:59:35] *** twilling has quit IRC[21:00:13] <Triskelios> it means the source for the OS/Net portion of build 136 is available, actually[21:00:57] <rcsheets`osu> ok[21:02:00] <popster> Thank you for clarifying, wasn't exactly sure.[21:03:14] <rcsheets`osu> thanks[21:03:14] *** freitas has left #opensolaris[21:04:26] *** fleyta has quit IRC[21:06:49] <[lewellyn]> and ON is what most people tend to care about the version of ;)[21:07:47] *** dajune has quit IRC[21:07:56] <bball> how can I list which devices (/dev/dsk/*) a particular zpool is using ?[21:08:52] <[lewellyn]> zpool status?[21:09:10] <bball> ty[21:10:19] *** Kedare has joined #opensolaris[21:11:09] *** vmlemon has quit IRC[21:12:58] *** sstallion has joined #opensolaris[21:13:55] <bball> how do I figure out the /dev/dsk/ path to a new disk that I just added to my system ?[21:14:22] *** dalibor has quit IRC[21:15:21] *** doug_ndndn has quit IRC[21:15:39] <Triskelios> bball: iostat -En[21:18:06] *** ingenthr has quit IRC[21:18:49] *** fOB has quit IRC[21:21:50] *** dajune has joined #opensolaris[21:23:19] *** Kush- has left #opensolaris[21:23:53] *** ma3x_ has quit IRC[21:28:39] *** disharmony has quit IRC[21:29:18] *** jfisc has quit IRC[21:31:05] *** doug_ndndn has joined #opensolaris[21:33:28] *** jayk- has joined #opensolaris[21:34:37] <jayk-> i have a zfs pool where aclmode and aclinherit are set to SOURCE local, and i want to change it so that SOURCE says default. how would i do that?[21:35:13] <Triskelios> zfs inherit aclmode[21:35:48] <jayk-> k[21:36:46] *** ma3x_ has joined #opensolaris[21:38:07] *** thana has quit IRC[21:38:07] *** forquare-macbook has joined #opensolaris[21:38:10] *** thana has joined #opensolaris[21:38:11] *** thana has joined #opensolaris[21:45:45] *** Garen has quit IRC[21:46:49] *** Garen has joined #opensolaris[21:47:46] *** MaxVandervelde has quit IRC[21:49:10] <bball> does anyone else have problems with "pkg" stalling during an install ?[21:49:41] <bball> I'm uisng an http_proxy[21:49:49] *** leito has quit IRC[21:52:24] *** MaxVandervelde has joined #opensolaris[21:54:08] *** Mdx4 has quit IRC[21:54:09] *** hajma has quit IRC[21:54:50] <sickness> bball: I think it's normal, pkg is a mess... are you using IPS 134 ?[21:55:17] <sickness> always use latest release, bugs are fixed at every release[21:55:48] *** hajma has joined #opensolaris[21:56:08] <CIA-21> David Hollister <David.Hollister at Sun dot COM>: 6936425 Maguro+SAS2: system stuck while running IOs with thread 'txg_quiesce_thread'[21:57:14] *** fastsolars has joined #opensolaris[22:00:51] <richlowe> if it's normal, I haven't see a bug against it[22:01:41] <alanc> there was just mail to pkg-discuss about the -na-2 mirror being down, which could cause slowness if you're trying to use it[22:02:03] * [lewellyn] kicks lpq[22:02:27] <richlowe> alanc: Yeah, saw that. but the mirror quality stuff should start ignoring it pretty quickly[22:02:35] <bball> had do try about 10 times to get wireshark installed[22:02:37] *** fastsolars has quit IRC[22:02:40] <[lewellyn]> richlowe: if you configured it properly[22:02:46] *** dajhorn has quit IRC[22:02:49] <[lewellyn]> which, of course, is underdocumented[22:02:49] <richlowe> alanc: (notice that Joanie didn't find a bug againstn it, either)[22:02:49] <bball> would usually stall between 8-12 MB of download[22:03:36] <richlowe> anyway, I'm just trying to encourage folks to file useful bugs if things go wrong.[22:03:48] <richlowe> rather than just think it's normal because, "pkg is a mess" and leave them to never get fixed.[22:04:05] *** kevku has quit IRC[22:04:05] <alanc> always a good plan (filing bugs that is)[22:04:28] *** andrnils_ has joined #opensolaris[22:04:37] <bball> is there a way to capture loopback network traffic ? for example I have an NFS mount[22:04:52] *** fastsolars has joined #opensolaris[22:04:52] <bball> where the server is my local machine[22:05:08] <bball> but wireshark doesn't see the traffic.. I presume because the packets are being short circutied through the networking stack[22:05:10] *** cocacola has joined #opensolaris[22:05:44] <Triskelios> snoop works on lo0 for me[22:06:17] *** andrnils has quit IRC[22:07:52] <[lewellyn]> loopback's always been special on solaris, for better or for worse[22:08:09] <bball> ok.. can see some stuff with snoop.. doesn't seem to be showing me everything[22:08:09] *** skullone has joined #opensolaris[22:08:13] *** panos has quit IRC[22:08:24] *** jfisc has joined #opensolaris[22:08:36] *** Beket has quit IRC[22:09:38] *** dalibor has joined #opensolaris[22:10:10] *** anibal has joined #opensolaris[22:12:08] *** radsy has joined #opensolaris[22:15:01] *** wrapster has quit IRC[22:17:15] *** syamajala has joined #opensolaris[22:17:37] *** tlyu has quit IRC[22:20:45] *** ingenthr has joined #opensolaris[22:20:50] *** hsp has quit IRC[22:21:59] *** Animal-X has joined #opensolaris[22:24:59] <Kapsel> how do I see what ip's my machine is doing traffic to/from? is there some kind of trafshow?[22:26:24] *** logic855 has quit IRC[22:28:03] *** mroconnor has quit IRC[22:29:13] *** scottdickson has left #opensolaris[22:30:25] *** q5sys has joined #opensolaris[22:32:34] *** gjl_ has joined #opensolaris[22:33:35] *** muthu has joined #opensolaris[22:34:02] <muthu> hi guys[22:34:16] <muthu> is ficon supported in opensolaris[22:35:55] <Triskelios> what's ficon?[22:36:20] <muthu> ibms fibre connectivity[22:36:41] <[lewellyn]> i'd imagine so, since there are storagetek devices which do ficon[22:36:55] <[lewellyn]> but i'm not sure, since i don't have the budget for ficon ;)[22:37:01] <muthu> :)[22:37:11] *** tlyu has joined #opensolaris[22:37:15] <muthu> not sure if its available in the opensolaris.[22:37:24] <[lewellyn]> you're welcome to donate some ficon stuff to me so i can tell you! ;)[22:37:36] <[lewellyn]> muthu: check the hcl and device driver utility.[22:37:59] <muthu> lewellyn: thx. lemme check that[22:38:12] <muthu> but i believe its more of protocol than a device[22:38:24] <muthu> so may not in hcl?[22:38:26] <[lewellyn]> you have no controllers? ;)[22:38:33] <muthu> nope[22:38:40] <[lewellyn]> but yeah, the storagetek T10000B supposedly works with opensolaris[22:39:17] <[lewellyn]> "Available today for Windows, Solaris, OpenSolaris and Linux operating system environments, the LUMINEX FICON and ESCON channel connectivity solutions provide high level software interfaces that vastly simplify the task of driver integration." says a press release, too[22:39:28] <muthu> oh cool[22:39:32] <muthu> link please?[22:39:32] <[lewellyn]> so even if it's not out-of-box, it's available[22:39:36] *** e1kg has joined #opensolaris[22:39:39] <[lewellyn]> http://findarticles.com/p/articles/mi_m0EIN/is_20091117/ai_n42123077/[22:39:44] <muthu> thx a lot.[22:39:50] <[lewellyn]> i assume that if you can do ficon, you can buy software ;)[22:40:11] *** windowsrefund has joined #opensolaris[22:40:13] <windowsrefund> hello[22:40:24] <muthu> not for us[22:40:29] * [lewellyn] goes back to fighting with BrightQ + CUPS + OpenSolaris[22:40:29] <muthu> for a customer[22:40:50] <muthu> custy gets all the good stuff :([22:40:53] <windowsrefund> if I touch /etc/dhcp.vnic1 and run svcadm restart network/physical:default, shouldn't that interface attempt to grab his settings via dhcp?[22:40:54] <[lewellyn]> muthu: you may want to search the various mailing lists and stuff too[22:41:00] <muthu> ok[22:41:16] * [lewellyn] doesn't do dhcp with network:physical, so *shrug*[22:41:25] <[lewellyn]> that's what nwam is for[22:41:44] <windowsrefund> oh[22:41:59] <windowsrefund> so nwam looks for the /etc/dhcp* files?[22:42:14] <Triskelios> nwam ignores vnics right now, afaik[22:42:17] *** fastsolars has quit IRC[22:42:55] <Triskelios> nwam also just uses its own configuration rather than /etc/{hostname,dhcp}.*[22:42:58] <[lewellyn]> oh vnics. i missed that[22:43:13] <[lewellyn]> nwam is so you don't have to configure things as much :)[22:43:48] <windowsrefund> I don't mind using nwam... I just want to know my vnic will dhcp when I reboot[22:43:57] *** MaxVandervelde has quit IRC[22:44:00] *** muthu has quit IRC[22:44:30] <Triskelios> DHCP on a vnic should work, provided there is a server on the vlan[22:44:35] <windowsrefund> right[22:44:37] <[lewellyn]> not with nwam, it won't. i don't really play with vnics[22:44:45] <[lewellyn]> listen to Triskelios :)[22:44:55] <windowsrefund> I'll try it[22:45:12] <windowsrefund> so just touch /etc/dhcp.vnic1 and ensure network:nwam:default is running?[22:45:37] <Triskelios> stick with physical:default for vnics[22:46:12] *** cato_ has joined #opensolaris[22:46:14] <windowsrefund> I'm getting confused[22:46:27] <windowsrefund> oh and it's network:nwam, sorry for the error[22:46:30] <[lewellyn]> windowsrefund: you have vnics. nwam is irrelevant at this time.[22:46:56] <windowsrefund> ok[22:47:39] <windowsrefund> I touched /etc/dhcp.vnic1 and network/physical:default is running but vnic1 is just sitting there with 0.0.0.0[22:47:48] <windowsrefund> no[22:47:51] <windowsrefund> nwam works[22:47:59] *** MaxVandervelde has joined #opensolaris[22:48:05] <[lewellyn]> nwam will configure only a single nic.[22:48:40] <windowsrefund> that's all I want[22:48:58] *** leito has joined #opensolaris[22:49:18] <[lewellyn]> one nic, period, in the machine? why vnics then?[22:49:40] <[lewellyn]> and why more than one vnic?[22:49:49] <Triskelios> and if you only have a single vnic, how is the dhcp server going to work?[22:50:06] <[lewellyn]> Triskelios: well, he has more than one, if he has vnic1[22:50:26] <windowsrefund> this is just way too complicated[22:50:37] <windowsrefund> over engineered even[22:51:55] <[lewellyn]> windowsrefund: i think it's more that you don't quite have what you want to do planned out well enough to get assistance, yet[22:51:59] <[lewellyn]> sstallion_work: ping?[22:52:24] <sstallion_work> [lewellyn]: pong[22:52:25] <sstallion_work> whats up ?[22:52:36] *** fastsolars has joined #opensolaris[22:52:36] <[lewellyn]> how much interest do you have in virtualized hardware? ;)[22:53:06] <fastsolars> i use virtualbox[22:53:09] <sstallion_work> depends on how painful it is :D[22:53:10] <sstallion_work> whats up ?[22:53:30] <windowsrefund> [lewellyn]: are you kidding me? I asked how to ensure a vnic uses dhcp and how to ensure that config survives a reboot[22:53:40] <windowsrefund> what's unplanned about such a simple thing?[22:54:39] <[lewellyn]> sstallion_work: i'm curious how useful the open-source hyper-v stuff (gpl) for linux would be for getting better opensolaris support.[22:54:55] <[lewellyn]> windowsrefund: you said that nwam worked fine when we both said it can't[22:55:08] <[lewellyn]> which indicates that your configuration is incorrect. or your expectations.[22:55:23] <windowsrefund> but still no simple answer to such a simple question[22:55:43] <[lewellyn]> as i said, i don't do vnics. so i can't help much.[22:55:48] <windowsrefund> which is fine[22:55:53] <[lewellyn]> in a single-nic situation, nwam is the answer.[22:55:57] *** richb has quit IRC[22:55:58] <[lewellyn]> in your case, it can't be[22:56:03] <sstallion_work> [lewellyn]: I suppose it depends on the hardware really - do you know offhand which hardware is emulated?[22:56:04] <windowsrefund> but don't start telling me that I may not have my stuff "planned"[22:56:16] <sstallion_work> or is this for para-virtualized driver support?[22:56:17] <[lewellyn]> sstallion_work: moment[22:56:19] *** solar-star has joined #opensolaris[22:56:33] <[lewellyn]> sstallion_work: http://nopaste.info/87525f9b31.html[22:56:47] *** disharmony has joined #opensolaris[22:56:56] <[lewellyn]> the synthetic hardware is what interests me[22:57:10] <[lewellyn]> integrated shutdown could be nice too :)[22:57:12] <Triskelios> windowsrefund: you've been unclear about what the rest of the vlan looks like[22:57:36] <windowsrefund> omg[22:57:37] <[lewellyn]> windowsrefund: by giving us just a small piece of the puzzle and making us guess about the rest, we can only assume that the rest is still nebulous[22:57:49] <[lewellyn]> all we know is that you have a vnic. and that you want to do dhcp.[22:57:51] <windowsrefund> how much simpler do I need to phrase the question?[22:57:59] *** raichoo has quit IRC[22:58:02] <windowsrefund> vnic, dhcp, survive reboot, how?[22:58:03] <windowsrefund> :)[22:58:39] <[lewellyn]> http://www.bing.com/search?q=dhcp+vnic+opensolaris&form=OSDSRC[22:58:47] <windowsrefund> bing?[22:58:48] <windowsrefund> :)[22:58:54] <[lewellyn]> all i'd be doing is reading those and telling you to try them[22:59:02] <[lewellyn]> so i'll short-circuit the feedback loop[22:59:22] *** popster has left #opensolaris[22:59:35] <disharmony> dunno if this question is oftenly asked, but is there any news on when we can expect 2010.03 to be released? Or is this unclear/unkown/uncertain at this moment?[22:59:36] *** cocacola has quit IRC[22:59:52] <CosmicDJ> disharmony: /topic :)[23:00:00] *** held has joined #opensolaris[23:00:03] *** richb has joined #opensolaris[23:00:06] <sstallion_work> [lewellyn]: hrmm. looks like paravirt support. Its certainly doable, provided that there are specifications online (not just source code).[23:00:07] *** gjl_ has quit IRC[23:00:18] <sstallion_work> If its not GPL, there might be a prayer of it getting integrated even.[23:00:35] <disharmony> CosmicDJ: ohhh thanks, the topic didn't fit in my irssi topic bar ;-) Sorry ;-)[23:00:43] <[lewellyn]> sstallion_work: it's gpl. it's the only way they could get it into the linux distros' kernels :P[23:00:54] <alanc> and yes, it is often asked, which is why we put the answer in /topic[23:01:00] *** Reedd has joined #opensolaris[23:01:05] <[lewellyn]> afaik, there are no specs outside of the sources[23:01:21] *** milek_ has joined #opensolaris[23:01:31] <[lewellyn]> and it amuses me that they distribute the download as a .exe and say you can't redistribute it[23:01:53] *** nitrile_ has quit IRC[23:02:06] <Reedd> is it possible to grow a physical partition on a disk and stick two more zfs pools in that extra space?[23:02:25] *** twisti has quit IRC[23:02:34] <sstallion_work> [lewellyn]: are there specifications or just code?[23:02:37] <Triskelios> windowsrefund: it's dependent on your other vnic having a working DHCP server, and you may experience delays before it's able to acquire an address if the server is not forced to be brought up at boot before that particular vnic[23:02:44] <Reedd> when i installed opensolaris i installed it to a 30 gb partition and now i would like to grow it to use my whole disk[23:02:52] <[lewellyn]> sstallion_work: i haven't downloaded the exe yet. lemme do so.[23:03:02] <Triskelios> Reedd: it's usually not a good idea to create multiple pools on one disk[23:03:25] *** nitrile has joined #opensolaris[23:03:38] <[lewellyn]> hahaha. the exe is a sfx archive with a .iso and .pdf inside. *sigh*[23:04:13] *** Kedare has quit IRC[23:04:18] <Reedd> i have 4 tb drives in a machine, i'm trying to keep the system install on a mirror across the first piece of the 4 drives then make the rest a data pool[23:04:30] <[lewellyn]> Reedd: that'll impact your performance[23:04:48] <Reedd> i'm not that worried about it this isn't for a business[23:04:55] <[lewellyn]> grab a couple of smallish CF cards and use those as the rpool mirror ;)[23:05:01] *** gary_ has joined #opensolaris[23:05:08] *** gary_ is now known as Guest74527[23:05:22] *** wdp_ has quit IRC[23:05:31] *** randw has joined #opensolaris[23:05:32] <Reedd> mb doesn't have the ability to have that it's only got one ide and it's being occupied[23:06:45] <[lewellyn]> sstallion_work: ooh... lemme pastebin this![23:07:27] <Guest74527> hi, have any of you guys tried installing osol on a samsung netbook ?[23:07:37] <[lewellyn]> sstallion_work: http://nopaste.info/a634ce5e2e.html[23:08:02] <[lewellyn]> no one ever mentioned that the thing was dual-licensed![23:09:17] *** solar-star has left #opensolaris[23:09:23] <sstallion_work> welp[23:09:27] <sstallion_work> that could pretty easily be ported then[23:09:42] <sstallion_work> even has a snowball's chance in hell to get integrated :P[23:09:49] <[lewellyn]> would it be a project that interests you? ;)[23:10:05] * sstallion_work mutters something about how deep his stack is[23:10:08] <Triskelios> Guest74527: what are the specs?[23:10:10] <[lewellyn]> likewise :P[23:10:13] *** Mdx4 has joined #opensolaris[23:10:27] <[lewellyn]> and i suspect that this isn't a beginner's project[23:10:35] <sstallion_work> possibly. Unfortunately I don't really have a machine I can dedicate to Hyper-V[23:10:45] <sstallion_work> its probably not as bad as you think[23:10:50] <sstallion_work> DDI/DKI is really well documented[23:10:58] <Guest74527> standard atom 1.6 ghz based jobby i believe (2gb ram, 160gb etc)[23:11:13] <[lewellyn]> i probably could set something up for you at some point in the not-too-distant future, if you wanted a hyper-v machine[23:11:16] <sstallion_work> If you don't mind doing a bit of debugging I can scope it and see how tough it would be to port[23:11:27] *** radsy has quit IRC[23:11:42] <[lewellyn]> matt@matt-osol:/mnt/src/drivers/netvsc$ wc -l netvsc_drv.c 731 netvsc_drv.c[23:11:53] <[lewellyn]> it's smaller than i had expected[23:12:04] <sstallion_work> thats surprisingly small - which component is that?[23:12:09] <[lewellyn]> the network driver[23:12:23] <sstallion_work> probably just a simple SHM driver[23:12:23] <[lewellyn]> as much as dnet's improved, i'd love gigabit bursts in my guests :)[23:12:30] *** windowsrefund has quit IRC[23:12:35] <sstallion_work> make no mistake - dnet is still a piece of crap :D[23:12:51] <[lewellyn]> i didn't try to insinuate otherwise :)[23:12:52] <sstallion_work> I just added a little lipstick (which quite a bit of help from gdamore)[23:12:58] <sstallion_work> s/which/with/[23:13:06] <[lewellyn]> but you can quickly tell the difference when you image-update from 111b to 134[23:13:13] <Triskelios> Guest74527: as long as the video chip isn't the GMA500, should be fine[23:13:14] <sstallion_work> good![23:13:21] <sstallion_work> there were a few things which were done' to improve performance[23:13:25] <sstallion_work> you should get near line-speed[23:13:37] <[lewellyn]> it's still less than ideal, and i think it still causes guest hangs for brief moments, but i can't prove it :([23:13:59] <[lewellyn]> and it's not often enough that i've cared to try to prove it[23:14:13] <sstallion_work> well[23:14:15] <Guest74527> Triskelios, thanks, ill check the video chipset[23:14:22] <sstallion_work> a paravirtualized driver is going to be quite a bit faster[23:14:26] <[lewellyn]> well yeah[23:14:28] <[lewellyn]> that's the idea[23:14:29] <sstallion_work> its essentially a shared memory transport[23:14:53] <[lewellyn]> since i'm using hyper-v with osol guest as lamp replacements, with ALL the files (and in some cases, mysql databases) on nfs[23:15:10] *** ingenthr has quit IRC[23:15:21] <sstallion_work> frames are just copied (or better yet, fed directly to the sending bits) rather than going through an emulated DMA engine[23:15:23] <[lewellyn]> a bit of a non-standard configuration, but it seems to perform better than linux hardware[23:16:04] <sstallion_work> it wouldn't be too tough to bang out a nemo driver for it[23:16:23] <Guest74527> Triskelios, just checked , its the GMA 950 .... is that still a no go ?[23:16:37] <Triskelios> Guest74527: that's fine. I'm using one now[23:16:50] <sstallion_work> [lewellyn]: any chance it could be funded? ;)[23:16:50] <Guest74527> ah, is it samsung ?[23:17:02] <[lewellyn]> ah. looks like you need the vmbus driver for the nic and storage controller[23:17:11] <sstallion_work> *nod*[23:17:19] <Triskelios> Acer Aspire One. had a similarly spec'ed Eee before that[23:17:28] <sstallion_work> thats going to be a generic "bus" used to communicate with Hyper-V[23:17:32] <[lewellyn]> sstallion_work: i wish i could guarantee more than a couple hundred dollars :([23:17:34] <Guest74527> did you have a lot of trouble setting it up ?[23:17:39] <Triskelios> no[23:17:51] <sstallion_work> [lewellyn]: nah, if its out of pocket for you no worries - I was just giving you some guff[23:17:53] <[lewellyn]> the vmbus stuff is still less than 2000 lines, too[23:17:59] <Macer> cloverfield cracks me up[23:17:59] <sstallion_work> yeah[23:18:17] <sstallion_work> probably looks somewhat like your basic STREAMS driver[23:18:17] <Macer> i need to seriously work on fixing the newer BEs of opensolaris[23:18:23] <Macer> like the ptmux problem and what not[23:18:28] <Macer> i've just been too lazy[23:18:29] <Guest74527> I assume that osol has a usb install option ?[23:18:32] <[lewellyn]> sstallion_work: well, it'd not be exactly out of pocket. i do have a bit of a budget in my business ;)[23:18:33] *** popster has joined #opensolaris[23:18:37] *** fleyta has joined #opensolaris[23:18:50] <[lewellyn]> but times aren't great right now. money's tighter than i'd like.[23:19:04] *** doug_ndndn has quit IRC[23:19:05] *** forquare-macbook has quit IRC[23:19:07] <Triskelios> Guest74527: yup, the images are linked from the release notes and also mirrored on genunix.org[23:19:16] <popster> What would be the sol equivalent of /proc/uptime?[23:19:17] <sstallion_work> [lewellyn]: seriously, no worries. If this is doable and you have hardware I can use to test with, we can call this more than even[23:19:35] <tsoome> man uptime?[23:19:41] <alanc> /bin/uptime ?[23:19:54] <sstallion_work> [lewellyn]: I can imagine the PSARC will be the most painful part.[23:19:58] <[lewellyn]> sstallion_work: i'll have to buy another box i can dedicate to it. and i'm not in a huge rush. it'd just make my life easier and give me some more options in the future[23:19:59] *** Mech0z_ has joined #opensolaris[23:20:02] *** Mdx4 has quit IRC[23:20:08] *** Mech0z has quit IRC[23:20:18] <[lewellyn]> and yeah. at least it's available with a BSD license! :)[23:20:28] <popster> Not the executable uptime, its a symlinked string output of the current uptime[23:20:29] <[lewellyn]> would you like a copy of the iso?[23:20:39] <sstallion_work> [lewellyn]: sure.[23:20:47] <Guest74527> thanks Triskelios[23:20:50] <sstallion_work> it doesn't sound like a tough port to do[23:20:52] <[lewellyn]> lemme scp it somewhere useful[23:21:08] <sstallion_work> and having better Hyper-V support would probably be seen as a "good thing"[23:21:13] <[lewellyn]> i actually didn't know the linux drivers were open source till the other day[23:21:42] <Triskelios> popster: but what you want is the uptime information, right?[23:21:43] *** longcat has joined #opensolaris[23:22:03] <sstallion_work> [lewellyn]: you could probably turn this into a project belonging to the driver community[23:22:16] <sstallion_work> I'll do the code if you want to manage the political side of it :P[23:22:32] <sstallion_work> I'll even +1 it :P[23:22:35] <longcat> I see the topic says "Dont know when" but is there a why?[23:22:52] <[lewellyn]> sstallion_work: http://www.greenviolet.net/~lewellyn/LinuxICv21.iso[23:22:56] *** Animal-X has quit IRC[23:22:58] <sstallion_work> longcat: releases are hairy things - there are a number of bugs being worked on[23:23:04] <alanc> popster: the equivalent is running /bin/uptime - Solaris doesn't have non-process-specific data in /proc[23:23:07] *** fastsolars has quit IRC[23:23:08] <[lewellyn]> sstallion_work: i'm almost afraid of the political side :)[23:23:13] <longcat> i imagine package renaming would be part of it[23:23:22] <sstallion_work> [lewellyn]: these days its a message to a mailing list and some wiki hacking ;)[23:23:33] <alanc> longcat: there is certainly a why, several of them[23:23:36] <sstallion_work> once its done, then we can sort through the PSARC bits and look for a champion to get it integrated[23:23:47] <sstallion_work> we could probably pick on gdamore or jmcp for that[23:23:49] <longcat> I guess i'll look at the blocking bugs on the release in the tracker[23:24:37] <sstallion_work> longcat: you won't find anything outside of the indiana bits - the necessary information for ON is internal. just know that its being worked on... we promise :)[23:24:42] *** jdoe_ has joined #opensolaris[23:24:53] <sstallion_work> [lewellyn]: 404[23:24:56] <longcat> That's cool. At least it doesn't seem like it's an oracle type issue with the release[23:25:05] <gdamore> what did I miss?[23:25:09] <[lewellyn]> sstallion_work: it's probably permissions. moment[23:25:23] <[lewellyn]> sstallion_work: try again :)[23:25:41] *** kris__ has joined #opensolaris[23:25:50] <[lewellyn]> sstallion_work: but yeah. with how many places have windows servers, it'd probably not hurt adoption if there was support.[23:26:03] <[lewellyn]> gdamore: i'm trying to convince sstallion_work to improve hyper-v support in opensolaris ;)[23:26:26] <[lewellyn]> turns out that microsoft's open source driver is dual-licensed[23:26:52] *** kris__ has quit IRC[23:27:00] <gdamore> so running as a hyper-v guest?[23:27:04] <longcat> http://defect.opensolaris.org/bz/show_bug.cgi?id=8314[23:27:15] <gdamore> (although why anyone would choose microsoft's vm tech is ... mysterious to me.)[23:27:23] <gavino> lol[23:27:35] <gavino> virtualization is a bit wakky in general as uix is a multiprocessing os[23:27:48] <gavino> separation discipile causes fear I suppose[23:27:56] <[lewellyn]> gdamore: it already runs nicely in hyper-v. i have production websites to prove it. ;) my main motivation is to ditch dnet[23:27:58] <gavino> discipline[23:28:01] *** jdoe has quit IRC[23:28:18] <[lewellyn]> gdamore: and if you have a windows server already, why not use it? ;)[23:28:31] <gdamore> s/not use it//[23:28:41] *** fastsolars has joined #opensolaris[23:28:45] <[lewellyn]> you can't always choose the available tools that a client has, so you have to make the best of them[23:28:58] <[lewellyn]> and hyper-v is actually a pretty nice product, from an administration view[23:29:28] <gdamore> i understand that this may be a desirable at some sites... it seems (from an Open Source stand point at least) that we ought to try to work well under all VMs.[23:29:29] <[lewellyn]> i can't wait till i get them to deploy scvmm here[23:29:35] *** robo has quit IRC[23:29:39] <sstallion_work> [lewellyn]: got it[23:29:45] <gdamore> of course, I'm an Oracle employee, and Oracle has its own VM technologies....[23:30:00] <[lewellyn]> gdamore: all the more reason to use microsoft's open source assets and integrate them[23:30:06] <gdamore> i only use virtualbox because I have to... because some tools are only available for Windows. yech.[23:30:08] <TommyTheKid> OVM ~= Xen right?[23:30:08] <sstallion_work> gdamore: its more of an academic exercise really[23:30:09] <[lewellyn]> (the working well part)[23:30:22] <gdamore> TommyTheKid: more or less, yeah.[23:30:26] <[lewellyn]> TommyTheKid: afaik, it's an older xen + patches[23:30:42] <gdamore> but there is also Virtualbox which is another route and also an Oracle product[23:30:56] <[lewellyn]> virtualbox makes a poor hypervisor platform[23:31:17] <TommyTheKid> virtualbox seems to be more geared torwards desktop virtualization[23:31:35] <[lewellyn]> and virtualbox doesn't have an enterprise-wide central management solution[23:31:36] <TommyTheKid> like when you have required training courses that require the use if Internet Exploder and XP[23:31:42] <macousi> it's getting better... very slowly[23:31:44] * TommyTheKid grumbles[23:31:55] <[lewellyn]> hell, afaik, it doesn't even have integrated P2V tools[23:32:36] * causality hasnt seen better than esx yet[23:32:42] *** TomJ has quit IRC[23:32:44] *** melbogia has quit IRC[23:32:50] *** TomJ has joined #opensolaris[23:33:01] <longcat> Heh i'd have to agree that esx is pretty much the bee's knees if you have the money[23:33:03] <[lewellyn]> causality: osol runs better in hyper-v than in esx, ime[23:33:17] <macousi> esxi is free, no?[23:33:22] <causality> esxi is shit[23:33:33] *** doug_ndndn has joined #opensolaris[23:33:34] <causality> [lewellyn] i've only rnu opensolaris inside esx[23:33:44] <macousi> says you[23:33:45] *** Guest74527 is now known as [Gary][23:33:45] <causality> i was speaking feature-wise, as a wohle[23:33:59] <causality> macousi yes, says i..[23:34:07] <causality> vmotion etc[23:34:39] *** gm152 has joined #opensolaris[23:34:44] <TommyTheKid> its interesting, but since it requires windows as a management platform, I veto it :p[23:34:50] <[lewellyn]> causality: hyper-v + scvmm is a really good combination to compete with esx. and i haven't seen any xen-based offerings which offer a similar featureset[23:35:18] * causality throws the anti-windows zeolots out in the rain[23:35:34] <[lewellyn]> TommyTheKid: as a microsoft partner, i can't automatically veto windows-based solutions. (and due to politics at customer sites, it's in my best interest to know about all the current offerings from microsoft, and to be trained for them)[23:35:36] <gavino> zelots![23:35:53] <macousi> if you're desperate and you've got no windows machines kicking around, i believe the web interface can accomplish most tasks[23:35:55] <[lewellyn]> gavino: i've overlooked your ban evasion so far. do you have anything useful to add?[23:36:02] <TommyTheKid> I dont like it, and wont use it for my production infrastructue[23:36:06] <gavino> during install how can i see how big the drive is? I have 2x36g drive and 16x140g drive[23:36:12] <TommyTheKid> s/tue$/ture/[23:36:26] <gavino> and I think I want to exclude the 36g because they slow thigns down...put os on them[23:36:26] <causality> TommyTheKid: that attitude is fine for small time stuff[23:36:32] <gavino> is ti pretty obvious in the installer?[23:36:50] <[lewellyn]> gavino: why don't you try it?[23:36:55] *** fastsolars has quit IRC[23:36:57] <gavino> I am about to[23:37:00] <TommyTheKid> thats funny causality, I was thinking windows was fine for small time stuff :p[23:37:06] <[lewellyn]> TommyTheKid: in the real world, your environment is heterogenous[23:37:08] <gavino> got my cd but hopes for no pitfalls[23:37:12] *** RoyK has joined #opensolaris[23:37:22] <gavino> freebsd and centos suck sofar[23:37:31] <causality> [lewellyn] you mean there's a world outside of irc where everyone uses unix?![23:37:32] <gavino> as does this dell hardware, opensol to rescue[23:37:45] *** RoyK has quit IRC[23:37:49] <[lewellyn]> causality: i said the opposite[23:37:52] <TommyTheKid> yes, your environment can be heterogenous, and can support any client on any platform, as long as its windows XP or 2000 with Internet Explorer ;)[23:38:12] <causality> [lewellyn] you can read my sentence in two ways :)[23:38:13] <TommyTheKid> I would applaud ESX if they didn't have stupid ass requirements like that[23:38:23] <TommyTheKid> the underlying tech is awesome[23:38:26] <[lewellyn]> TommyTheKid: funny. the client i'm at today is a "windows shop", but only 4 physical machines are windows[23:38:32] <[lewellyn]> (in the server room)[23:38:38] <causality> TommyTheKid i dont think vmware care about the very small percentage that wouldnt use their product for that reason[23:38:48] <[lewellyn]> oh you meant esx, not hyper-v[23:39:01] <TommyTheKid> and thats why I choose OpenSolaris and whatever xVM is called these days :p[23:39:14] <causality> TommyTheKid your attitude has lead you to an inferior product :)[23:39:15] *** kvanals has quit IRC[23:39:19] <TommyTheKid> I really think ESX is probably the very best product[23:39:43] <[lewellyn]> wake me up when xVM has a central management interface that allows you to P2V boxes at will, and migrate them to instances all over the enterprise.[23:39:46] <causality> real men use the esx console![23:39:53] <TommyTheKid> thats my personal opinion, not the opinion of my employer of course ;)[23:40:13] * [lewellyn] wishes he could use that line[23:40:18] * causality uses the right tool for the job where possible[23:40:28] <[lewellyn]> likewise[23:40:47] <[lewellyn]> and if a client has an underutilized windows server with hyper-v, that makes it the right tool ;)[23:40:56] <crichardso> esx yuck[23:41:07] *** jdoe_ is now known as jdoe[23:41:10] <crichardso> i like xenserver better[23:41:19] <crichardso> esx does some funny stuff sometimes[23:41:32] <TommyTheKid> i hate to be this way, but this conversation has really gone off course, can we somehow draw this back to Open Solaris? :p[23:41:59] <crichardso> ok ;p xvm is better than esx ;p[23:41:59] <[lewellyn]> TommyTheKid: this whole thing started because i was trying to pester sstallion_work into porting the linux hyper-v stuff to opensolaris ;)[23:42:04] <causality> opensolaris can run virtualbox to launch the vmware infrastructure manager client[23:42:05] <causality> :)[23:42:07] <popster> I rebooted my OpenSolaris machine a couple hours ago...[23:42:14] <TommyTheKid> ahhh[23:42:32] <TommyTheKid> causality: heh, exactly, that was how I managed my ESX-i stuff :)[23:42:35] *** kvanals has joined #opensolaris[23:42:49] <[lewellyn]> then gdamore chimed in and started all this and wandered off. :)[23:42:59] <echobinary> anyone know of any good (FREE!) backup programs that do incrementals?[23:43:12] *** aszeszo has joined #opensolaris[23:43:14] <TommyTheKid> uh, zfs send -i ?[23:43:15] <echobinary> (other than snapshots, etc)[23:43:16] * [lewellyn] goes to lunch[23:43:20] <Spencer_tt1> [lewellyn]: do you get certified by MS to do MS work[23:43:21] <TommyTheKid> ok[23:43:44] <[lewellyn]> Spencer_tt1: being a partner is optional. *shrug*[23:44:02] <MaxVandervelde> Is there anyone here that could give me a hand with an openGrok Install? I can't for the life of me get the thing configured properly.[23:44:03] <[lewellyn]> again, if i have clients with windows needs, it's definitely in my best interests to know how to serve those[23:44:17] <causality> [lewellyn] dont you need to have a certain number of certified people to retain partner status?[23:44:20] <[lewellyn]> even if that means running osol under hyper-v :)[23:44:27] <[lewellyn]> causality: depends on which level of partner you are[23:44:33] <causality> i'm thinking gold partner[23:44:39] <Meths> echobinary: What size of environment? OSs being backed up?[23:44:39] *** kvanals_ has joined #opensolaris[23:44:43] <[lewellyn]> i'm a registered partner. i plan on being a certified partner next year[23:44:52] <Spencer_tt1> [lewellyn]: I just find the whole certification thing certifiable, no pun intended - some of us can't find time to justify such expenditure :)[23:44:54] <[lewellyn]> bigger than that and i need to do some hiring ;)[23:45:14] <causality> [lewellyn] iirc certified partner requires two mcps+[23:45:23] <causality> last time i checked was 2000[23:45:31] *** kvanals has quit IRC[23:45:33] <echobinary> Meths: initially one zpool to another, but ultimatly I want to back up laptops (running any variant of a MS OS) from other places in the network as well[23:45:38] <[lewellyn]> Spencer_tt1: registered partners have free online training and webcasts and stuff. no need for msce or anything[23:46:00] <[lewellyn]> causality: you can get certified with a sole proprietorship these days. it's just a pain in the ass.[23:46:04] <[lewellyn]> and really. lunchtime! ;)[23:46:11] <TommyTheKid> mmm f00d[23:46:16] *** kvanals_ has left #opensolaris[23:46:19] <Spencer_tt1> yet again time is the main variable :) enjoy your lunch :)[23:46:57] *** kvanals has joined #opensolaris[23:47:47] *** melbogia has joined #opensolaris[23:49:03] * sstallion_work is back.[23:49:34] <sstallion_work> gdamore: sorry, leweylln was trying to convince me to port the Hyper-V paravirt bits to ON[23:49:37] *** Animal-X has joined #opensolaris[23:50:26] <Spencer_tt1> MS advocacy[23:50:55] *** Animal-X has quit IRC[23:51:24] *** Animal-X has joined #opensolaris[23:52:17] <Meths> echobinary: Could look at Bacula for the whole thing but takes some setting up. Or something like Unison to sync the laptops and zfs send/snapshots/unison/rsync to cater to the server.[23:52:38] *** LaidBack_01 has joined #opensolaris[23:52:41] <echobinary> hmm[23:52:48] * echobinary checks out Unison[23:53:59] <LaidBack_01> hello, I've just installed nagios on an opensolaris zone. I already have a working email system here in this facility and would just like to have this relay mail through that IP. I'm looking for some type of smtp relay I'd suppose, can someone point me at the right software to install/look for?[23:54:53] *** gebi has quit IRC[23:55:15] <sstallion_work> blech[23:55:21] <sstallion_work> [lewellyn]: this is just a basic packet driver[23:56:14] <sstallion_work> [lewellyn]: I see nothing in here for setting multicast filteres much less changing the ethernet address[23:56:18] <sstallion_work> filters rather[23:57:06] <sstallion_work> it will require some love to make this a fully GLDv3 compliant driver[23:57:52] *** eleusis has quit IRC[23:57:58] *** Mdx4 has joined #opensolaris[23:58:26] *** eleusis has joined #opensolaris[23:58:26] *** eleusis has joined #opensolaris[23:59:10] <Meths> LaidBack_01: Think most setups just let nagios use the mail server on the local machine. If it needs to go somewhere else just set up the forwards/configure the MTAs to talk to each other.