October 19, 2007 
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 | 31
October 19, 2007 [00:00:32] <dsch04> Hmmm, Amd [00:00:39] <dsch04> Any Itel boards? [00:00:42] <dsch04> Intel? [00:01:08] <dsch04> Prferably with onbaord VGA? [00:04:37] *** SirFunk has joined #opensolaris [00:07:17] *** plocher has quit IRC [00:08:40] *** spurton has joined #opensolaris [00:09:00] *** dosiu has quit IRC [00:15:32] <RElling> I don't see any for Intel... http://www.nvidia.com/docs/IO/35384/LC_NFORCE_INTEL_web.pdf [00:16:55] *** yippi has joined #opensolaris [00:19:02] <elektronkind> heh, there's a Sun employee who has the name "Quaker Fang" [00:19:35] <elektronkind> If my last name were Fang, I'd want my first name to be Judge. [00:20:18] <jmcp> heh [00:25:16] <RElling> how about "Cardinal"? [00:25:31] <TrogL> too common up here [00:27:57] <e^ipi> one of the girls I work with has the last name of "power" [00:28:13] <e^ipi> I keep telling her she should name her first son "max" [00:28:22] <elektronkind> RElling: then I'd have to have a comfy chair in my office ;) [00:30:37] *** tomww has quit IRC [00:30:46] *** tomww has joined #opensolaris [00:31:09] *** Hydroxide has joined #opensolaris [00:31:28] <Hydroxide> hey. I accidentally got my raidz volume into a state where some files got corrupted [00:31:31] <Hydroxide> I'm trying to tell which ones [00:31:40] <Hydroxide> zpool status -xv is hanging or taking a long time [00:31:51] *** nicksieger_ has joined #opensolaris [00:31:52] *** _mw46_ has joined #opensolaris [00:31:53] <Hydroxide> (it is a huge volume) [00:32:12] <Hydroxide> what can I do? and should I bring the machine into single user mode or is it ok to allow it to be accessed to whatever extent it can be? [00:32:56] *** nicksieger_ is now known as nicksieger [00:33:01] *** nicksieger has left #opensolaris [00:34:02] *** yippi has quit IRC [00:35:37] <RElling> Hydroxide: iostat will show if it is just busy working on the disks [00:39:13] *** mega has joined #opensolaris [00:39:20] <JBeck> e^ipi: I had a classmate in high school whose last name was Licht [00:39:31] <JBeck> her brothers were both planning to have a son name Darrell [00:42:52] <stevel> when i worked for network operations @ UCSD we had a freshman call us for help whose first name was Nimrod [00:43:02] *** dnilsson has quit IRC [00:43:49] <Hydroxide> RElling: probably is doing that because it was resilvering [00:43:53] <Hydroxide> RElling: and scrubbing etc [00:44:11] <Hydroxide> RElling: but I'm just trying to get it in a state where I can run zpool status and similar commands and get useful info [00:44:15] <Hydroxide> RElling: without increasing the corruption [00:44:46] <Hydroxide> RElling: what happened is that someone else accidentally removed one drive (hot-swap) then put it back in, and I cleared the error state, and while it was resilvering I did the same mistake to another drive [00:45:03] <Hydroxide> RElling: the data that was on the drive should still be there [00:45:34] <Hydroxide> RElling: my main concern is to once again be able to issue these commands and also to not lose any more data than has happened [00:45:35] <RElling> Hydroxide: argv! OK, iostat should show that the disks are read I/O bound: iops should be reads, and %busy should be 100% or so [00:46:59] *** nrubsig has joined #opensolaris [00:46:59] *** ChanServ sets mode: +o nrubsig [00:47:37] *** tenex has quit IRC [00:47:45] <nrubsig> Today's news "mantools" rewrite in progress. [00:47:52] <nrubsig> s/news/news:/ [00:48:25] *** Pietro_S has quit IRC [00:48:50] <Hydroxide> RElling: well, now I can't even log in - all my shells are hung on I/O and it won't let me ssh in because I tried to go into single user mode (which is also not actually getting there) [00:49:07] <Hydroxide> RElling: I think rebooting forcefully will at least be necessary for me to get into a shell, but will that hose the data more than it's already hosed? [00:50:23] <RElling> Hydroxide: that might happen if the system is having difficulty accessing a disk, too. Are the disk blinkenlights blinking? [00:52:19] <Hydroxide> RElling: the two disks that were mucked with have their lights solidly on [00:52:26] <Hydroxide> RElling: and have for a while now [00:52:54] <RElling> I'd hang loose for a little while [00:53:51] <Hydroxide> RElling: how long is a "little while"? just so I know when to get more worried [00:54:35] <RElling> on the assumption that the disk lights mean that I/O is occurring, I wouldn't want to disrupt that progress. [00:55:06] <RElling> OTOH, if the disks were actually broken, then it may be a futile wait... are these consumer disks? [00:55:15] <Hydroxide> RElling: the disks were not broken. they were just removed accidentally [00:55:38] <Hydroxide> but how long of a wait would ou say is reasonable? it's probably already been at least 10-15 minutes [00:55:57] <Hydroxide> and I could be misinterpreting the light - the user manual is really really bad and tiny for this hard drive enclosure [00:56:04] <RElling> are any of the other disks busy? [00:56:05] <Hydroxide> no [00:56:09] <Hydroxide> just the two that were mucked with [00:56:38] <Hydroxide> the enclosure manual says that orange light blinking = hard disk access, but it doesn't actually mention the case of orange light solid [00:57:12] <RElling> for a reconstruction, I would expect reads from the surviving disks and writes to the two bad boys [00:57:24] <Hydroxide> yeah. I really don't know what's going on. [00:57:43] <Hydroxide> and I have no way to type any commands any more without rebooting [00:57:48] *** e^ipi has quit IRC [00:58:22] <Hydroxide> I'm not worried about new data right now - I'm just worried about the hard-powercycle-reboot itself causing further corruption. on the other hand, I may have no alternative [00:59:01] <RElling> I wouldn't worry about losing new data, but the unknown question is whether a reboot will fix the problem. You may have little other choice [00:59:43] <Hydroxide> well, yeah, I think I don't [01:00:04] <Hydroxide> are you reasonably sure about no additional corruption from the reboot? it seems pretty hosed [01:00:30] <Hydroxide> (I understand I'm not paying you for a legally binding guarantee) [01:00:40] <RElling> the hosed pool may not be importable, but the other pools should be ok. [01:01:02] <Hydroxide> that's the only pool unfortunately :( [01:01:15] <Hydroxide> but I don't think the damage was in the pool-wide metadata [01:01:20] <Hydroxide> will I be able to run zpool -xv on it? [01:01:25] <Hydroxide> err, zpools tatus -xv [01:01:27] *** rcorreia has quit IRC [01:01:30] <Hydroxide> *status [01:01:44] <RElling> you can only do that if it imports, but I don't think it will [01:01:57] <Hydroxide> why not? [01:02:22] <Hydroxide> well you know what, at this point I don't have any choice [01:02:31] <Hydroxide> I clearly can't do anything without rebooting [01:02:33] <Hydroxide> here goes [01:02:34] <RElling> if enough devices are hosed, then it won't import. [01:02:53] *** e^ipi has joined #opensolaris [01:02:56] <RElling> btw, which version of the OS? [01:03:26] <Hydroxide> this is actually solaris 10 11/06, not opensolaris ... I hope you don't feel I've abused the channel [01:03:44] <Hydroxide> it just seemed like the differences wouldn't be relevant to this particular problem [01:04:03] <Hydroxide> and I don't know of a regular solaris IRC channel [01:04:11] <RElling> I guess we could move to #zfs, but we've added some better debugging for problematic pool imports [01:04:25] <Hydroxide> in opensolaris you mean? [01:04:41] <RElling> I just don't remember when they showed up... pretty sure it was post-11/06. [01:04:46] *** vmlemon has quit IRC [01:05:27] *** joevandyk__ has joined #opensolaris [01:06:02] <Hydroxide> gotcha [01:06:17] <RElling> is it back yet? [01:06:29] <Hydroxide> no, it had problems booting up and wants me to go to the failsafe mode and update the boot archive [01:06:37] <Hydroxide> the system disk was ufs not zfs, btw [01:06:50] <RElling> argv! I hate it when that happens. [01:07:07] <Hydroxide> when does it happen? [01:07:19] <Hydroxide> I'm not at all new to *NIX in general, but I am new to *Solaris [01:08:35] <Hydroxide> rebooting into non-failsafe - we'll see what happens. the system itself shouldn't depend on the zfs [01:08:49] <Hydroxide> (it's data being served, but not used by the system itself) [01:09:01] *** hsn_ has quit IRC [01:09:35] <RElling> it has to do with updating the kernel that grub starts... it can be annoying, but isn't supposed to happen often [01:10:18] <Hydroxide> I may have done an smpatch update since the last reboot [01:10:22] <Hydroxide> that probably would have done it [01:10:26] <RElling> yep [01:10:43] <RElling> the command in question is bootadm [01:10:54] <Hydroxide> ah. [01:11:09] <Hydroxide> hm, now that I think about it, I probably should go to failsafe mode [01:11:21] <Hydroxide> since normal mode seems to be having trouble booting up [01:11:28] <Hydroxide> make sense? [01:11:39] <Hydroxide> oh no, it is proceeding, but slowly [01:11:43] <Hydroxide> very slowly [01:12:02] <RElling> maybe, you should be able to check on pool from failsafe mode [01:12:21] <RElling> if it is slow, you might be back where you were, more or less, blinkenlights? [01:12:23] *** joevandyk has quit IRC [01:12:44] <Hydroxide> this time I'm going into the GUI so I can add and remove windows as needed :) [01:13:00] <Hydroxide> most lights are green actually [01:13:08] <Hydroxide> with occasional orange blinks in the two that were mucked with [01:13:13] <Hydroxide> no solid orange lights [01:13:17] <RElling> ok [01:13:24] *** nrubsig has quit IRC [01:13:54] <Hydroxide> um? wow. it seems to consider everything to be fine, but resilvering [01:14:00] <Hydroxide> no corruption after all? [01:14:12] <RElling> hopefully :-) [01:14:17] <Hydroxide> no drives did actually fail, they were just removed temporarily [01:14:32] <Hydroxide> well, zpool status -x says "errors: No known data errors" [01:14:37] <Hydroxide> does that mean we can assume all data is fine? [01:14:39] <sommerfeld> so it's making sure none of them missed any updates. [01:14:49] <Hydroxide> or should I wait until the scrub is done to conclude that? [01:14:59] <sommerfeld> i'd let the resilver run to completion, then start a scrub, just to be sure. [01:15:07] <Hydroxide> well, I'm going to run it to completin [01:15:13] *** ZeleniZub__ has quit IRC [01:15:31] <sommerfeld> (resilver only reads what it needs to in order to reconstruct the suspect disk. scrub reads everything). [01:15:53] <Hydroxide> of course, with 2.57TB (no I do not mean GB) used, that will take a while :) [01:16:08] <sommerfeld> how many disks? [01:16:17] <JBeck> sommerfeld: one of the ZFS engineers told me last week that a scrub should not be needed after a resilver... [01:16:24] <JBeck> ... as they supposedly do the exact same thing [01:16:36] <sommerfeld> jbeck: i am more paranoid than that. [01:16:52] <JBeck> :-) [01:17:03] <sommerfeld> resilver will read all the metadata but won't verify all the leaf blocks. [01:17:14] <Hydroxide> sommerfeld: I think it's 8 [01:17:21] <Hydroxide> sommerfeld: but 2 are spares [01:17:26] <Hydroxide> sommerfeld: (raidz1 + 2 spares) [01:17:36] <Hydroxide> err, raidz1 including one redundant + 1 hot spare [01:17:44] <Hydroxide> or something like that [01:17:50] <Hydroxide> 8 physical disks, definitely [01:20:33] <Hydroxide> if the resilver and scrub both show no errors, can I assume that no files were lost entirely as a result of this? [01:20:37] *** _mw46_ has quit IRC [01:20:38] <Hydroxide> nor corrupted? [01:22:19] *** e^ipi_ has joined #opensolaris [01:22:20] *** e^ipi has quit IRC [01:22:49] <dsch04> I intend to run my storage servers with a pair of smaller drives, mirrored, for the OS installation. [01:23:01] <dsch04> Can I use zfs for this? [01:23:09] <dsch04> Is it possible to install to zfs? [01:24:01] <dsch04> Or do I need to do an install onto UFS and then copy it to a zfs? [01:24:28] <dsch04> Can I use whole disk zfs, or do I need to use slices for (e.g.) grub ? [01:24:37] <dsch04> Are there any up-to-date guides for this? [01:25:31] *** mog has left #opensolaris [01:31:21] *** hile_ has quit IRC [01:32:52] *** m0le has joined #opensolaris [01:34:11] <dsch04> Aha, found this [01:34:13] <dsch04> http://www.opensolaris.org/os/community/zfs/boot/zfsboot-manual [01:35:45] *** sommerfeld has quit IRC [01:35:48] <dsch04> So, assuming I have 2x80GB disks to use for the O/S, I can do something like: [01:36:05] <dsch04> - install OS to disk 1 [01:36:26] <dsch04> - create zpool on disk 2 [01:36:31] *** tenex has joined #opensolaris [01:36:37] <dsch04> - copy OS from disk 1 to disk 2 [01:36:59] <dsch04> - boot from disk 2 (zfs) [01:37:04] <tenex> what is the "best practices" way to ensure that ./configure will identify and use the Sun Studio compiler rather than gcc when both are in my path? [01:37:13] <tenex> (as well as make) [01:37:24] <dsch04> - add disk 1 to the zpool as a mirror [01:37:59] <tenex> I can't seem to find anything but 'it just works' when ./configure scripts seem to always latch-on to gcc ;) [01:40:27] *** Kush- has quit IRC [01:41:00] *** joevandyk has joined #opensolaris [01:42:42] *** JBeck has quit IRC [01:47:55] *** joevandyk__ has quit IRC [01:48:04] *** e^ipi_ has quit IRC [01:49:44] *** webar7 has joined #opensolaris [01:50:25] *** l1s has quit IRC [01:50:57] *** m0le has quit IRC [01:50:59] *** estibi_ has quit IRC [01:51:03] <webar7> when did solaris introduce "zones" [01:51:16] <webar7> which release [01:53:01] <elektronkind> solaris 10, the first release of it [01:53:02] <wesolows> S10 [01:53:13] *** ichigo has quit IRC [01:53:22] *** uebayasi has joined #opensolaris [01:57:54] *** e^ipi has joined #opensolaris [02:00:04] *** _Megaf has joined #OpenSolaris [02:01:19] *** kspath_ has left #opensolaris [02:04:34] *** e^ipi_ has joined #opensolaris [02:05:10] *** e^ipi has quit IRC [02:07:38] <webar7> hmm [02:07:42] <webar7> ok [02:07:45] <webar7> tnx [02:12:18] *** simford has joined #opensolaris [02:12:32] *** simford has joined #opensolaris [02:13:06] *** __hsilva has joined #opensolaris [02:20:41] *** charlesnw has quit IRC [02:21:22] <dsch04> Quickie: where is wget installed? Or what's the Solaris equivalent? [02:24:03] <Jiko> dsch04: try /usr/sfw/bin/wget [02:24:10] <dsch04> thx [02:25:14] *** lacaAFK is now known as laca [02:27:06] *** _hsilva has quit IRC [02:27:15] *** __hsilva is now known as _hsilva [02:32:17] *** mega has quit IRC [02:34:39] *** joevandyk__ has joined #opensolaris [02:37:39] *** polk__ has quit IRC [02:38:44] *** polk__ has joined #opensolaris [02:38:54] *** plocher has joined #opensolaris [02:40:19] *** plocher_ has joined #opensolaris [02:41:48] *** joevandyk has quit IRC [02:47:42] *** friendly12345 has joined #opensolaris [02:48:43] *** Gman has quit IRC [02:53:47] *** phs has joined #opensolaris [02:55:30] *** Tpenta has quit IRC [02:56:12] *** joevandyk__ has quit IRC [02:56:25] *** _Megaf has quit IRC [02:58:35] *** plocher has quit IRC [03:03:21] *** yippi has joined #opensolaris [03:04:28] *** e^ipi has joined #opensolaris [03:04:51] *** RainDoctor has joined #opensolaris [03:05:00] *** e^ipi_ has quit IRC [03:05:19] *** e^ipi has left #opensolaris [03:08:25] *** comay has joined #opensolaris [03:09:48] *** echaz has joined #opensolaris [03:13:25] <movement> steleman: ping [03:13:39] <steleman> movement: pong! [03:14:20] <movement> I just replied to your comment on the u_int* bug. I wasn't sure what you meant by grepping the headers? [03:14:37] <steleman> you asked for a platform having these non-standard typedefs ? [03:14:55] <movement> ah. no, I asked for a platform that had those but NOT the standard ones. sorry I wasn't clear. [03:15:15] <steleman> oh Linux has the Standard ones in stdint.h [03:15:21] <movement> yeah, exactly [03:15:23] *** stevel has quit IRC [03:15:31] <steleman> the problem is that sys/types.h has the non-standard ones, and these are they ones "they" chose [03:15:38] <movement> I don't know of a reasonable explanation for why somebody would prefer u_int* once they're aware [03:16:01] <steleman> "it works on Linux. i won't change my code just because it won't compile on Solaris". [03:16:04] <movement> I've sent patches numerous times and it's always been ignorance (in the non-pejorative sense) and they've always been immediately applied [03:16:10] <movement> steleman: !!! [03:16:12] <echaz> movement: legacy code [03:16:17] <movement> steleman: there are maintainers like that? [03:16:20] <movement> echaz: hmm? [03:16:24] <steleman> movement: oh yeah. [03:16:37] <movement> steleman: then I'm 100% against it :( [03:16:46] <echaz> that is why some would prefer that header file. [03:17:01] <movement> echaz: I don't get you. [03:17:06] <steleman> movement: most people usually give up arguing, and either do -Du_int32_t=uint32_t -Du_int16_t=uint16_t and so on [03:17:11] <steleman> :-) [03:17:16] <elektronkind> aww [03:17:24] <steleman> or hack config.h and add them at the bottom and hope it works [03:17:31] <movement> steleman: seems much more reasonable. preferably with a direct quote from the idiot maintainer? [03:17:34] <echaz> if you want to compile your code from 1988, and you don't want to be bothered with getting all the header files updated, you just use that one [03:17:52] *** nrubsig has joined #opensolaris [03:17:52] *** ChanServ sets mode: +o nrubsig [03:17:53] <movement> echaz: OK. Is that a real issue? [03:18:05] <echaz> it was for me at work.... [03:18:12] <movement> I'm not sure solaris should ever be integrating unmaintained software, echaz [03:18:14] <echaz> I used to maintain code that was that old [03:18:21] <echaz> up untill last year [03:18:26] *** webar7 has quit IRC [03:18:49] <steleman> not to mention that typing u_int16_t is harder than typing uint16_t so i really don't get it [03:19:05] <steleman> but they seem to love this stuff [03:19:08] <steleman> it's all over the place [03:19:11] <movement> steleman: well apart from the morons you mention, it's just genuine mistake in my experience. [03:19:47] <steleman> movement: i would *in theory* agree with the excuse that, before the uint*_t were standardized they had their own "thing" [03:19:57] <steleman> but now there really is no excuse for it [03:20:20] <movement> it's just plain hard to remember offhand which is std I think. [03:20:29] <jamesd> wonders when xen or what ever sun wants to call it, will all my solaris box (u20 939 based chip, dual core 2.4 ghz opteron) to run windows :-) [03:20:32] <steleman> movement: the one which is easier to type :-) [03:21:07] <echaz> you could always #define it to whatever you want :) [03:21:14] <echaz> or typedef even. [03:21:23] <movement> jamesd: is that an M2 box? [03:21:23] <wesolows> that rapidly becomes a nightmare [03:21:24] <echaz> typedef u_int16_t a [03:21:30] <jamesd> movement, nope :-( [03:21:35] <movement> jamesd then never [03:21:53] <movement> you /need/ amd-v to run windows [03:22:06] <wesolows> movement: that's actually very good news for him and everyone else; there was a risk he was going to run *windows*! [03:22:12] <jamesd> movement, i figured as much... so it stays a linux+vmware server box. [03:22:14] <movement> heh :) [03:23:20] <steleman> echaz: the discussion here is about whethere or not Solaris should have an "inttypes_gnucompat.h" header in /usr/gnu/include which contains these non-standard typedefs [03:24:35] <steleman> and there are strong arguments against it, and one convenience argument for it [03:26:51] *** yippi has quit IRC [03:26:57] <nrubsig> erm... can anyone access src.opensolaris.org right now ? I'm getting a "Lock at /opt/apache-tomcat-5 dot 5.20/temp/lucene-d6daff9915baac0851b88101f8511da5-commit.lock" each time... ;-( [03:27:41] <jamesd> nrubsig, its a conspiracy by the java coders to make there apps crash when they see your ip# [03:28:13] <nrubsig> jamesd: groan [03:28:26] <jamesd> nrubsig, it gives me the same error [03:28:29] <nrubsig> jamesd: please... don't tease me today [03:29:29] <g4lt-mordant> no if we were going to tease you, we'd say that we set up ksh2007 to be in java ;P [03:30:47] * nrubsig makes a note to add a FAQ [03:31:10] <nrubsig> g4lt-mordant: and for the log: The next thing after ksh93 is "kash", not "ksh2007" [03:31:42] <g4lt-mordant> "gimme the KASH, maaaaaan" [03:32:21] <g4lt-mordant> hey, what can I say, it's the prelude to the milla joviovich nude scene [03:32:32] <nrubsig> g4lt-mordant: techicially "kash" is a beefed-up ksh93, nothing more... :-) [03:34:28] <Bartman007> nrubsig: I thought ksh2001 was next... [03:34:39] <Bartman007> Hello, Dave. [03:34:46] <jamesd> g4lt-mordant, and the part that would kick him while he's down is that sun marketing would put a $1million dollars into convincing the world that the java based ksh is faster and less bloated than ksh93. [03:39:18] *** e^ipi has joined #opensolaris [03:39:37] <nrubsig> comay: ping! [03:39:43] *** snuff-away has quit IRC [03:39:50] * nrubsig is frustrated about comay today... [03:39:59] <jbk> evening [03:41:17] *** nachox has joined #opensolaris [03:41:33] <nachox> hi people [03:42:36] <jamesd> hi nachox [03:43:59] <g4lt-mordant> nrubsig, when AREN'T you? [03:44:47] <wesolows> comay is the most helpful, mild-mannered guy you'll ever meet; if he's frustrating you, you might consider becoming a monk and not interacting with humans any longer [03:45:24] <e^ipi> lol [03:45:31] <nachox> hahaha [03:46:25] <nrubsig> wesolows: erm... if I had to rate persons I'd say there is at least one person who is even more mild-mannered... [03:47:10] <jamesd> yeah but they hung him on a cross over 2000 years ago ;-) [03:47:23] <nrubsig> jamesd: no. [03:47:23] <jbk> haha [03:47:32] <nrubsig> jamesd: it's a "she" technicially [03:47:41] <wesolows> I'm mild-mannered, too...when unconscious [03:47:46] <jbk> ok brain fart... what's the file that lists the stuff to put in the boot ramdisk? [03:48:10] <jamesd> wesolows, why dont you just say every saturday night at bar close [03:48:13] <wesolows> /boot/solaris/filelist.ramdisk [03:48:33] <wesolows> jamesd: because I don't restrict my drinking to one night a week [03:48:36] <e^ipi> siddharta gautama? [03:48:36] <e^ipi> boot_archive? [03:48:44] <jbk> wesolows: thanks [03:49:05] <jamesd> wesolows, you said unconsious.... you dont get plastered on every day ending with y do you? [03:49:51] <wesolows> jamesd: I take my fifth amendment right [03:50:43] <jamesd> wesolows, i bet you mean... i'll take the fifth and slam it [03:51:12] <wesolows> well, yes :-) [03:51:42] <g4lt-mordant> I'll take the fifth of vod^W^Wamendment [03:52:38] <jamesd> i'll take the fith of 35 year old single malt scotch thanks. [03:53:49] <wesolows> if you have a bottle that old, I'll be glad to help you drink it [03:54:10] <wesolows> older than 21 years tends to be outside my budget [03:54:30] <g4lt-mordant> paedo!!! [03:54:51] <boyd> Anyone running Solaris paravirtualised DomU under xen? [03:55:00] *** e^ipi_ has joined #opensolaris [03:56:02] *** phs has quit IRC [03:56:24] *** phs2 has joined #opensolaris [03:56:52] <jbk> boyd: not yet [03:56:53] *** e^ipi has quit IRC [03:59:02] <boyd> jbk: I don't suppose you have any ideas on how to pass boot options to the DomU? (like -s for example) [03:59:36] *** e^ipi_ is now known as e^ipi [03:59:41] <jamesd> bets movement does unless he has hard coded them in dev code ;-) [04:01:03] <jbk> no clue [04:02:00] <jbk> heh if figures that like 2 weeks before snap upgrade is supposed to be released, i figure out how to do everything manually :) [04:02:14] <jbk> err it [04:03:10] *** phs2 has quit IRC [04:04:50] *** phs2 has joined #opensolaris [04:05:09] <friendly12345> Hi, I've got another xen question: [04:05:10] *** delewis has quit IRC [04:05:33] <friendly12345> Is it possible to run Solaris Express Developer Edition 9/07 as Dom0? [04:06:32] <boyd> I don't think the xen bits are in that, so no [04:07:13] <friendly12345> I think I read somewhere that the next OpenSolaris release or the one after might have Dom0 capability all fixed up [04:07:35] <friendly12345> add some time for it to go into Solaris Express, and ... [04:08:19] <friendly12345> they'll probably be an indiana release with it by then [04:15:09] *** plocher_ has quit IRC [04:16:19] <myrkraverk> someone said snv 75 would have xen [04:16:25] <myrkraverk> I don't know for sure though [04:16:40] <jbk> i think it's on the putback list [04:16:43] <friendly12345> Yeah, read something along those lines [04:16:52] <myrkraverk> jbk, putback list? [04:17:53] <jbk> that's probably not quite the right term for it.. but http://www.opensolaris.org/os/community/on/flag-days/all/ [04:19:03] *** stevel has joined #opensolaris [04:19:04] *** ChanServ sets mode: +o stevel [04:23:15] *** master_of_master has joined #opensolaris [04:23:54] <g4lt-mordant> myrkraverk, yeah, they have to respin 75, it'll be 75a [04:24:59] *** Kush- has joined #opensolaris [04:26:01] <jbk> so it should be well laundered by the time we get it :) [04:27:48] *** spurton has quit IRC [04:31:33] *** Tpenta has joined #opensolaris [04:32:10] *** ChanServ sets mode: +o Tpenta [04:34:04] <nrubsig> Tpenta: Welcome almighty, slayer of dragons, monsters and managers, lord of the southern shores of the great australian island, high priest of the coding circle, inventor of dtrace bourne probes, great god of the beloved... and the slave who's looking at the new ARC case stuff in a few mins... [04:34:40] * Tpenta sighs deeply at having to drop back to snv_65 on his u60 as gnome looks stuffed in 75 for this box [04:34:50] <bda> nrubsig: You're so weird you should be in the movies. [04:35:02] <nrubsig> bda: *hicks* [04:35:35] *** master_o1_master has quit IRC [04:37:22] <nachox> 32% to go and i'll have my very own sxde :) [04:38:02] <nachox> bda, and you havent seen his photo ;) [04:38:05] * nachox hides [04:38:07] <bda> ha. [04:38:12] * bda was quoting the Guide. :) [04:39:53] <nachox> Tpenta, how much ram does the u60 have? [04:40:04] <Tpenta> 2gb [04:40:16] <Tpenta> bug-buddy is SEGVing [04:40:41] <nachox> when did jds started including bug-budy? [04:40:49] <jbk> heh.. [04:40:54] <jbk> probably some sort of irony in that [04:41:11] *** echaz has quit IRC [04:41:37] <nachox> last time i asked i was told someone at sun was against it, the reason was to protect the customer's privacy i think [04:42:18] <Tpenta> protects my privacy well, ... session doesnt start, and no gnome-panel [04:42:19] <brendang> oh, did nrubsig convince Tpenta to do the ARC stuff? :) [04:42:19] <wesolows> bug-buddy...who comes up with these names? [04:42:37] <Tpenta> not yet [04:42:40] <stevel> it's not that different from Bugster or Bugtraq :) [04:42:59] <Tpenta> oh brendan, i did dtrace+javascript+postgresql+php at two developer days this week [04:43:10] <wesolows> true, those are fairly lame [04:43:13] <elektronkind> hey brendang, might you have a second to entrain a php/dtrace PECL module query? [04:43:17] <brendang> Tpenta: sweet. did you use any of the /Code examples from the toolkit? [04:43:20] <elektronkind> er entertain [04:43:26] <wesolows> it's a steel-cage death match between bugster and bug-buddy for lamest name [04:43:27] <Tpenta> i did use the toolkit :) [04:43:41] <elektronkind> the .99 DTT rocks with much glory [04:43:46] <Tpenta> it does [04:43:51] <brendang> elektronkind: maybe; there is the /Php subdirectory now in DTT [04:44:06] <Tpenta> "How many debugging tools have you used recently that are actually fun?" [04:44:14] <nachox> we use mantis at work to track those things, i hate it [04:44:30] * Tpenta has to grab some lunch as I have to be somewhere in 20 minutes [04:44:35] <wesolows> what do you hate about it? [04:44:40] <elektronkind> brendang: yeah, that's what I'm trying to use, but it looks like that no matter what rev of solaris I run the PECL extension on, the probes don't fire although they're listed for each httpd PID in 'dtrace -l' output [04:44:41] <Tpenta> damn not enuff time to fire up the bbq [04:44:50] <wesolows> Tpenta: use LOX [04:44:51] <brendang> well, I am a bum for not getting .99 out sooner - Michelle from docs gave me a deadline to get it done for the next OpenSolaris starter kit CD. [04:45:00] <Tpenta> oh and also nice, the dtrace probes are in postgresql in s10u4 [04:45:03] <elektronkind> brendang: just wondering it you've heard of that or have any idea, general or not [04:45:04] <jbk> are there any restrictions on who can propose projects? [04:45:15] <wesolows> uhh...since when do the docs group give engineers deadlines on private projects? [04:45:29] <brendang> elektronkind: I did notice some oddness with the php provider - some probe types had corrupt args. Try the /Code/Php examples from the toolkit [04:45:41] <Tpenta> keith, i gues sthey always could, it's just previously they were quietly ignored ;) [04:45:44] <elektronkind> brendang: cool, will do [04:45:58] <brendang> wesolows: one opensolaris person nagging another [04:46:11] * wesolows nods [04:46:18] *** tenex has quit IRC [04:46:44] <jbk> (just wondering if i'm gonna get flamed if i propose something) [04:46:57] <wesolows> jbk: anyone can propose a project [04:47:02] <jbk> ok [04:47:04] <brendang> hopefully in DTT 1.0, I'll have subdirectories for C and C++. that'll provide 10 language directories in total... [04:47:05] <wesolows> jbk: the Group in question may vote it down... [04:47:08] <e^ipi> what're you gonna propose? [04:47:19] <wesolows> jbk: and they may flame you, but not because you aren't entitled to propose a project. [04:47:27] <nachox> about mantis? it feels ackward working with it, it's like the author placed things where i dont expect them, it's probably just me though [04:48:01] <jbk> :) [04:50:09] <wesolows> hopefully by the time DTT 1.0 is out C++ will be deader than latin and we can all forget we've ever heard of it [04:50:47] <brendang> there is too much C++ in the world for that [04:51:03] <e^ipi> oh can it... C++ isn't that bad [04:51:53] <e^ipi> it's not a catch-all language, but anyone who thinks there is a language that falls in that category is lying [04:51:54] <jamesd> e^ipi, your right, compared to java or cobol. [04:52:19] <e^ipi> jamesd: or the GTK "let's make C object-oriented" approach ... [04:52:21] <wesolows> it's all assembly in the end [04:52:25] <stevel> ... please .... we already have OS wars here. can we not have language wars too... ? [04:52:28] <wesolows> so I nominate that the catch-all language [04:52:33] <wesolows> stevel: they're fun [04:52:38] <stevel> i'll +1 that [04:52:41] <wesolows> stevel: actually, ragging on c++ is fun [04:52:58] <wesolows> and it's such a shitty little bitch of a language; it really deserves it [04:53:13] <nrubsig> Does anyone here speak chinese ? [04:53:32] <jamesd> nrubsig, no, not even the 1 billion chinese people in china [04:53:45] <e^ipi> heh, pedant [04:53:53] *** stevel changes topic to "assembly: pwn'ing you for 20+ years" [04:54:03] <nrubsig> jamesd: the dragon shell eat you (cooked, with mushrooms) [04:54:08] <e^ipi> which assembly? [04:54:15] <nachox> someone should just kill stroustrup [04:54:20] <stevel> e^ipi: all variations on a theme :) [04:54:20] <nrubsig> stevel: erm, I hope you still remeber the original topic [04:54:25] * jbk dons flame retardant suit [04:54:27] <e^ipi> x86 hack assembly, or MIPS assembly [04:54:32] <e^ipi> or weird itanium assembly [04:54:33] <stevel> nrubsig: it was getting old and boring anyway ;-) [04:54:35] <jamesd> nrubsig, and i bet with lee kum key oyster sauce and cineese hot mustard. [04:54:48] <nrubsig> stevel: umpf [04:55:12] * nrubsig mourns the loss of opensolaris.pastebin.org - thanks to ---> YOU <--- stevel !! [04:55:35] <nrubsig> stevel: you're guilty of causing the worlds pain, wars and whatever else, too! [04:55:36] <nrubsig> burn him! [04:55:47] <stevel> and you're guilty of triggering my highlight too many times in a row [04:55:55] <nrubsig> heh [04:56:00] <nrubsig> stevel: *BEEP* [04:56:04] * nrubsig runs [04:56:09] *** nrubsig was kicked by stevel (*BEEP* THIS) [04:56:18] <stevel> huh [04:56:19] <stevel> no auto-rejoin [04:56:20] <stevel> i guess [04:56:29] *** nrubsig has joined #opensolaris [04:56:29] *** ChanServ sets mode: +o nrubsig [04:56:30] <wesolows> stevel: I'm appreciating your newly-assertive irc personality [04:56:31] <nachox> isnt that abusing of your op status? :) [04:56:40] <stevel> wesolows: :-P [04:56:50] <nrubsig> RAAAABAAAAHHHH... stevel kicked me... *whine* [04:56:51] <stevel> nachox: no, abuse would be if it were unwarranted [04:56:58] <wesolows> or banning him [04:57:09] <stevel> that too [04:57:26] <jamesd> or usermod -s /bin/sh nrubsig [04:57:54] <jbk> heh [04:58:05] *** nrubsig has left #opensolaris [04:58:06] <jbk> i've been known to set people's shells to emacs :) [04:58:17] <nachox> that is sick [04:59:56] <nachox> 1 hour to go :) [05:00:04] <jbk> until? [05:00:09] *** nrubsig has joined #opensolaris [05:00:09] *** ChanServ sets mode: +o nrubsig [05:00:14] <g4lt-mordant> no, sick is setting their shell to ed [05:00:16] <nrubsig> groan [05:00:19] * Tpenta had someone who worked for him who logged into vi [05:00:24] <nachox> i get SXDE [05:00:28] <jbk> ahh [05:00:39] <nrubsig> Tpenta: "vi" as login interface ? [05:00:44] <stevel> if your shell is set to /bin/vi, and you execute :sh - does it just spawn vi again? [05:00:49] <Tpenta> it was the shell he worked from [05:00:56] <nrubsig> heh [05:01:00] <Tpenta> no idea exactly how he did it, but he lived in vi [05:01:16] <elektronkind> one could always !sh out of it [05:01:28] <nrubsig> Tpenta: feel free to file the RFE to add /usr/bin/vi to getusershell(3c)'s whitelist [05:01:44] <Tpenta> no thank you [05:01:55] <nrubsig> Tpenta: that arc case... it will be... interesting... :-) [05:02:19] * wesolows files it closed approved automatic, puts back, and resigns [05:02:31] <brendang> nrubsig: what ARC case is it? [05:02:42] <brendang> nrubsig: is this the ksh93 demos? [05:02:49] *** stevel changes topic to "Latest SXCE 74 | Latest ON 75 | Starter kits: http://get.opensolaris.org | Clipboard: http://opensolaris.pastebin.ca/ | Twitter: http://twitter.com/opensolaris" [05:02:52] <nrubsig> brendang: add /bin/vi to getusershell(3c)'s whitelist per Tpenta's request [05:03:18] <brendang> what?? why would /bin/vi be a valid shell? [05:03:25] <nrubsig> brendang: ask Tpenta [05:03:49] <brendang> Tpenta: who uses /bin/vi as their default shell? [05:04:02] <wesolows> brendang: you do now [05:04:03] * stevel heads to dinner [05:04:06] *** stevel has quit IRC [05:04:14] <jamesd> brendang, the same type of people that enjoy /bin/sh as there shell [05:04:19] <Tpenta> i had someone work for me at csiro who lived in vi [05:04:37] <e^ipi> speaking of which actually... is it possible to have someone log in to a machine which automatically dumps them to a tip session ? [05:05:11] <e^ipi> that is, if I wanted to give someone a login on my server so that they could be at the console of a different machine? [05:05:15] <elektronkind> add it to the .login/.profile/.bashrc [05:05:20] <nrubsig> e^ipi: yes... /etc/profile or /etc/env.d/ are your friends... :-) [05:05:23] <e^ipi> so, yes then [05:05:40] <elektronkind> tip hardwire ? [05:05:41] <e^ipi> I could alternately buy a terminal server, but that requires purchasing something [05:06:13] <nrubsig> elektronkind: (after some thinking...) you want either .bashrc or .kshrc AFAIK since you only want to affect interactive shells [05:06:26] <elektronkind> tis true [05:08:08] <g4lt-mordant> right, because csh isn't interactive ;P [05:08:32] <nrubsig> elektronkind: and you need some kind of "lock" around "tip" to avoid a mess-up between two instances [05:08:52] <nrubsig> elektronkind: and you need a way to kick a hung "tip" [05:11:35] *** Teltariat has quit IRC [05:11:54] <nrubsig> Tpenta: piiiing! [05:12:57] *** tombhadAC has joined #opensolaris [05:25:19] <brendang> nrubsig: did you enjoy the summit? ... I couldn't hang around - busy with ZFS... [05:25:41] <e^ipi> dtracing zfs? [05:26:19] <brendang> zfs performance [05:26:30] <jamesd> brendang, zfs does just about eveything fast, except scrub while taking snapshots every 15 minutes [05:26:38] <e^ipi> so... don't do that? [05:26:40] <elektronkind> ugh scrub [05:26:49] <elektronkind> I wish there was a way to rate limit scrub [05:27:18] <tsp> my zfs scrub is hung, I tried to abort but it's stuck [05:27:27] <jamesd> i just modified my snapshot script, so that it exits if the disk is being scrubed. [05:27:28] <jbk> i believe there are a few areas where it doesn't perform well [05:28:01] <elektronkind> just being able to set a priority would be nice [05:28:22] <elektronkind> instead of having go "ooooh IO! nomnomnom" [05:28:23] *** hsn_ has joined #opensolaris [05:29:26] <jamesd> zpool status | grep "scrub in progress" && exit [05:29:28] *** phs2 has quit IRC [05:29:54] <e^ipi> gzip compression is also quite useless on zfs [05:31:04] <jamesd> i'm waiting for rzip, i hear its quite effective on log files... and zfs is allready using tons of memoy what a bit more? [05:31:31] <elektronkind> heh [05:31:43] <elektronkind> ARC is already the pirate of the high seas when it comes to memory [05:32:10] <jamesd> then explain why ZFS runs so crappy on a 512MB box... [05:32:35] <nrubsig> brendang: yes... but I had trouble withe the machine pool: Installer bugs (Sarah Jelinek has four of them on her plate), missing cables and simple not enough time in one chunk to get the stuff up&&running. [05:32:44] <elektronkind> <arc> yaaar! thar be unclaimed pages about! quick, mates, pull yonder an' pillage! [05:32:55] <nrubsig> brendang: good point is: I did all the stuff I wanted to do with the T2000 and U45 [05:33:37] *** spurton has joined #opensolaris [05:33:41] <nrubsig> brendang: I'm preparing an email to list what went ok and what didn't work. [05:34:10] * tsp wonders how much more performance he would get if he moved from zfs to ufs on his 256mb box [05:34:13] <brendang> nrubsig: so they can fix it for the next one in 6 months? [05:34:43] <nrubsig> brendang: no fixing needed, just adjustments in the organisation. Next time we try to be better prepared for stupid hiccups. [05:34:54] <nrubsig> brendang: and we really had _stupid_ hiccups. [05:34:55] <jamesd> tsp, lots but no snapshots, in-line compression, end to end checksums and much more would be missing [05:35:00] <tsp> when I'm running, say, maildrop, I think it's loading maildrop from disk every time it is called instead of being cached in ram, at least on this box [05:35:01] <brendang> elektronkind: yes, I'm working on the ARC. [05:35:22] <elektronkind> yay [05:35:30] *** akolb has joined #opensolaris [05:35:31] * elektronkind roots for brendang [05:35:33] <tsp> what is the modern command to show how much the ARC is using? I tried a mdb command, forgot which one now, but it didn't work [05:35:35] <nrubsig> brendang: one of them being the detail that he had to troubleshoot stuff and didn't had more than three hours until the building was locked-down with no way to stay around anymore. [05:36:02] <nrubsig> brendang: which technically delayed the setup of the server by 1 1/2 days [05:36:33] <nrubsig> brendang: next fun was: T2000 was there but no matching serial cable to access the SC [05:36:48] <nrubsig> which SUCKS [05:37:22] <nrubsig> and I didn't had any tools to craft my own cable [05:38:14] <nrubsig> brendang: and some smaller items [05:38:47] <e^ipi> tools? [05:38:47] <e^ipi> just grab some cat5, a pair of nail clippers and go to town [05:39:00] * nrubsig kicks e^ipi [05:39:02] <brendang> nrubsig: fair enough; I've done plenty of presentations where missing cables or poorly setup AV gear is a real pain. things need to be tested the day before. [05:39:39] <tsp> ah here it is - the failing command is echo "arc::print -d size" | mdb -k, with a failed to dereference symbol error [05:39:44] <nrubsig> brendang: that was the next problem... I spend too much time in MPK17 discussing things with April&Don and then spend >= two hours stuck on the streets [05:39:55] <nrubsig> six hours late [05:39:58] *** plocher has joined #opensolaris [05:40:06] <brendang> elektronkind: well, I did spend a number of weeks looking at improving the ARC hit rate and concluded that it was pretty good as it was - and additionaly algorithmic complexity was arguably not worth it. So, the ARC does work pretty well as it is. [05:40:10] <nrubsig> and technically we would've needed these six hours [05:40:44] <brendang> tsp: kstat -pn arcstats [05:40:57] * nrubsig attacks plocher and strangles him with 8+2 tentacles and drags the victim into his cave for digestion... [05:41:29] <jbk> it would be nice if ::memstat had some knowledge of the arc cache [05:41:37] <elektronkind> brendang: my only real beef with ARC is that it can be a little overbearing on consumption and not pull back fast enough in the face of pressure. I've had ARC bump httpd processes out to swap before and so I have to manually cap its max size on boot [05:42:00] <brendang> elektronkind: but there do seem to be cases where it can starve system memory, which I might be able to help with. [05:42:21] <elektronkind> yeah, that's generally what I've seen as well [05:42:41] <brendang> elektronkind: yep, I've seen the ARC hammer memory to the point where the system was unusable (ok, the system *was* DEBUG, which makes a big difference) [05:43:15] *** sfire||mouse has joined #opensolaris [05:43:36] <tsp> I'm guessing zfs:0:arcstats:size 40395544 is what I want [05:44:49] <brendang> tsp: yep [05:45:03] <brendang> tsp: those kstats are pretty useful, and were added recently (thanks Bryan!) [05:45:20] <tsp> hmm that seems pretty small, what could be using the other 200+ mb ram [05:45:46] *** friendly12345 has quit IRC [05:46:05] <jamesd> perhpas that is whats left after zfs takes its chunk... ;-p [05:46:28] *** ICU has quit IRC [05:46:47] *** ICU has joined #opensolaris [05:48:35] <brendang> tsp: try ::kmastat from mdb -k [05:49:58] *** hsn_ has quit IRC [05:51:37] *** tombhadAC has quit IRC [05:53:32] *** tombhadAC has joined #opensolaris [05:55:14] <tsp> hmm, interesting stats if I knew waht they meant :) [05:56:10] <jamesd> tsp, sounds like you need to pick up a book by brendan gregg its quite good [05:57:06] <e^ipi> I found the other one more entertaining [06:00:55] <tsp> jamesd: Ah, the solaris performance and tools book? Sounds interesting if there was an etext version of it [06:01:07] *** akolb has quit IRC [06:03:01] <jamesd> tsp, well it is interesting... and brendang is one of the authors of it :-) [06:10:06] *** sfire||mouse has quit IRC [06:14:45] *** nachox has quit IRC [06:18:24] *** akol1 has joined #opensolaris [06:20:39] *** chrisso_ has joined #opensolaris [06:20:53] <chrisso_> bl0rting commences at 5pm Mr Tempt. [06:28:14] *** Triskelios has quit IRC [06:30:41] <JWheeler> Hi folks, I'm trying to burn a 700MB ISO, I understand that I need to feed cdrw -C from the man page, but I'm quite sure what units it wants... Does anyone know? [06:31:59] *** kadath has quit IRC [06:34:40] <JWheeler> oh... just plain -C did it, I think [06:49:41] *** comay has quit IRC [07:00:29] *** plocher has left #opensolaris [07:13:59] *** akol1 has left #opensolaris [07:17:36] *** halton has joined #opensolaris [07:19:33] *** laca has quit IRC [07:29:01] *** jteo has quit IRC [07:29:36] *** e^ipi_ has joined #opensolaris [07:29:37] *** e^ipi has quit IRC [07:29:38] *** cmihai has joined #OpenSolaris [07:33:21] <WickedWicky> morning [07:34:51] *** tsoome1_ has quit IRC [07:39:36] *** friendly12345 has joined #opensolaris [07:40:01] *** e^ipi_ has quit IRC [07:40:04] *** e^ipi has joined #opensolaris [07:45:39] *** SirFunk has quit IRC [07:46:11] *** Vanuatoo_ has joined #opensolaris [07:52:40] *** phrost6 has quit IRC [07:57:35] <razrX> morning all [07:59:04] *** Vanuatoo_ has left #opensolaris [08:00:04] *** TwoPiece has joined #opensolaris [08:00:35] <TwoPiece> i just know it. Project Indiana will kick ass. [08:01:44] <g4lt-mordant> s/kick/suck/ [08:02:17] <TwoPiece> why do you say that? [08:03:44] <g4lt-mordant> because project indiana's major claim to fame has been to tell the rest of the cimmunity how they're doing it wrong [08:04:47] <TwoPiece> so you prefer belenix and nexenta, etc? [08:06:00] <TwoPiece> regardless, there's a lot of side projects in indiana that will benefit opensolaris anyways. [08:07:04] <TwoPiece> i don't see how it's insulting. [08:07:40] *** tsoome has joined #opensolaris [08:07:45] <TwoPiece> (as you seem to suggest) [08:09:02] <g4lt-mordant> TwoPiece, IT may not be insulting, but the first dozen press releases for PI were basically "yuou all suck, and here's how cool our stuff's gonna be" [08:09:43] <TwoPiece> so you're just deriving it from context, then? [08:10:48] <g4lt-mordant> what else is there? show me the code [08:11:48] *** tsoome1 has joined #opensolaris [08:11:51] <TwoPiece> so what you're saying is that you can't see what they're doing? [08:12:28] <g4lt-mordant> what I'm saying is EXACTLY WHAT I FUCKING SAID, no more, no less [08:13:03] <WickedWicky> dude! you said the F word! [08:13:35] <g4lt-mordant> WickedWicky, well, if someone wants to put words in my mouth I have to let SOME out [08:13:44] <WickedWicky> that's true [08:13:46] <WickedWicky> fuck eh! [08:14:13] <WickedWicky> yes, more coffee then to the office, thank god is friiiiiday [08:18:10] *** tsoome1_ has joined #opensolaris [08:19:21] <madhatter> So true [08:21:00] *** seanmcg has quit IRC [08:21:19] *** seanmcg has joined #opensolaris [08:21:26] <friendly12345> I have major gripes with the way solaris 'resolves' dependencies during install [08:21:56] <gdamore> hmmm... new one with opengrok: Error: Lock obtain timed out: Lock at /opt/apache-tomcat-5 dot 5.20/temp/lucene-d6daff9915baac0851b88101f8511da5-commit.lock [08:22:16] <gdamore> (trying to search on symbol ndi_devi_online.... [08:22:33] *** cmihai has quit IRC [08:23:25] *** tsoome1__ has joined #opensolaris [08:23:34] *** seanmcg has quit IRC [08:23:48] *** tsoome has quit IRC [08:23:55] *** seanmcg has joined #opensolaris [08:23:56] *** Gropi has joined #opensolaris [08:25:01] *** phimic has joined #opensolaris [08:26:44] *** RainDoctor has quit IRC [08:32:58] *** tsoome has joined #opensolaris [08:35:50] *** tsoome1 has quit IRC [08:38:55] *** tsoome1 has joined #opensolaris [08:40:17] *** tsoome1__ has quit IRC [08:40:45] *** tsoome1_ has quit IRC [08:43:19] *** TwoPiece has quit IRC [08:54:34] *** spurton has quit IRC [08:55:08] *** tsoome has quit IRC [08:56:48] *** Chihan has joined #OpenSolaris [08:57:33] *** jonkri has joined #opensolaris [08:59:39] *** Cyrille has joined #opensolaris [09:05:38] <flyingparchment> S10U4 doesn't need the encryption kit, right? [09:05:57] *** Netwolf has quit IRC [09:06:26] *** cmihai has joined #OpenSolaris [09:06:59] *** cydork has joined #opensolaris [09:07:08] *** mikefut has joined #opensolaris [09:17:52] *** timelyx has joined #opensolaris [09:18:24] <trochej> Coffeee [09:18:29] <trochej> I go hunting [09:22:09] *** cydork has quit IRC [09:25:58] *** echelog has joined #opensolaris [09:28:30] *** nrubsig has quit IRC [09:29:59] <trochej> re [09:32:50] *** timsf has joined #opensolaris [09:33:27] <timsf> morning all [09:38:25] <trochej> mornin [09:39:18] <quasi> hey hey [09:40:16] <trochej> Drat, my daughter overslept and I got late to work :/ [09:40:57] <cmihai> I fail to see the connection. [09:40:58] <timsf> Lucky you - my daughter underslept, and had us up at 5am! [09:41:26] <timsf> (add to that, I arrived back in Ireland at 4pm yesterday, and my brain is still on PST) [09:41:33] <Berny_> lucky me... my daughter slept til 8 this morning [09:41:34] <trochej> Khe\ [09:41:50] <trochej> cmihai: You have small choldren? [09:41:54] <trochej> s/cho/chi/ [09:42:02] <cmihai> No, not really. [09:42:08] <cmihai> Well, not that I know of... [09:42:11] <trochej> cmihai: Get yourself one, you'll see the connection [09:42:18] <trochej> :) [09:42:38] <Berny_> how old are your kids? [09:43:07] <trochej> 2 and -1,5 [09:43:08] <trochej> :) [09:43:23] <trochej> Which means, the second is in making :) [09:43:38] <cmihai> Hm... last I checked, it took up to 9 months. [09:43:53] <Berny_> planning... [09:44:00] <Berny_> not yet ordered so to say [09:44:12] <timsf> trochej, yeah - 9 month build time, no incremental option :-) [09:44:12] <cmihai> Oh, right. [09:44:17] <cmihai> heh [09:44:21] <Berny_> hehe my is a little over 2 as well [09:52:07] *** The-spiki has joined #opensolaris [09:57:49] *** WobbleU has joined #opensolaris [09:59:46] <trochej> It takes 1,5 year, 'cause we [09:59:55] <trochej> It takes 1,5 year, 'cause we're practising first [10:00:02] <trochej> So when the time comes, we do it roight :) [10:02:01] *** peteh has joined #opensolaris [10:03:43] <cmihai> WOOHOO! [10:03:47] <cmihai> Complete data recovery [10:04:10] <cmihai> I _love_ ZFS [10:04:20] <timsf> me too [10:04:32] <cmihai> Turns out the controller was dead [10:04:33] <cmihai> not the disk [10:04:35] <cmihai> But didn't show it. [10:04:42] <cmihai> Changed the controller [10:04:49] <cmihai> and ZFS is back up and running :D [10:06:24] <jteo> woot. [10:06:56] <Berny_> cool [10:07:04] <Berny_> saved you some time :-) [10:07:42] <cmihai> Nevermind the time, the backup window was too large, hence tons of data loss :-) [10:08:19] <Berny_> hehe quite lucky then [10:08:33] <cmihai> Yeah [10:08:50] <cmihai> I'm lucky I didn't get any data corruption either, ZFS took care of that. [10:09:01] *** WobbleU has left #opensolaris [10:09:26] *** Pietro_S has joined #opensolaris [10:09:38] *** C_REATiVE_ has joined #opensolaris [10:10:17] *** KermitTheFragger has joined #opensolaris [10:13:13] *** bengtf has quit IRC [10:18:07] *** NikolaVeber has joined #opensolaris [10:24:07] *** migi has joined #opensolaris [10:24:53] *** logic has joined #opensolaris [10:24:54] *** smtms has joined #opensolaris [10:26:53] *** FallenHitokiri has joined #opensolaris [10:28:10] *** srirama has joined #opensolaris [10:32:13] *** bengtf has joined #opensolaris [10:33:57] *** timelyx has quit IRC [10:39:42] *** logic_ has quit IRC [10:41:32] *** friendly12345 has left #opensolaris [10:42:47] *** Dar has joined #opensolaris [10:44:20] *** uebayasi has quit IRC [10:46:44] *** yongsun has joined #opensolaris [10:50:14] *** tsp has quit IRC [10:50:19] *** bnitz has joined #opensolaris [10:52:04] *** tsp has joined #opensolaris [10:52:23] *** tsoome1 has quit IRC [10:53:32] *** C_REATiVE_ has quit IRC [10:55:37] *** larryone has joined #opensolaris [10:56:31] *** ICU has quit IRC [10:57:52] <flyingparchment> my webconsole isn't working. when i send a GET request, it just closes the connection with no output. what's broken? [10:58:15] <larryone> morning [10:58:43] <larryone> is this the best place on freenode to discuss general solaris stuff? [10:58:53] <flyingparchment> here or #solaris [10:58:58] *** larryone has quit IRC [10:59:29] *** larryone has joined #opensolaris [11:00:15] <flyingparchment> oh, it's an ssl service. duh [11:02:33] *** simford has quit IRC [11:03:02] *** sniffy_ is now known as sniffy [11:11:26] <dsch04> Morning all [11:11:45] <dsch04> I've decided I've had enough of my new mobo - I'm looking for an alternative. [11:12:18] <dsch04> So, can anyone recommend me a board with the following features: [11:12:41] <dsch04> 1. supported by OpenSolaris/Solaris Express [11:12:50] <dsch04> 2. Core2Duo support [11:13:27] <dsch04> 3. >= 10 onboard SATA ports [11:13:28] <dsch04> or [11:13:39] <dsch04> 4. Some onboard SATA ports + PCI-X slot [11:13:53] *** ichigo has joined #opensolaris [11:13:56] <dsch04> 5. onboard VGA (preferable) [11:14:35] *** Bahra has quit IRC [11:14:56] <dsch04> 6. DDR2 800 RAM (PC6400) [11:15:01] <dsch04> Any ideas? [11:19:38] *** yongsun has quit IRC [11:21:02] *** sbahra has joined #opensolaris [11:25:38] *** tombhadAC has quit IRC [11:27:06] *** tsoome has joined #opensolaris [11:27:39] *** polk__ has quit IRC [11:31:52] *** dlg has joined #opensolaris [11:33:17] <ofu> dsch04: the ich8 and 9 have 4 sata-ports, ich8/9r and do 6 ports [11:34:09] <ofu> i use intel dg33tl, but some things do not work: ahci, onboard-nic and onboard-sound [11:34:22] <dlg> pwned [11:34:51] <ofu> if you need so many sata-ports, a pcie-controller seems to be a better/cheaper solution [11:36:30] <palowoda> ofu: Strange I thought ahci and the onboard nic should work. Are there known bugs? [11:37:36] <ofu> i had no time to search the database [11:37:56] <palowoda> what build are you running? [11:38:00] <ofu> nv75 [11:39:01] *** calumb has joined #opensolaris [11:39:40] <dlg> palowoda: im pretty sure the solaris ahci driver only attaches on controllers that say theyre ahci [11:39:46] *** cypromis has joined #opensolaris [11:39:52] *** cypromis has quit IRC [11:39:53] <dlg> a lot of ahci controllers say theyre raid in some configurations [11:40:07] *** cypromis has joined #opensolaris [11:41:44] <ofu> the intel bios lets me configure the sata ports as legacy, ahci and raid [11:42:46] *** sbahra has quit IRC [11:42:51] *** mega has joined #opensolaris [11:44:42] *** e^ipi_ has joined #opensolaris [11:44:43] *** e^ipi has quit IRC [11:46:05] *** mega has quit IRC [11:47:25] <dsch04> ofu, Can you point me at a cheap PCI-E SATA controller? [11:47:46] <ofu> hmmm, lsi? [11:47:59] [11:48:57] <tomww> PCI-X in a real -X Slot isn't the worst think.. [11:49:09] <dsch04> ? [11:49:20] <tomww> if you run it as 64 bit [11:49:33] * tomww leave chat the train reaches the station [11:49:37] <dsch04> Oh, I don't have a problem with pci-x [11:50:46] <ofu> has anyone heard about xfx sata-controllers? they are 34E for 5 ports (pci only) [11:50:47] <dsch04> Or pci-e [11:50:51] <palowoda> dsch04: 4 port pci-e sata http://www.addonics.com/products/host_controller/ [11:51:06] <dsch04> I want 10 ports [11:51:33] <palowoda> Get a motherboard with two pci-e slots and use a couple on the motherboard. [11:51:34] <dsch04> That would require at least two cards (assuming at least two ports onboarD) [11:51:56] <dsch04> Also, is that controller chipset supported? [11:52:43] <palowoda> Don't know the older addonics pci cards where tested. The company does active testing on solaris. [11:54:09] <palowoda> The other two companies LSI and Areca. [11:56:06] *** calumb is now known as calAFK [11:56:06] *** KermitTheFragger has quit IRC [11:56:42] <palowoda> http://www.areca.us/products/pcie341.htm but I head some grumbles that it wasn't working on build 72 where it was working previous on the storage-discuss list I think. [12:00:33] <dsch04> The thing is, I don't want/need expensive raid controllers. [12:00:43] <dsch04> I'm going down the OpenSolaris route because of zfs [12:00:58] <dsch04> I just need SATAII connections - lots of them! [12:01:51] <palowoda> Back to the supermicro pci-x based sata controller unless they came up with pci-e models now. [12:02:11] <dsch04> Yeah, hence my specifying a mobo with PCI-X slot [12:03:05] <palowoda> most likely some of the supermicro server board too. [12:03:20] <palowoda> not cheap. [12:03:48] <palowoda> That or Tyan. [12:04:00] <ofu> what about cooling? [12:04:15] <ofu> if you plan 10 or more disks, this might be an issue [12:04:24] <ofu> no tyan please [12:04:44] <palowoda> Yeah I avoid tyan nowdays. [12:04:49] * JoergB is back (gone 17:41:43) [12:05:01] <dsch04> I have Coolermaster cases, with 4-in-3 drive cages, each with a fan [12:05:20] <ofu> they are very loud... [12:05:39] <ofu> try supermicro mobile racks, they do 5 disks in the size of 3 cdroms [12:05:50] <ofu> buy a silent 90mm fan and everything is fine [12:06:17] <dsch04> ofu - they're not loud. Practically silent [12:06:25] <dlg> w [12:06:27] <dlg> oops [12:07:10] *** calAFK is now known as calumb [12:11:25] *** virgee has joined #opensolaris [12:11:26] <tsp> is there a way to unload all these kernel modules such as zones and dtrace in order to free ram? [12:11:45] *** sbahra has joined #opensolaris [12:12:07] *** KermitTheFragger has joined #opensolaris [12:12:49] <palowoda> Sure use modunload and watch the sparks fly. [12:13:59] <ofu> how much ram do they eat up? [12:16:07] <Pietro_S> tsp: I think that dtrace probes can't be unloaded ... [12:17:35] <palowoda> zones either [12:21:23] <tsp> I'm not sure how much ram they eat up, but I think unloading fibre channel or something freed up a few hundred k [12:21:54] <palowoda> You really want to chew up some kernel memory fire up zfs. [12:22:26] <Pietro_S> rather migrate from JDS to xfce or something like it to save memory, than unloading hardware drivers ... [12:23:26] <Pietro_S> does someone know where I can download that nice OpenSolaris template for OpenOffice Presenter (slides)? [12:24:11] <tsp> palowoda: I'm already running zfs on 256Mb that's why I'm trying to tune it down a bit :) [12:24:18] *** halton has quit IRC [12:25:02] <trochej> tsp: I believe that if you're not consuming dtrace probes, they are not residing in the memory. [12:25:04] *** halton has joined #opensolaris [12:25:09] <trochej> That's why they call it dynamic [12:25:36] *** halton has left #opensolaris [12:26:53] <palowoda> I don't know if one could create a stripped down kernel in order to get a reasonable system to work with zfs and such a small footprint of memory. [12:27:21] <trochej> Not me, for sure. :) [12:27:38] *** dsch04_ has joined #opensolaris [12:27:59] <palowoda> Heck I'd recommend 2G of memory just to start using zfs. [12:28:19] <trochej> Naaah, for a home desktop 1 GB is quite fine :P [12:28:37] <dlg> hrm [12:29:58] *** larryone has quit IRC [12:30:07] <tsp> I'm at a slight advantage since I don't use X, but sometimed ed can take 10 seconds to load even so [12:30:51] <timsf> 768 is enough for ZFS root on this laptop, [12:31:05] <timsf> and I've got a 256mb machine which acts as my home backup server : just depends on what you [12:31:10] <timsf> 're doing with it.. [12:32:31] <jteo> it depends on your workload. ARC is a monster. [12:32:52] <cmihai> Don't really see the debate. [12:33:12] <cmihai> Hell, the prices for 2GB DDR2 are the same as for 1GB DDR1 here ;P [12:33:32] *** cypromis has quit IRC [12:33:47] *** cypromis has joined #opensolaris [12:35:01] <trochej> Not here [12:35:05] *** tsoome has quit IRC [12:38:32] *** tsp has quit IRC [12:38:56] <palowoda> 2G DDR2 800mhz is about 85.00 here. Can't complain. [12:44:28] *** calumb has quit IRC [12:45:14] *** dsch04 has quit IRC [12:47:11] *** mikefut has joined #opensolaris [12:49:20] *** calumb has joined #opensolaris [12:51:38] *** cypromis has quit IRC [12:52:34] *** sparc-kly_ has joined #opensolaris [12:55:06] <JWheeler> Hi folks, I'm working my way through compiling the latest mplayer under solaris. Unfortunately it's now making EXTENTSIVE use of stdint.h, and I'm trying to compile this for BW, thus solaris 8. Solaris 8 uses inttypes.h instead. So I'm going through all the header files one by one and putting in an ifdef block. Is there somehow a faster way? [12:55:31] <JWheeler> I'm need to generically replace one line, with 5 lines, in a bunch of .h files essentially, can something like sed replace 1 line with 5? [12:55:49] <dme> Create a stdint.h that just does #include <inttypes.h> ? [12:56:09] <JWheeler> Where would I need to put it? [12:56:19] <JWheeler> So the lines are: "#include <stdlib.h>" [12:56:36] <dme> Somewhere in the include path - depends how mplayer is built. Look at the -I arguments to the compiler. [12:56:39] <JWheeler> pathwise, where is it looking? [12:56:44] <JWheeler> hmm [12:57:15] <trygvis> I'd rather make a complete patch as you can send that back to the project [12:57:34] *** tsp has joined #opensolaris [12:57:44] <dme> Sure, either works. [12:58:06] <JWheeler> so it's not just automatically looking in /usr/include for that kind of header, the -I flags given to the compiler decide where to look, rather then the C code itself? [12:58:28] <dme> The compiler will look in /usr/include automatically. -I adds to the list of places to look. [12:58:47] <JWheeler> Ah, I see [13:00:05] <tsp> I added this line to /etc/system: set zfs:zfs_arc_max = 0x2000000 [13:00:06] <JWheeler> Ok, so with a complete patch being the 'better' of the two solutions, I'm after a way to save my edit,make,edit,make,edit,make cycle :) [13:00:11] <tsp> but it doesn't seem to ahve an affect [13:01:16] <dme> JWheeler: "Life as a porting monkey" :-) [13:01:30] <JWheeler> heh, well I'm fighting the good fight... :) [13:01:43] <JWheeler> What I'm doing really is simple enough... but there must be a better way! [13:02:46] <Cyrille> well you could write a script which does your replacement (reads the files line by line and replaces #include<whatever.h> with your #ifdef block and then run a find -exec to apply the script to all headers... [13:03:16] <JWheeler> how would that script work? [13:03:20] *** sponix has joined #opensolaris [13:03:23] <JWheeler> the find -exec bit I know [13:04:05] <sponix> when is b70 out ? [13:04:32] <trochej> sponix: like five releases ago? [13:04:53] <Cyrille> the script, given a file name, would read the file line by line, if the line isn't the "#include <something.h>" you want to replace it outputs it to a temp file, if it is it outputs your replacement block to the temp file, at the end it moves the temp file to the original one. [13:04:59] *** salamanders has joined #opensolaris [13:05:45] *** NikolaVeber has quit IRC [13:05:54] <sponix> trochej: Crap, I meant b75... [13:06:03] *** NikolaVeber has joined #opensolaris [13:06:12] <JWheeler> hmm, I might be able to do that with perl - yes I know perl CAN do it, but can I do it? ;) [13:06:15] <sponix> Is it out now, saw b74 earlier today [13:06:37] <Pietro_S> sponix: today? [13:06:40] <trochej> sponix: b75 is out, sxce 75 isn't yet [13:06:43] <quasi> sponix: it is out as ON [13:06:44] <trochej> sponix: /topic [13:07:04] <dsch04_> What's ON? [13:07:19] <jteo> JWheeler: do or do not. there is no try. [13:07:44] <sponix> trochej: Thanks, I missed that in the topic, looking up I see it now [13:07:45] <trochej> dsch04_: Nevada [13:07:46] <JWheeler> snoogens! [13:07:50] <Cyrille> JWheeler, man read, man if, man echo, man mv ;-) [13:08:00] <JWheeler> ROFL [13:08:13] <JWheeler> I just don't think I could live with myself doing that [13:08:18] <trochej> dsch04_: http://www.opensolaris.org/os/community/on/ [13:08:41] <Cyrille> doing what? [13:09:17] *** hallu_ has joined #opensolaris [13:09:27] <trochej> read man, maaan [13:09:51] <jteo> i prefer women. but since i'm married that's a moot point. [13:11:36] <sponix> trochej: thanks for helping me out, even though my questions were more brain fart related [13:12:01] *** sparc-kly has quit IRC [13:13:04] <trochej> jteo: I know married men, who prefer other men, but are married for some reasons, so you know, says nothing :) [13:13:25] <jteo> trochej: touche. [13:14:09] <Giaco> where are cronjobs logged ? [13:14:16] <trochej> sponix: No problem, I am new to os, too :) [13:15:00] <sponix> trochej: I have one sparc solaris 8 box at work, other than that I only have a few months of play time [13:16:37] <palowoda> Giaco: man crontab [13:18:29] *** RainDoctor has joined #opensolaris [13:18:49] <jteo> mmm..pkg(5) looks fun. python too! [13:19:10] <Giaco> the /var/cron/log file is too short [13:19:27] <trochej> sponix: I have one intel solaris laptop at home [13:20:35] <palowoda> Only one? [13:20:37] *** tsoome has joined #opensolaris [13:21:49] <trochej> palowoda: Yes, I can't afford two private personal computers [13:22:01] *** hallu has quit IRC [13:22:04] <sponix> trochej: right now I'm still waiting on parts, so I'm at a bit fat zero until newegg gets me my stuff [13:22:23] *** jamesd_ has joined #opensolaris [13:22:27] <jteo> trochej: join the club. i only have one solaris box too. ;) [13:22:46] <sponix> trochej: soon, I'll have a sempron 3000+ with 1G Ram and 6 x 500G Drives in raidz [13:23:05] <sponix> jteo: hell of a lot better than none ;) [13:23:18] <jteo> sponix: be patient. ;) [13:23:56] <trochej> :) [13:24:15] <trochej> I used to have some at work, but since I switched I work with linux [13:24:15] <sponix> jteo: well, I'm half tempted drop it in vbox or vmware, just so I feel better when I run netbeans ;) [13:24:31] <jteo> sponix: zealotry clouds perception. ;) [13:25:18] <sponix> I began wanting to migrate from Linux and FreeBSD to Solaris from the time they started to offer free downloads [13:25:37] *** Dar has quit IRC [13:26:33] *** mikefut has quit IRC [13:26:47] <sponix> Not that I don't like FreeBSD, or Linux.. Just time for something new [13:27:04] <cmihai> That's not new, UNIX is ancient [13:27:07] <cmihai> Try OpenVMS [13:27:10] <cmihai> it's all the new thing I hear [13:27:21] <palowoda> CPM [13:28:16] <sponix> I didn't mean new development wise, I meant new to me [13:28:33] <sponix> OpenVMS and CPM aren't _new_ either :P [13:28:56] <cmihai> Newer than UNIX :-) [13:29:09] <trochej> Pan9? [13:29:14] <trochej> s/Pan/Plan/ [13:29:56] <jteo> now that's kinky [13:29:59] <jteo> AROS? [13:30:05] <trochej> Or, you can try AfrOS [13:30:17] <trochej> Unfortunately, no repo exists [13:30:23] <trochej> But THAT was a portable system [13:30:42] <trochej> It even worked on .gif [13:30:43] <trochej> :) [13:31:59] <trochej> It was barely usable, however [13:32:24] <trochej> The only thing that it did was print blin king butink... and stop :) [13:35:07] *** pablomh has quit IRC [13:35:24] *** pablomh has joined #OpenSolaris [13:36:08] *** jamesd has quit IRC [13:37:27] *** pablomh has quit IRC [13:41:58] <flyingparchment> hmm, when i try to log in with ssh using ldap+kerberos, the ssh server just closes the connection [13:42:57] <jteo> root login? [13:43:02] <Giaco> what's wrong in this command: /usr/sbin/logadm /opt/coolstack/var/logs/cache.log -C365 -a 'pkill -HUP squid' -z 0 ? [13:43:07] <flyingparchment> jteo: no, normal user [13:43:11] <jteo> hmm. [13:43:54] <flyingparchment> ah, it does log 'Failed keyboard-interactive for ...' - but if i enter the wrong password, it gives a normal error to the client.. [13:45:07] *** nostoi has joined #opensolaris [13:45:33] *** __hsilva has joined #opensolaris [13:45:48] <flyingparchment> oh, my fault. i forgot to add the host principle to kerberos [13:48:54] <jteo> god there is so little documentation regarding IPS. [13:49:29] <salamanders> could the harddisk be found on solaris10 x86 if i set the harddisk as "None" in BIOS ? [13:49:30] <timsf> yeah, it's a work in progress - haven't been able to find out much muself [13:49:44] <dlg> salamanders: try it [13:50:10] <jteo> timsf: couldn't you just walk over and ask sch? :P [13:50:38] <salamanders> there is no sol10 in my hand now :(. [13:50:38] <timsf> Yep, with a 10hr transatlantic flight - I'll get my coat :-) [13:51:36] *** timeless is now known as timeless_reading [13:51:40] <jteo> timsf: true. ;) [13:51:58] *** timeless_reading is now known as timely_changelog [13:55:59] <jteo> timsf: what precisly is the definition of an "image" ? [13:57:35] <timsf> My reading of it was a collection of installed packages on a running OS (or zone) [13:57:55] <timsf> but I'm not the expert here - wait a few hours till sch or danek show up [13:58:16] *** _hsilva has quit IRC [13:58:24] *** __hsilva is now known as _hsilva [13:58:52] <jteo> ah. [13:59:04] <jteo> i'm guessing the same thing too. [13:59:15] <jteo> time to do a hg clone and see what's going on... [14:00:13] <kjetilho> what's the name of the memcheck thingy (ala Purify or valgrind) in Sun Studio? [14:00:54] <timsf> Yeah - I'm thinking I should be on pkg-discuss to know more. There's some docs in the respository, [14:01:10] <timsf> but they're not complete (and it's hard to know where to start reading) [14:01:58] <jteo> timsf: i'm grokking those too. -sigh- [14:02:56] <trygvis> JWheeler: http://rafb.net/p/3nL8G471.html [14:03:22] <kjetilho> found it, libumem :-) [14:03:28] <jteo> ah. image = "collection of installed package instances" [14:03:33] <trygvis> that'll add the ifdef around the #include [14:03:40] <timsf> jteo - thanks! [14:04:17] <timsf> "Images can be of three types: full images, capable of providing a complete system; partial images, which are linked to a full image (parent image), but do not provide a complete system on their own; and user images, which contain only relocatable packages. (See NOTES on user images.)" [14:04:30] <timsf> from ./src/man/pkg.1.txt [14:04:31] <jteo> yup. [14:04:41] <jteo> i'm reading image.txt from the gate. [14:05:11] *** snuff-away has joined #opensolaris [14:18:55] <jteo> this prototype is rough. [14:19:28] *** MattMan has joined #opensolaris [14:20:11] *** kaiwai has joined #opensolaris [14:20:22] <kaiwai> 'allo :) [14:20:27] <jteo> wb kaiwai [14:20:44] <trochej> oh, kaiwai [14:20:51] <trochej> /leave [14:20:52] <trochej> :) [14:20:54] <kaiwai> you sound disappointed :) [14:21:02] <trochej> I do? :) [14:21:05] <kaiwai> trochej: its ok, I'm now a Mac user as of today :) [14:21:13] <kaiwai> well, back to Mac more correctly [14:21:24] *** phs2 has joined #opensolaris [14:21:58] <kaiwai> but my server is still Solaris though :) [14:24:08] <trochej> kaiwai: You finally use computer ptoper for your sexual orientation :) [14:24:35] <kaiwai> of course - apparently Apple is filled to the brim with limp wristers such as I :) [14:25:08] <bda> Statisically, that is untrue. [14:25:13] <kaiwai> anyway, I'm happy - so no more whinging and whining from me no longer :) [14:25:24] <bda> kaiwai: You'll just find things to complain about with the Mac. [14:25:31] <bda> Also... did you get one new? [14:25:40] <kaiwai> yeap [14:25:49] <kaiwai> 18months flirtation with a PC, now back to Mac :) [14:25:50] <bda> Why didn't you wait until the end of the month? [14:26:00] <kaiwai> why? [14:26:02] *** Gropi has left #opensolaris [14:26:05] <bda> Leopard. New iLife. [14:26:08] <kaiwai> I've got a free upgrade to Leopard :P [14:26:16] <bda> You won't get the iLife upgrade, afaik. [14:26:17] <kaiwai> and I've got iLife already [14:26:19] <kaiwai> '08 [14:26:22] <bda> Oh, ok. [14:26:25] <bda> Nevermind, then. :) [14:27:32] <kaiwai> the good side, I can go back to learning C as well, maybe one day I'll learn Objective-C [14:28:12] <dlg> mmm c [14:28:57] <kaiwai> yeah, my brain is too broken to learn complex things like object orientation [14:29:21] <holcomb> or maybe not broken enough [14:29:24] <FallenHitokiri> maybe I searched on the wrong locations but I found no useful informations - is there a way to convert a ext3 drive to zfs? [14:29:29] <jteo> FallenHitokiri: no. [14:29:32] *** wms has joined #opensolaris [14:29:35] <FallenHitokiri> jteo: thanks [14:29:45] <jteo> FallenHitokiri: besides a complete backup and restore. [14:29:48] <trochej> FallenHitokiri: Only by crewating a pool on one, but you loose al data [14:29:59] *** calumb has quit IRC [14:30:09] *** calumb has joined #opensolaris [14:30:27] <FallenHitokiri> okay,... now I have to find a way how to backup 800gb data... [14:31:27] <cmihai> FallenHitokiri, compress it ;_) [14:31:32] <cmihai> FallenHitokiri, what kind of data are we talking about? [14:31:53] <jteo> now wtf is distro-import. [14:32:02] <jteo> this is like some text adventure game. [14:32:03] <FallenHitokiri> cmihai: everything. videos from dv-tapes, images, backups,... [14:32:18] <ofu> no compression [14:32:26] <cmihai> FallenHitokiri with regular types of data, you get really good compression. Give rzip a try. http://unixsadm.blogspot.com/2007/10/rzip-great-compression-rates-for-logs.html . Even for stuff like backups and media, you can still get maybe 2:1 compression overall. [14:33:06] <cmihai> ofu, that really depends. I've gotten compression up to 10 times of what bzip can do for certain file types. [14:33:35] <ofu> 10x? wow [14:33:51] *** kaiwai has left #opensolaris [14:33:55] *** cypromis has joined #opensolaris [14:33:59] <sponix> rzip is _slow_ but it does rock on some data [14:34:00] <FallenHitokiri> cmihai: sounds good... if I get 2:1 I would not need to move 800gb to my office fileserver... thanks [14:34:44] <cmihai> Read that article mate. Rzip is made by the bloke who made rsync... and works on Samba. It uses something like rsync, with a 900MB history buffer (compared to 32k for gzip) for the first stage, and bzip (or LZMA for lrzip) for the second. Results are amazing. [14:34:49] <cmihai> And it's actually faster than bzip :P [14:34:56] <cmihai> sponix, rzip isn't slow. [14:35:01] <jteo> Tridgell? [14:35:22] <cmihai> It does have a high memory usage though. 500-900MB (due to the history buffer). [14:35:27] <cmihai> And you can't use pipes. [14:35:28] <FallenHitokiri> cmihai: is it really this hard to be faster than bzip? [14:35:35] <cmihai> :-) [14:35:41] <ofu> cmihai: your blog sounds interesting, rss available? [14:35:44] *** pablomh has joined #OpenSolaris [14:35:53] <cmihai> ofu, sure, click the link on top (the rss icon) [14:36:03] <cmihai> Or: http://unixsadm.blogspot.com/feeds/posts/default?alt=rss [14:36:36] <FallenHitokiri> thanks :) rzip is running,... time to get back to work [14:36:50] <cmihai> :-) [14:37:12] <pablomh> hi [14:37:34] <pablomh> is there a static % of memory dedicated to kernel/userland ? [14:37:35] <jteo> cmihai: another fan earned. :P [14:37:41] <cmihai> :-D [14:37:44] *** yongsun has joined #opensolaris [14:37:50] <jteo> pablomh: not that i know of. [14:38:02] <cmihai> Yes [14:38:11] <cmihai> pablomh, http://unixsadm.blogspot.com/2007/08/getting-detailed-system-information-on.html [14:38:14] <cmihai> First line [14:38:26] <ofu> cmihai: found the rss-feed in the sourcecode of the page, but i couldnt find the button to click on, strange [14:38:29] <cmihai> If you want to find out memory usage distribution [14:38:35] <yongsun> guys, do you know how xVM booting can work on a ZFS root fs? [14:38:44] <cmihai> ofu, oh, if you use firefox or something, it shows on top of the bar, next to the address [14:38:45] *** yongsun has quit IRC [14:38:55] <ofu> aaah [14:38:57] <cmihai> ofu, I'll add a static link though. [14:39:03] *** yongsun has joined #opensolaris [14:39:31] <ofu> i looked at the web page, not the address line [14:39:33] <pablomh> cmihai, the ::memstat one ? [14:39:41] <yongsun> guys, do you know how to boot xVM on a ZFS rootfs? [14:39:44] <cmihai> pablomh, yeah, though I might have misunderstood what you've wanted. [14:40:13] <cmihai> I've read statistic [14:40:41] <cmihai> pablomh, are you actually talking about the virtual address space? [14:41:41] <pablomh> I know in linux there's a % of virtual space, say 1/3 [14:41:44] <cmihai> Would be fun if rzip was also multithreaded. 7zip is, but even so, on a dual CPU system it's twice as slow as rzip for half the compression power ;-( [14:41:47] <pablomh> 1/3 for kernel [14:41:53] <pablomh> 1/4 [14:42:00] <LuckyLuke> hi there. I'm trying to start learning/using git and I see it isn't in blastwave. Is there any convenient way to put it on SXCE or I have to take the Way of the Source(tm) ? [14:42:01] <pablomh> and 3/4 for userland processes [14:42:02] <cmihai> pablomh, like in Windows, there is a 2GB reserved for system and 2GB for user processes? [14:42:08] <cmihai> And you can have 3/1 with /3GB mode? [14:42:09] <pablomh> yes [14:42:20] <pablomh> not 1/3 [14:42:29] <pablomh> 1/4 [14:42:31] <pablomh> typo [14:42:54] <pablomh> is there in solaris the same 2GB for system/kernel and 2GB for user ? [14:42:55] <cmihai> Well, this would only matter on a 32 bit system. [14:42:59] <pablomh> or is more dynamic ? [14:43:09] <pablomh> curiosity :) [14:43:09] <Pietro_S> LuckyLuke: in SFE there is [14:43:28] <Pietro_S> LuckyLuke: SFEgit.spec does the job ;-) [14:43:30] <cmihai> pablomh, you may want to take a look at the Solaris Kernel Tuning manual on docs.sun.com [14:44:32] *** sponix has quit IRC [14:44:47] <cmihai> pablomh, this: http://docs.sun.com/app/docs/doc/817-0404 [14:45:02] <pablomh> cmihai, thx for the tips :) [14:45:11] *** FallenHitokiri has quit IRC [14:45:12] <LuckyLuke> Pietro_S: please forgive my noob-ness but, what does SFE actually is? [14:45:23] <cmihai> pablomh, but like I've said, on a 64 bit system it wouldn't make a difference, you have a 64 bit address space (probably more like 40 bit in real life due to the lack of a proper 128 interlock exchange) and you'll never reach such a limit in real life. [14:45:24] *** larryone has joined #opensolaris [14:45:52] <cmihai> On a 32 bit system, where you have 4GB of virtual address space, you may want to allocate as much as 3GB to userspace processes. [14:50:07] *** bengtf has quit IRC [14:53:34] <jteo> mmm..the summit sounded like it was a blast. [14:54:13] *** calumb is now known as calLNCH [14:54:39] <flyingparchment> why is there no /etc/netgroups? [14:55:10] <tsoome> so u can make one yourself [14:55:23] <flyingparchment> making one won't do anything [14:55:30] <flyingparchment> what i mean is, why is there no files repository for netgroups? [14:55:32] <bda> Are you sure you want to uninstall zone lunix (y/[n])? y [14:55:32] <bda> The ZFS file system for this zone has been destroyed. [14:55:33] <bda> grrr [14:55:38] <bda> That's really annoying. :) [14:56:40] <kjetilho> flyingparchment: why would you want a file based netgroups ? [14:57:09] <kjetilho> distribute it via rsync? sounds a bit backward [14:57:13] <Pietro_S> LuckyLuke: take look here http://pkgbuild.sourceforge.net/spec-files-extra/ [14:57:50] <Pietro_S> LuckyLuke: it's repository of build recepies - so you can build software automagically ;-) [15:00:14] <LuckyLuke> ok, thank you very much. [15:01:35] *** yarihm has joined #OpenSolaris [15:03:42] *** tsp has quit IRC [15:04:38] *** sponix has joined #opensolaris [15:08:57] <nico> did anyone noticed load problems with iscsi ? [15:09:36] <cmihai> That would really depend. 2xGbE (aggr), doing VMware infrastructure on a RAIDZ array. [15:09:41] <cmihai> Load is quite low. [15:10:17] *** laca has joined #opensolaris [15:11:10] *** chrisso_ has quit IRC [15:12:16] *** bengtf has joined #opensolaris [15:13:14] *** deather_ is now known as deather [15:13:30] <nico> cmihai: I'm testing zfs & iscsi and after creating target, I loose control over the box during several minuts [15:14:00] <nico> bash-3.00# w [15:14:05] <nico> bash: fork: resource unavailable [15:16:04] *** jonkri has quit IRC [15:16:16] <flyingparchment> hm.. i'm mounting home directories from a zfs filesystem, and 'quota' is hanging on the client side. nfs/rquota is enabled. what did i miss? [15:16:31] <cmihai> nico, sounds like you're hitting a process limitation mate... [15:16:38] <cmihai> can you monitor ps -ef | wc -l? [15:16:53] *** tsp has joined #opensolaris [15:16:55] <nico> cmihai: nope, I can't even type :) [15:16:55] <cmihai> I see no reason you'd hit some kind of ulimit or system limit... though I guess that's possible [15:17:12] *** jonkri has joined #opensolaris [15:17:13] <cmihai> nico, see http://docs.sun.com/app/docs/doc/817-0404 [15:17:20] <cmihai> You may have to tune some kernel parameters [15:17:35] <nico> cmihai: gonna read this, thanks [15:20:31] <Symmetria> hrm, if I've manually changed the ndd settings which are defaults, without rebooting hte box is there a command I can issue to reset htem all to default [15:20:31] *** jcea has quit IRC [15:20:44] <Symmetria> (without having to look on another solaris box and find the defaults) [15:21:33] <PerterB> "uadmin 2 0" ;) [15:21:54] <tsp> wow my box flies now [15:22:20] <sponix> Anyone in here serve 100+ users with samba ? [15:22:31] *** virgee has quit IRC [15:23:12] <cmihai> sponix, got about 900 here I think. [15:23:57] <nico> sponix: yes [15:24:25] <Symmetria> errr PerterB I dont want to reboot the box [15:24:38] <Symmetria> I thought I was pretty clear on that [15:24:54] <PerterB> oh, I must have skipped over that bit [15:26:04] *** virgee has joined #opensolaris [15:26:54] *** cga has joined #opensolaris [15:27:35] *** nostoi has quit IRC [15:29:54] <quasi> I think the only way to get back to defaults without comparing to another box is a reboot [15:30:14] *** NikolaVeber_ has joined #opensolaris [15:31:47] *** derchris has quit IRC [15:32:32] *** orion has joined #opensolaris [15:34:48] <timsf> Unload the driver and load it again [15:35:05] <timsf> (might work - dunno) [15:35:12] *** Fish has joined #opensolaris [15:35:33] <timsf> Or dig around the source to find out what the defaults are... [15:36:36] <sponix> cmihai nico : I plan to have about 1.75-2TB of a raidz pool shared out to roughly 100 users any tips ? [15:37:08] <cmihai> sponix, nope, would work just fine. I've got a similar setup for user profiles for about 1000 [15:37:26] <cmihai> sponix, look into NIC bonding [15:37:27] *** mrman has joined #opensolaris [15:37:30] <cmihai> dladm create-aggr stuff [15:37:38] <ofu> should everybody have his own zfs filesystem? [15:38:09] <flyingparchment> ofu: yes. at least, there's no reason not to do that, and it's very flexible [15:38:14] *** mrman has left #opensolaris [15:38:22] <cmihai> ofu, sure mate, you can set quotas and reservations much easier. [15:38:24] <cmihai> It's how I do it. [15:38:24] <flyingparchment> ofu: (unless you have many thousands of users, then you might hit some performance issues) [15:38:25] *** mrman has joined #opensolaris [15:38:40] <sponix> Not sure how many ways I will share it out, figure vsftpd, samba, nfs, and debating an http solution [15:38:52] <cmihai> NFS would be best. [15:38:55] <cmihai> Easy to script too [15:38:59] <cmihai> Get a list of users [15:39:22] <cmihai> and zfs create bla bla zfs set sharenfs=rw or whatever [15:39:23] <sponix> 99% of the users will be on windows, so samba is a must [15:39:30] <cmihai> Windows has NFS :P [15:39:40] <cmihai> But yeah, Samba would work fine also. [15:39:57] <cmihai> Actually, I do Samba, NFS, sftp and vsftpd anyway. [15:40:02] <sponix> and box only has one nic, 10/100, but the fileshare will be off a wireless AP, so it will max out bw wise before the NIC gets close [15:40:36] <sponix> windows has nfs ? [15:40:39] <cmihai> Yeah, sure. [15:40:42] *** orion has left #opensolaris [15:40:51] <cmihai> It's part of Microsoft Services for UNIX (Now SUA / Interix). [15:40:56] <ofu> hmmm, perhaps i should do some testing with many zfs filesystems in one box [15:41:03] <cmihai> Comes by default in Windows 2003 and Windows 2008, and you can download it in Windows XP [15:41:11] <sponix> wow, didn't know that [15:41:16] <cmihai> You can share it just like SMB [15:41:27] <cmihai> Once you install SFU, it just adds a tab next to sharing [15:41:39] <cmihai> for NFS.. you can also "browse" the network as you would with SMB. Quite nice really. [15:41:40] <ofu> i know our windows admins had a few problems with sfu, but i have no details [15:41:49] <cmihai> Maybe they didn't map root -2 -2 [15:41:56] <cmihai> Oh well, they should read the manpage :P [15:42:05] *** timsf has quit IRC [15:42:37] *** capitano__ has joined #opensolaris [15:42:46] <cmihai> I [15:43:15] <cmihai> I've come to realize most complains about Windows aren't really founded. All the tools are there, all the scripting and command line tools are there.. you just have to know where to look :-) [15:44:09] * ofu is still searching for the messages on a windows host [15:44:20] <ofu> no, event manager is sometimes not detailed enough [15:44:54] <Symmetria> ok quick question. whats the command to loopback mount a directory from the base system into a zone again? [15:45:09] *** phimic has quit IRC [15:45:30] <ofu> on windows? net use \\bla foo: [15:45:46] * cmihai throws a boot at ofu [15:45:58] *** zeax has quit IRC [15:46:01] <Symmetria> errr no, mounting a solaris base directory into a zone [15:46:07] <cmihai> subst x: c:\bla [15:46:08] <cmihai> :P [15:46:19] <sponix> My number one complaint about Windows is uptime, it still wants you to pull drives offline, or restart all the freaking time ... Not much for serving imho [15:46:39] <cmihai> Huh? I can pull drives online just fine. [15:46:59] <cmihai> Got a Hitatchi Thunder FC storage with FC-AL disks, no problems pulling disks online [15:47:06] *** NikolaVeber has quit IRC [15:47:13] <sponix> Anyway... I'll look into SFU, if its easy enough an end user can install/configure I'll make screen snaps of exactly how to do it and then just drop that in a webpage for apache... to easy [15:50:12] *** tsoome has quit IRC [15:51:32] <sponix> the NFS/SFU is worth checking into.. Might still go with Samba just because the users are more use to setting seeing windows shares [15:53:02] <cmihai> Yeah, good point. [15:53:12] *** zeax has joined #opensolaris [15:53:36] *** zeax has joined #opensolaris [15:58:05] <sponix> plus they show SFU35SEL_EN.exe as a 200Meg Download, that doesn't support Vista... Sadly enough Vista is popular here [15:58:30] *** timsf has joined #opensolaris [15:59:21] <cmihai> sponix, it's in Vista mate [15:59:32] <cmihai> It's included in Vista Enterprise and Ultimate [15:59:35] <cmihai> They call it SUA [15:59:48] <cmihai> Though you can download / install it on any other. [15:59:48] <flyingparchment> hm, nfs and netgroups seem to hate me... if i share a filesystem using a netgroup, it lets anyone mount it [16:00:04] *** Giaco has quit IRC [16:00:13] *** calLNCH is now known as calumb [16:03:55] *** e^ipi_ has quit IRC [16:03:58] *** e^ipi has joined #opensolaris [16:04:08] *** vmlemon has joined #opensolaris [16:05:40] *** kadath has joined #opensolaris [16:05:43] <sponix> cmihai: thanks, I [16:05:43] <sponix> cmihai: thanks, I'll look into it [16:05:43] <sponix> honestly, debating just hosting filezilla over http, and using proftpd with ratios :P [16:05:49] <sponix> Most of the users are tards (idiots) though, so an ftp client may be way to complex for them [16:06:09] <cmihai> sponix, use SMB and add it to their profile. [16:06:18] <cmihai> It will automagically mount as a X: or whatever drive [16:06:25] <cmihai> If you use Active Directory, it's a breeze [16:06:31] <cmihai> They'll just see it as another disk :-) [16:06:44] <cmihai> No fuss, no mess, no work on their behalf. [16:06:56] <cmihai> I call it idiot-proofing :-) [16:09:03] *** tsoome has joined #opensolaris [16:09:21] <jteo> valued added service. [16:09:23] <jteo> *value [16:10:30] *** bengtf has quit IRC [16:11:19] *** larryone has quit IRC [16:11:41] <cmihai> Though, surprisingly, some of our idiots tend to be quite resilient. [16:12:10] *** postwait has joined #opensolaris [16:12:27] *** postwait has joined #opensolaris [16:25:50] *** timsf has quit IRC [16:28:02] *** Giahomo has joined #opensolaris [16:32:30] *** yongsun has quit IRC [16:33:11] *** bengtf has joined #opensolaris [16:33:31] *** tfb has joined #opensolaris [16:33:41] *** timsf has joined #opensolaris [16:34:03] *** loke__ has quit IRC [16:39:04] *** stevel has joined #opensolaris [16:39:04] *** ChanServ sets mode: +o stevel [16:40:28] *** srirama is now known as srirama|away [16:41:48] <jteo> wb stevel. [16:41:52] <stevel> morning jteo [16:42:01] *** joevandyk has joined #opensolaris [16:42:02] *** phs2 has quit IRC [16:43:59] *** bengtf_ has quit IRC [16:45:00] <timsf> hey stevel [16:45:07] <stevel> morning tim [16:45:11] <stevel> or evening/afternoon for you i guess [16:45:22] <timsf> is it? uh - no idea.. been mainlining caffiene all day [16:45:39] <timsf> seems to be working thus far [16:46:22] <jteo> jetlag a bitch eh timsf [16:46:38] <timsf> if men were meant to fly (etc. etc.) [16:47:42] <Cyrille> they'd be full of kerosene? [16:48:38] <timsf> 7 days is just the right amount of time to give jet lag. My body clock's dancing to a very weird circadian rhythm. Some sort of experimental jazz methinks [16:48:49] <stevel> could be worse. could be acid jazz [16:50:04] <Cyrille> nice... [16:52:56] <timsf> Acid jazz usually has some sort of constant rhythm though : I'm thinking more 60s experimental stuff [16:53:08] <tsp> hmm try Reggaeton [16:53:15] <timsf> you know, not so much jazz as random musical poking about [16:53:16] <stevel> so... just acid. [16:53:17] <stevel> :-) [16:53:19] <timsf> (which I can't stand) [16:53:20] <timsf> Yeah.. [16:53:26] <jteo> that's noise [16:54:28] *** migi has quit IRC [16:54:30] <timsf> snoop -a [16:55:01] <timsf> "Solaris, the musical" I can see it now! [16:55:06] *** derchris has joined #opensolaris [16:55:22] <stevel> .... are you sure you're not on acid? [16:55:24] <stevel> :-P [16:55:41] * timsf goes off to make sure that really was coffee... [16:55:47] <tsp> timsf: snoop -a is quite useful unless you get more packets than you can handle them you cannot shut it up [16:56:14] <timsf> Saved me a few times trying to find a machine in a lab (before we had all that nice programmable led stuff) [16:56:24] *** phs2 has joined #opensolaris [16:57:21] <tsp> just loop an audioplay :) [16:59:21] <Giahomo> does someone use sarg o light squid under solaris ? [17:02:28] *** tinman2k has joined #opensolaris [17:05:04] *** jerkface_ has quit IRC [17:08:01] <axisys> ottom: hey do know how to block access to some websites on sunray.. I get complain from security team about sunray users going to some sites that they should not [17:08:47] <axisys> ottom: should I install java webproxy server or some simpler soln out there? [17:09:25] *** e^ipi_ has joined #opensolaris [17:09:59] *** e^ipi has quit IRC [17:12:17] *** nostoi has joined #opensolaris [17:14:05] *** logic has quit IRC [17:14:08] *** logic has joined #opensolaris [17:17:00] *** phs2 has quit IRC [17:17:40] *** phs2 has joined #opensolaris [17:19:09] *** logic_ has joined #opensolaris [17:19:12] *** pauliukas has joined #opensolaris [17:19:23] <pauliukas> Something doesn't make sense here. [17:19:25] <pauliukas> I ordered the DVD... [17:19:30] <pauliukas> It's version 57 on the DVD. [17:19:35] <pauliukas> And now we're at 74, lol. [17:20:31] <tsp> at least Opensolaris ships free dvds, I don't know many others that would take the time to do that [17:20:48] <pauliukas> Ubuntu does it. [17:21:08] <pauliukas> But my real Solaris CDs are actually becoming more useful. [17:21:09] <cmihai> I think they only ship cds [17:21:30] <pauliukas> As I'm looking into using Solaris in production. I don't want to upgrade it each week [17:21:44] <cmihai> Nobody forces you to... [17:21:47] <pauliukas> lol [17:21:52] <cmihai> And even with SXCE, it's 2 weeks [17:21:53] <pauliukas> Yeah, but then you feel old. [17:22:11] <cmihai> I am old! [17:22:14] <pauliukas> It's the same feeling as I have with Fedora. [17:22:18] <cmihai> (no, not really) [17:22:18] <tsp> I'm still on snv_69 with a zfs pool from the 50's :) [17:22:23] <pauliukas> I want to run something stable an not upgrade every day [17:22:31] <Cyrille> That's called Solaris 10 [17:22:34] <g4lt-mordant> good god, I'm sort of annoyed now that I brought up my sb100 to nv55a, it used to be nv19 [17:22:49] <cmihai> Isn't that pre-ZFS? [17:22:52] <tsp> big jump [17:23:17] <g4lt-mordant> cmihai, yeah, but this is a SB100, not exactly known for its disk perfomance anyways [17:23:59] <tsp> damn my U5 is caching metadata for a bunch of files and not giving back the ram so everything swaps again [17:24:04] <jteo> i'm on 61. i'm middle-aged! [17:25:21] <tsp> I tuned the arc cache to be a max of 64MB, but I don't think I can tune how much metadata it caches [17:25:23] *** cypromis_ has joined #opensolaris [17:25:58] *** mikefut has joined #opensolaris [17:26:49] *** cypromis has quit IRC [17:27:53] *** phs2 has quit IRC [17:29:36] *** virgee has joined #opensolaris [17:34:38] *** jamesd_ has quit IRC [17:35:29] *** salamanders has quit IRC [17:39:10] *** cypromis_ is now known as cypromis [17:42:47] <gdamore> wow. I just got an OpenSSL announcement. what is weird about this is, I subscribed to the openssl lists during my *first* stint at Sun. For 4.5 years mail from openssl to me was bouncing, and they still didn't remove me from the lists! [17:44:47] *** jfndi has joined #opensolaris [17:44:52] <cmihai> gdamore, the vulnerability thing? [17:45:01] <gdamore> 0.9.8g. [17:45:13] <gdamore> i didn't look at it closely; i don't care about openssl anymore. [17:45:29] <gdamore> (5 years ago I did, but I was working on Sun crypto stuff back then.) [17:48:47] *** phs2 has joined #opensolaris [17:49:31] *** phs2 has joined #opensolaris [17:50:26] *** peteh has quit IRC [17:51:14] *** slowhog has joined #opensolaris [17:54:19] *** charlesnw has joined #opensolaris [17:58:49] *** jfndi has left #opensolaris [18:01:29] *** nostoi has quit IRC [18:03:16] <flyingparchment> "The SNMP master agent fails to start on Solaris 10U2 (and higher) due to the TCP fusion feature added in Solaris 10U2." < broken :-( [18:04:58] <HCoyote> you can dturn off tcp fusion. [18:05:03] <HCoyote> s/dturn/turn/ [18:05:18] <flyingparchment> i know, but it's still broken ;) [18:05:29] <HCoyote> hey, I'm just an idea man. ;-) [18:06:22] <flyingparchment> what does tcp fusion do exactly? i assume on an unloaded system i'm not going to notice any issues turning it off.. [18:06:41] *** WickyReally has joined #opensolaris [18:06:42] <WickyReally> howedy [18:08:50] <flyingparchment> ah, faster loopback communications [18:09:05] <TrogL> back. somebody was helping me yesterday with ORACLE box taht woulnd't swap. any update? [18:10:53] <cmihai> Still no ideas :-) [18:11:18] <cmihai> Spent all day recovering a crashed RAID array [18:11:21] <pauliukas> A box that doesn't swap... [18:11:25] <TrogL> cmihai: I thought of running multiple instances of my ram-gobbling C program. will try that now [18:11:28] <pauliukas> cmihai: Wow. That's doable? [18:11:31] <pauliukas> What kind of RAID? [18:11:41] <cmihai> RAIDZ [18:11:50] <pauliukas> RAIDZ can crash? [18:11:53] <cmihai> The controller was dead though. [18:12:07] * pauliukas goes to the corner and shivers [18:12:20] <pauliukas> Isn't RAID-Z all software? [18:12:30] <jteo> his controller died. unrelated to software. [18:12:36] <cmihai> Well, thankfully, I didn't loose anything, on resilver it was all back up and running :-). [18:12:49] <g4lt-mordant> did you lose anything either? [18:12:51] <pauliukas> The normal SATA/SAS controller? Or does RAIDZ have some sort of hardware controller thing? [18:12:52] <cmihai> I think there was some data corruption, but RAIDZ / zfs fixed that. [18:12:59] <cmihai> g4lt-mordant, nope, all there. [18:14:06] <pauliukas> Wow. You're a computer g0d. [18:16:04] <WickyReally> or just lucky [18:16:08] <pauliukas> heh [18:16:24] <WickyReally> I woreship him though, he is kinda god-is [18:16:25] <pauliukas> What were to happen if I had a 4 disk RAID-Z. And 2 of the 4 disks would crash. [18:16:30] <pauliukas> Would it be recoverable or no? [18:16:41] <WickyReally> if it is the controller, yes [18:16:44] <WickyReally> if it are two disks, no [18:16:55] <pauliukas> O [18:17:05] <pauliukas> I've had a friend who had a RAID5 and 2 disks crashed. [18:17:14] <pauliukas> And his backups weren't working. He was [18:17:15] <WickyReally> you can survive a double disk failure, if the data and disks are still accessible [18:17:17] <pauliukas> "fucked" [18:17:36] <pauliukas> Uhhh [18:17:37] <cmihai> Hence, RAIDZ-2 and proper backups :-) [18:17:37] <timsf> Raidz-2 would allow for two disks failing [18:17:49] <pauliukas> Yeah? But how many drives does it need? [18:17:53] <timsf> 3 [18:17:56] <WickyReally> 3 [18:18:00] <pauliukas> Does ZFS have something equivalent to RAID10? [18:18:01] <stevel> Three [18:18:02] <stevel> :-D [18:18:06] <timsf> Trois [18:18:09] <WickyReally> drie [18:18:10] <WickyReally> tres [18:18:12] <pauliukas> How does that work? Magic [18:18:13] <pauliukas> ? [18:18:13] <HCoyote> pauliukas: yes. mirror. [18:18:22] <stevel> San [18:18:25] <stevel> (chinese) [18:18:26] <sickness> 3 is stupid ghgh, better go for mirroring then :P [18:18:29] <pauliukas> HCoyote: RAID10 is mirror plus striping. Or the other way around. [18:18:33] <cmihai> RAID10 is a mirror of stripes... or a stripe of mirrors, depends how you see it. [18:18:34] <WickyReally> jedna i dva (one and two) :P [18:18:37] <cmihai> 0+1 [18:18:38] <pauliukas> Well, I'd want that for performance. [18:18:46] <HCoyote> pauliukas: you attach multiple mirror vdevs to a pool. it becomes 0+1. [18:18:48] <cmihai> Yes, RAID-10 gives good performance [18:18:49] <stevel> 11 (binary!) [18:18:52] <jteo> all ZFS pools are always striped. [18:18:57] <cmihai> And you can fail two disks if they're the right onesw [18:18:58] <pauliukas> Sweet. [18:18:59] <WickyReally> stevel: you win :P [18:19:07] * stevel bows [18:19:10] <pauliukas> Now, here's another question... [18:19:15] [18:19:16] <pauliukas> Can ZFS help me create a ghetto array? [18:19:17] <sickness> "all your ZFS pools are belong to us" [18:19:18] <WickyReally> III [18:19:20] *** Giahomo has quit IRC [18:19:27] <pauliukas> As in... Hook up multiple different sized drives. [18:19:28] <stevel> oooh. roman numerals. [18:19:32] <pauliukas> And make it all redundant. [18:19:41] <WickyReally> I'm going to fetch a taco [18:19:42] <jteo> woohoo! dirt slinging in pkg-discuss! [18:19:42] <WickyReally> bbl [18:19:51] <cmihai> Nope. You can have JBOD though. [18:20:00] <pauliukas> And what's JBOD? [18:20:00] <cmihai> You probably want block level sort of RAID [18:20:05] <cmihai> just a bunch of disks [18:20:12] <cmihai> It's what it says. [18:20:13] <cmihai> Like a ccd [18:20:17] <tfb> pauliukas: yes, it can, but you will prbably lose some space [18:20:18] <cmihai> concantenated disk [18:20:35] <pauliukas> tfb: As in it will all be limited to the space of the smallest drive? [18:20:38] <cmihai> Yep [18:20:47] <pauliukas> Yeah, I was trying to avoid that. [18:20:52] <cmihai> I figured. [18:20:52] <pauliukas> The Drobo can do this... [18:20:55] <pauliukas> But it's like $600. [18:20:58] <cmihai> The Drobo is junk. [18:21:03] <jteo> really? [18:21:03] <pauliukas> I'd rather build a whole new computer if I were to do it. [18:21:09] <tfb> each mirror will probably be as small as the smaller drive. [18:21:22] <tfb> But, you can construct the mirrors from slices [18:21:42] <pauliukas> The whole sexy thing about ZFS is the pooling stuff. [18:21:58] <tfb> so you could have (say) s0 on one drive mirrored against another, and s1 against yet anothe r [18:21:59] <pauliukas> I thought it would be cool to just assemble all the old drives that are lying on my desk into one machine. [18:22:04] <pauliukas> And just let it go ctazy. [18:22:05] <tfb> but you would go mad trying to get it right [18:22:09] <pauliukas> lol [18:22:34] <timsf> Watch out, therein lies madness. [18:22:38] <tfb> and given what disks cost now ... [18:22:51] <tfb> timsf: yes. [18:22:53] <timsf> (when you start trying to compose a pool with lots of weird devices...) [18:22:57] <pauliukas> I'm trying to save up money for other stuff :-P [18:22:58] <timsf> http://blogs.sun.com/timf/entry/zfs_in_perfect_harmony [18:23:06] <jteo> is that the caffeine talking tim? [18:23:17] <tfb> don't they give disks away with food yet? [18:23:25] <pauliukas> hah. no. [18:23:28] <tfb> I thought they did [18:23:35] <timsf> Probably. [18:23:40] <pauliukas> A 250GB drive is still about a hundred. [18:24:14] <tfb> ah, obviously I eat the wrong kind of food [18:24:25] <jteo> so the question is: is nourishment more important than porn? [18:24:29] *** spurton has joined #opensolaris [18:24:37] <pauliukas> I just wanted a nice array to keep my backups. [18:24:44] <pauliukas> But I guess I'll just use my S3 account. [18:24:55] <pauliukas> And save up my money for a new Solaris server. [18:24:56] <sickness> jteo: sure it is :P [18:25:08] <g4lt-mordant> tfb yeah, I got a RAID-2 with my breakfast cereal last week [18:25:13] <pauliukas> lol [18:25:14] <sickness> well, sorry, I misunderstood the question :| [18:25:24] <sickness> pr0n indeed is more important... [18:25:32] <timsf> Hardware raid2 with you cereal - isnt that a bit flakey ? [18:25:40] <timsf> ba-dum, tcsh [18:25:40] <pauliukas> OH... Excellent point! [18:25:41] <TrogL> geez, I should start eating breakfast again [18:25:46] <timsf> (Im here all week folks) [18:25:53] <jteo> i didn't know people still used RAID-II. [18:26:00] <pauliukas> What's RAID2? [18:26:07] <TrogL> bug spray [18:26:14] <pauliukas> ah... [18:26:17] <pauliukas> hah. [18:26:43] <TrogL> and if you believe that... [18:27:08] <pauliukas> I know that RAID is a bug spray. [18:27:19] <pauliukas> I just thought you meant a RAID array at level 2. [18:28:51] <tfb> jteo: well, the question is really whether nourishment is more important than yet more ISO images of Solaris releases [18:29:33] <jteo> tfb: only men who run solaris even ponder such questions. [18:29:54] * tfb thinks: man can not live on Solaris releases alone [18:30:17] <libkeiser> but the bits are soooo tasty [18:31:25] <g4lt-mordant> pauliukas, they didn't just jump from raid 1 to raid 5 one day.... [18:31:32] <pauliukas> Yeah, no kidding. [18:31:37] <pauliukas> I'm reading up on Wikipedia. [18:32:25] *** mikefut has quit IRC [18:36:19] *** nostoi has joined #opensolaris [18:36:28] <flyingparchment> is making sendmail use ldap as simple as 'aliases: files ldap' in nsswitch.conf? [18:37:25] *** centrinia has joined #opensolaris [18:38:28] *** WickyReally has quit IRC [18:39:47] *** tfb has left #opensolaris [18:40:26] *** loky has joined #opensolaris [18:41:53] *** loky has quit IRC [18:42:18] *** loky has joined #opensolaris [18:44:48] *** centrinia has quit IRC [18:46:45] *** cmihai has quit IRC [18:50:33] *** pjlv has quit IRC [18:52:49] *** dlynes_laptop has joined #opensolaris [18:54:18] *** Cyrille has quit IRC [18:55:47] *** sparvu has joined #opensolaris [18:59:04] *** bondolo has joined #opensolaris [18:59:05] *** bnitz has left #opensolaris [19:03:09] *** KermitTheFragger has quit IRC [19:06:46] *** phs2 has quit IRC [19:07:38] *** phs2_ has joined #opensolaris [19:08:06] *** flyingparchment has quit IRC [19:08:26] *** virgee has quit IRC [19:09:21] *** flyingparchment has joined #opensolaris [19:11:08] *** phs2_ has quit IRC [19:13:19] *** edvil has left #opensolaris [19:14:32] *** bunker has joined #opensolaris [19:14:48] *** phs2 has joined #opensolaris [19:16:03] *** WickedWicky has quit IRC [19:17:22] *** RealWickedWicky has joined #opensolaris [19:17:25] <RealWickedWicky> hmmm [19:17:34] <RealWickedWicky> that upgrade didnt go as smooth as I was expecting it to go [19:19:40] *** deather_ has joined #opensolaris [19:22:56] <RealWickedWicky> I compiled onnv_76 , bfu-ed, the machine is back [19:23:13] <RealWickedWicky> but everything that needs libcrypto.so dies by complaining the librarie doesnt exist [19:23:29] <RealWickedWicky> library even [19:24:05] <TrogL> is the library libs in your LIBPATH? [19:24:18] <TrogL> crle [19:24:44] <RealWickedWicky> I might hope so, I didnt touch crle [19:25:08] <TrogL> they usually end up in /usr/local/ssl/lib - which isn't norammly on the path [19:27:34] <RealWickedWicky> the particulair library is in /usr/lib [19:27:48] <RealWickedWicky> and crle states /lib:/usr/lib is in the path (system default) [19:27:59] <TrogL> in that case, dunno [19:28:02] <holcomb> isn't libcrypto in /usr/sfw/lib? [19:28:23] <RealWickedWicky> nope, it's in /usr/lib [19:28:26] <RealWickedWicky> I just checked [19:28:31] <RealWickedWicky> it's libcryptoutil.so.1 btw [19:28:35] <RealWickedWicky> (the one it's complaining about) [19:28:40] *** sponix has quit IRC [19:28:49] <RealWickedWicky> even more interesting is that my sshd binary seems to have disapeared [19:28:56] <flyingparchment> is there any way to put ssh public keys in ldap using sun ssh? [19:29:12] <jteo> that's why they call it BFU. [19:29:23] <RealWickedWicky> or actually, it's in /usr/lib/ssh/ [19:29:31] <RealWickedWicky> and I can ssh to the box after doing a ./sshd [19:29:41] <RealWickedWicky> but svcadm sshd start returns me to the command line, and no ssh [19:30:27] <wesolows> look at the ssh service log [19:30:32] <RealWickedWicky> Oct 19 19:27:28 lilith sshd[161326]: [ID 522540 auth.error] libpkcs11: Cannot load PKCS#11 library /usr/lib/security/pkcs11_softtoken_extra.so. dlerror: ld.so.1: sshd: fatal: libcryptoutil.so.1: open failed: No such file or directory. See cryptoadm(1M). Skipping this plug-in. [19:30:56] <RealWickedWicky> Oct 19 19:28:22 lilith sshd[161894]: [ID 591423 auth.error] libpkcs11: Cannot load PKCS#11 library /usr/lib/security/pkcs11_kernel.so. dlerror: ld.so.1: sshd: fatal: libcryptoutil.so.1: open failed: No such file or directory. See cryptoadm(1M). Skipping this plug-in. [19:32:37] <TrogL> Ok, so find libcryptoutil.so.1 [19:32:43] <RealWickedWicky> it's in /usr/lib [19:33:04] <RealWickedWicky> # cd /usr/lib [19:33:04] <RealWickedWicky> # ls libcrypto* [19:33:04] <RealWickedWicky> libcryptoutil.so libcryptoutil.so.1 [19:33:15] <RealWickedWicky> Default configuration file (/var/ld/ld.config) not found [19:33:15] <RealWickedWicky> Platform: 32-bit LSB 80386 [19:33:15] <RealWickedWicky> Default Library Path (ELF): /lib:/usr/lib (system default) [19:33:27] *** jonkri has quit IRC [19:33:28] * TrogL looks puzzled [19:33:40] <RealWickedWicky> and starting sshd with ./sshd works [19:33:52] <TrogL> hang on, try... [19:33:58] <TrogL> ld -l libcryptoutil.so.1 [19:34:25] <RealWickedWicky> # ld -l libcryptoutil.so.1 [19:34:25] <RealWickedWicky> ld: fatal: library -llibcryptoutil.so.1: not found [19:34:25] <RealWickedWicky> ld: fatal: File processing errors. No output written to a.out [19:35:00] <RealWickedWicky> libcryptoutil.so.1: ELF 32-bit LSB dynamic lib 80386 Version 1, dynamically linked, not stripped, no debugging information available [19:35:08] *** jcea has joined #opensolaris [19:35:25] <TrogL> sorry, ldd -l [19:35:26] *** JWheeler has quit IRC [19:35:40] <RealWickedWicky> # # ldd -l libcryptoutil.so.1 [19:35:41] <RealWickedWicky> libc.so.1 => /lib/libc.so.1 [19:35:41] <RealWickedWicky> libm.so.2 => /lib/libm.so.2 [19:36:06] * TrogL shuts up [19:36:19] <RealWickedWicky> please dont [19:36:20] <RealWickedWicky> :P [19:37:38] *** JBeck has joined #opensolaris [19:37:38] *** ChanServ sets mode: +o JBeck [19:37:43] *** spurton has quit IRC [19:37:50] *** JWheeler has joined #opensolaris [19:39:14] *** deather has quit IRC [19:41:09] *** ZeleniZub__ has joined #opensolaris [19:43:16] *** sommerfeld has joined #opensolaris [19:43:16] *** ChanServ sets mode: +o sommerfeld [19:44:30] <TrogL> RealWickedWicky: I dunno. I'm stuck. I still think it's something to do with the path to the libraires being different somehow. [19:44:46] <RealWickedWicky> I'm recompiling the source [19:44:48] <RealWickedWicky> let's see [19:45:04] <TrogL> ls -l /usr/lib | grep libcrypoutil [19:45:23] <gdamore> very cool. alanc just dropped by and dropped off an SB2500 for me. [19:45:35] <flyingparchment> pah. i want free computers too [19:46:13] <RealWickedWicky> $ ls -l /usr/lib | grep libcrypto [19:46:13] <RealWickedWicky> lrwxrwxrwx 1 root root 18 Oct 19 11:20 libcryptoutil.so -> libcryptoutil.so.1 [19:46:29] *** cydork has quit IRC [19:47:57] <cygnusecks> is dduvall in the house? [19:48:12] <dduvall> In the hizouse, actually ... [19:48:28] <cygnusecks> hey danek, this is james falkner.. just had a quick question.. you ever use any python IDEs? [19:48:37] <dduvall> Nope. [19:48:53] <cygnusecks> ok.. I tried a few out earlier to day and not much success other than glorified syntax highlighting [19:49:04] <cygnusecks> so was wondering if anyone else such as yourself had used any that were "worth it" [19:49:31] <dduvall> I've heard of people working on some, but I don't remember their names. [19:49:41] *** RaD|Tz has joined #opensolaris [19:49:42] <dduvall> IDEs don't really excite me. [19:49:43] <stevel> vim is my co-pilot [19:49:46] <cygnusecks> there's a big list on wikipedia, but most of em suck [19:49:53] <dduvall> stevel rules [19:50:12] <dduvall> Sorry, r00l3z [19:50:15] <cygnusecks> yeah, mine too but I really like to have debuggers handy where I can step through code, and also TAGS-like abilities [19:50:18] <cygnusecks> vim do that? [19:50:44] <dduvall> vim has the ability to be led by the nose by a debugger, but not the other way 'round, so far as I know. [19:50:47] <cygnusecks> I like to jump around without having to hunt for symbols, etc. and get back to where I was [19:50:50] <dduvall> It does do tags, of course. [19:51:00] <cygnusecks> python too? [19:51:06] <dduvall> You just need exuberant ctags to generate python tags. [19:51:11] * stevel doesn't debug python code ;-) [19:51:16] <cygnusecks> I've used TAGS for C and java, but not for python [19:51:17] <flyingparchment> i find IDEs useful for java, but not much other languages [19:51:30] * dduvall thinks that says something about java [19:51:43] <flyingparchment> yes, it's very much a feature of the language ;) [19:52:02] <cygnusecks> heh.. <backs away slowly> [19:53:35] *** IvanR_ has joined #opensolaris [19:53:45] <jteo> oldschool. [19:54:01] <jteo> dduvall: is there actually a description of the native package manifest for IPS? [19:54:27] <dduvall> Not officially -- it's a private interface at the moment. [19:54:38] <jteo> ah. [19:54:46] <jteo> there's supposed to be one i presume? [19:55:12] <dduvall> But it's one action per line. Action name followed by a content hash if the action has content, followed by space-separated "key=value" pairs of metadata. [19:55:21] *** MattMan has quit IRC [19:55:34] <dduvall> We'll eventually have some sort of "prototype"-like file that people will write when developing packages. [19:55:36] *** g4lt-mordant is now known as g4lt [19:55:42] <jteo> intriguing. [19:55:49] *** srirama|away has quit IRC [19:56:12] <cygnusecks> and dduvall there's no serialized package form yet, right? there was the email from the pkgbuild author stating that he was going to start outputting "IPS binaries" [19:56:25] <cygnusecks> hadn't seen anything like that in the source, so... [19:56:31] <dduvall> I saw. Presumably he means making pkgbuild publish to an IPS repo. [19:56:52] <cygnusecks> yeah thats what I was wondering.. guess I'll ask him for clarification [19:56:58] *** g4lt is now known as g4lt-mordant [19:58:20] *** NikolaVeber_ has quit IRC [20:03:07] *** srirama has joined #opensolaris [20:10:46] *** mega has joined #opensolaris [20:12:19] *** Gman has joined #opensolaris [20:12:20] *** ChanServ sets mode: +o Gman [20:12:27] *** calumb has quit IRC [20:22:43] <g4lt-mordant> is the sb100 smartcard stuff part of ON? [20:23:41] <Berny_> lo glynn [20:23:53] *** cmihai has joined #OpenSolaris [20:26:37] <stevel> g4lt-mordant: i thought smartcard was being taken out [20:26:43] <stevel> you should ping bubbva [20:27:20] <g4lt-mordant> bummer, because zfs crypto + smartcard = win [20:29:23] *** mw46 has joined #opensolaris [20:29:54] <g4lt-mordant> especially if you could pass a token to unlock remote systems, as in I put my smartcard in my sunray, which then unlocks $HOME and I do work, out goes the smartcard, nobody has a chance to get my files [20:31:15] <flyingparchment> that's awesome until you lose the card ;) [20:31:41] <stevel> g4lt-mordant: http://www.sun.com/bigadmin/content/eof/s10u3eof.html [20:31:47] <cmihai> There are recovery mechanisms. If you want them. [20:31:53] <stevel> scroll down to "Features That Might Be Removed in a Future Release" [20:32:27] <flyingparchment> haha, admintool [20:32:35] <cmihai> Take encryptions with certificates for one. If you have your own root CA for example, you can opt for being able to recover lost certificates. Add the need for 3 out of 5 administrative accounts to perform such a recovery, and you're set. [20:34:46] <cmihai> Happens if you have a root CA that gives users certificates for use on encrypting filesystems / files / mails / etc. Sooner or later, they'll either lose them, or quit and you need to recover _your_ data from their machines. [20:35:21] <g4lt-mordant> for might, read "will"? [20:35:31] <sommerfeld> also known as "key escrow". [20:35:45] <g4lt-mordant> so what'll happen with SRSS if the smartcard dies? [20:35:50] <sommerfeld> g4lt-mordant: my understanding is that the existing smart card framework is unloved. [20:35:56] <sommerfeld> SRSS doesn't use the smart card framework [20:36:10] <sommerfeld> there's another one out there which is loved but not funded for integration. [20:36:40] <g4lt-mordant> is it open though? if it's open, maybe I can spare some lovin ;P [20:36:53] <cmihai> sommerfeld, yep. Some companies even use escrow agencies. [20:36:56] <cmihai> Though I don't trust that... [20:37:27] <sommerfeld> the critical thing is to escrow only encryption keys and not signature keys. [20:37:48] <cmihai> Yeah. [20:38:03] <cmihai> Actually, you shouldn't even have recovery for signature keys. [20:38:04] <ottom> sommerfeld: SRSS exposes smartcards to applications via the PC/SC API [20:38:09] <cmihai> Just for the encryption keys. [20:39:18] <sommerfeld> right, PC/SC & MUSCLE was the thing that's inbound. [20:39:39] *** Gman has quit IRC [20:39:40] <sommerfeld> .. but unfunded. speak up on security-discuss if you want to help. [20:39:44] <ottom> g4lt-mordant: if your card dies then you can alias a new card to the old one if you've configured SRSS to do card registration. Otherwise you can go in behind the scenes and tweak a few files to allow a new card to get your old session. That's totally unsupported, of course. [20:40:40] <ottom> unfunded, but implemented. For Sun Ray on Solaris, anyway. Not on Linux yet. [20:43:13] *** l1s has joined #opensolaris [20:45:09] <sommerfeld> that annoying organizational disconnect again (too many cases where sunray capabilites have needlessly diverged from non-sunray) [20:47:34] *** webmink has joined #opensolaris [20:48:33] *** RaD|Tz has quit IRC [20:49:36] <cmihai> I might have brought this up before, but do you have a CISSP sommerfeld? [20:49:52] <sommerfeld> hah. no. [20:50:09] <cmihai> Hm... my mistake :-) [20:50:24] <cmihai> How about... work on NetBSD? [20:50:48] <sommerfeld> i tend to spot mistakes made by people who flaunt their CISSP credentials in public. [20:50:55] <sommerfeld> cmihai: used to. ENOTIME. [20:51:29] <cmihai> I see. I had this weird deja-vu feeling, that's all :-). I remember seeing your name in such a context before. Oh well. [20:52:34] <jteo> infinite work..finite .. [20:52:56] <sommerfeld> (my theory after seeing a bunch of people claiming CISSP credentials demonstrate ignorance on certain matters of security was that the certification had negative value) [20:53:59] <cmihai> Well, some tend to be ingorant on technical matters. The CISSP is more.. managerial in nature. [20:54:32] <cmihai> Did the courses and read a bunch of books / material, and have spotted quite a few mistakes or vague subjects. [20:54:51] <cmihai> Talked about getting the cert, but they said I wasn't "experienced" enough. [20:54:57] <sommerfeld> there was an interesting social experiment at an IETF one year. the IETF uses color-coded dots on the conference badge to indicate various roles you have (working group chair, iab member, iesg member, local host, etc.) [20:56:38] <sommerfeld> someone bought a bunch of dot stickers of an otherwise unused color, called them "clue dots" and gave them to people he trusted to be clueful with instructions to give them out to people they trusted to be clueful. [20:57:05] <cmihai> heh [20:57:27] <cmihai> Let me guess, when you data mined the results, the CISSP's had none? [20:57:41] <sommerfeld> no CISSP's in sight. [20:58:10] <cmihai> You know, I could try the same social experiment here, with a twist. [20:58:12] <cmihai> But give it to clueless people [20:58:14] <cmihai> and make the clue dots brown [20:58:26] <cmihai> Hint: brown-nosers :D [20:58:35] <sommerfeld> but the experiment was terminated early when some people active in the routing area were upset that people they considered clueless had clue dots. [20:58:48] <cmihai> Heh [20:59:04] <cmihai> Yeah. A "double blind" experiment would work fine here. [20:59:37] <cmihai> Give the dots to both gormless and clued people... or count the number of dots attributed to a person. You got me thinking :-). [21:00:34] <flyingparchment> does N1 system manager manage the hardware, or the OS, or both? [21:00:55] <elektronkind> both [21:01:03] <flyingparchment> can i use it to manage solaris on a non-sun system? [21:01:16] <elektronkind> it knows how to talk to the iLOM/SP/ALOM and gets hardware status from that [21:01:40] <elektronkind> I haven't tried that with it. [21:01:43] *** polk__ has joined #opensolaris [21:01:59] <flyingparchment> hmm [21:02:18] <elektronkind> then again, I don't really use it in the course of day to day operations... I only have it running on a ancilliary box just to see what it's all about and how it works. [21:02:26] *** webmink has quit IRC [21:02:27] <flyingparchment> is it any good? :) [21:02:29] <sommerfeld> cmihai: anyhow, my point was that the CISSP is very much like the clue dot... [21:02:39] <flyingparchment> a central place to manage SPs and such might be useful.. [21:03:14] <elektronkind> flyingparchment: for central SP management, it's great for that. for the OS side, you can manage installs from it (linux installs, too) [21:03:39] <elektronkind> it's designed so that your operations for the included machines are based around it though [21:04:08] <elektronkind> but it's good if you just want to ignore the OS side of things and just use it for SP management [21:04:25] <elektronkind> its built-in CLI is pretty nice [21:05:33] <flyingparchment> maybe i can use it an an excuse to buy sun instead of dell next time, too ;) [21:05:50] <cmihai> sommerfeld, sadly, I am in dire need of such clue dots. Being young I have a hard time explaining what I know to bored human resources blokes. Hell, imho they're only useful to make your resume more attractive to the HR blokes. [21:06:49] <cmihai> I've seen too many certified, trained and experienced individuals who simply lacked basic troubleshooting skills. [21:10:11] *** delewis has joined #opensolaris [21:10:55] *** loky has quit IRC [21:11:38] *** nostoi has quit IRC [21:11:57] *** cga has quit IRC [21:12:29] *** cga has joined #opensolaris [21:13:26] *** sahafeez has joined #opensolaris [21:13:45] *** cga has quit IRC [21:15:19] *** mega has quit IRC [21:18:14] <holcomb> damn hr [21:18:50] <axisys> anyone setup transparent proxy with java web proxy? looking for an example [21:19:04] <axisys> transparent web proxy that is [21:19:20] <jteo> yes damm HR. [21:20:12] *** yippi has joined #opensolaris [21:21:13] *** locy has joined #opensolaris [21:25:32] *** charlesnw has quit IRC [21:26:35] *** charlesnw has joined #opensolaris [21:27:43] *** Wez has joined #opensolaris [21:28:31] <Wez> hey folks, can anyone describe in more detail the actual issue in this bug: 6268715 [21:29:02] <Wez> I'm seeing port_getn() always fail with EINVAL on solaris x86, and the same code works fine on opteron and sparc [21:30:04] <flyingparchment> port_getn is broken [21:30:10] <flyingparchment> are you calling it with a timeout of 0? [21:30:25] <Wez> 0 and 0.9999something [21:30:38] <Wez> but also see the same issue when its 1.999something [21:30:49] <Wez> seems to be independent of the timeout [21:31:43] <flyingparchment> hmm, not seen that one [21:31:48] <flyingparchment> can you show trivial testcase? [21:32:35] <Wez> haven't tried yet; I was hoping that the test case from that bug number would repro it, but the web page doesn't make that available [21:32:44] <Wez> http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6268715 [21:33:12] <Wez> if I set the timeout to NULL, then the process blocks until I connect to it, and then fails with EINVAL [21:34:31] <Wez> If I run the process under truss, it seems to work [21:34:56] *** cypromis has quit IRC [21:36:40] *** Fullmoon has quit IRC [21:38:01] *** yarihm has quit IRC [21:41:22] *** aka_druid has quit IRC [21:43:32] *** aka_druid has joined #opensolaris [21:45:46] *** phs2 has quit IRC [21:48:26] <flyingparchment> hmm, apparently the N1 management server can _only_ run N1 [21:48:32] <flyingparchment> that's kind of lame [21:49:05] *** Fish has quit IRC [21:49:21] *** Fish has joined #opensolaris [21:51:47] <sommerfeld> s/kind of// [21:52:07] <flyingparchment> i wonder if it's really a requirement or just a good practice [21:56:05] <flyingparchment> how can i see what program is using a particular port? [21:57:50] *** mikefut has joined #opensolaris [21:59:46] *** Shiv_1 has joined #opensolaris [22:03:19] *** stratism has joined #opensolaris [22:03:28] <flyingparchment> hm, lsof works [22:04:18] *** noobuntu has joined #opensolaris [22:07:38] <jteo> oh that N1 thing. ugh. :( [22:07:47] <flyingparchment> jteo: not good? [22:07:55] <cmihai> flyingparchment, http://unixsadm.blogspot.com/2007/09/matching-tcp-port-to-process-pid.html Matching a TCP port to a process PID - the reverse of that :-) [22:08:05] <jteo> flyingparchment: it never seemed very well thought out to me. [22:10:45] *** phs2 has joined #opensolaris [22:10:52] *** cmihai has quit IRC [22:11:18] *** wms has quit IRC [22:11:19] *** cmihai has joined #OpenSolaris [22:18:23] *** phs2 has quit IRC [22:19:27] <Shiv_1> Anyone aware of the release date of Indiana? [22:21:38] *** jonkri has joined #opensolaris [22:21:57] <Shiv_1> It is strange that there is so much of push to call it "THE distro" when the odds are there to see. [22:23:06] <wesolows> Yes, it is a bit mysterious. [22:23:14] <cmihai> Next Friday 13, if there is a full moon. [22:23:14] *** tinman2k has quit IRC [22:25:11] *** phs2 has joined #opensolaris [22:26:56] *** phs2 has quit IRC [22:27:05] *** phs2 has joined #opensolaris [22:27:24] *** hsn_ has joined #opensolaris [22:27:32] *** jmcp has quit IRC [22:28:13] *** jonkri has quit IRC [22:28:24] *** wnorrix has joined #opensolaris [22:29:42] *** phs2 has quit IRC [22:40:27] <g4lt-mordant> next friday 13 is next year.... [22:44:55] <Hydroxide> hey, is it incorrect for a ZFS pool to be in status ONLINE but resilvering at the same time? and should a resilver of a 2.5-3 TB pool take minutes or days? it's taking days for me, with wildly varying (both up and down) estimates of percentage completed and time remaining [22:45:42] <sommerfeld> Hydroxide: it is not incorrect. And days is more typical. And the estimates are .. about as accurate as ufsdump's, for roughly the same reason. [22:45:43] <Teknix> a pool can be online while resilvering [22:45:55] <Hydroxide> sommerfeld: what is that reason? [22:46:09] <Teknix> and in my experience, it takes days [22:46:10] <Hydroxide> Teknix: ok, thanks [22:47:10] <sommerfeld> disk throughput primarily depends on number of seeks rather than number of blocks moved. while walking the tree of blocsk top-down, sometimes you hit areas that force a lot of seeks. [22:47:26] <sommerfeld> and sometimes you hit areas where there are lots of contiguous blocks [22:47:36] <flyingparchment> Email Alert Mail Address = emailuser1 at n1lab dot sfbay.sun.com [22:47:41] <flyingparchment> that's an interesting default config option [22:48:27] <sommerfeld> a scrub on a 2.12TB pool I have here takes about 36 hours. resilver takes about the same amount of time. [22:49:16] <sommerfeld> (2.12T allocated, 5.87T total size; 44 10K rpm 146GB fiberchannel disks ) [22:50:58] *** jcsmith has quit IRC [22:53:17] *** phs2 has joined #opensolaris [22:53:59] *** vmlemon has quit IRC [22:54:31] *** Shiv_1 has quit IRC [22:57:19] *** Fish has quit IRC [23:10:56] *** phs2_ has joined #opensolaris [23:11:18] *** phs2_ has quit IRC [23:11:58] *** phs2_ has joined #opensolaris [23:12:59] *** p__m has joined #opensolaris [23:15:11] *** storycrafte1 has joined #opensolaris [23:16:11] <TrogL> I'd like to post a thread about my "won't swap" problem, but I don't know which forum. It's a swap problem, but I have a sneaking supicious ZFS might have something to do with it (hogging or competing with ORACLE for RAM) [23:16:14] <TrogL> ideas? [23:16:29] <cmihai> the mailing lists [23:16:34] <TrogL> which one? [23:16:37] <cmihai> bugs? [23:16:41] <cmihai> discuss? [23:19:01] *** yippi_ has joined #opensolaris [23:22:57] *** bunker has quit IRC [23:22:59] <stevel> wwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwww [23:23:03] <stevel> uh [23:23:07] <stevel> sorry. ignore that [23:24:07] <e^ipi_> beagle on keyboard? [23:24:10] <l1s> http://mail.opensolaris.org/mailman/listinfo @ TrogL [23:24:16] *** e^ipi_ is now known as e^ipi [23:24:32] <stevel> e^ipi: nope, was me still missing sloppy focus and thinking i was in vim :) [23:24:49] <stevel> wondering why it wasn't, realising i was typing in my IRC client instead, and then accidentally hitting Return instead of Backspace [23:25:09] <stevel> iow, a perfect storm of retardedness and typos [23:25:22] * stevel often creates perfect storms of retardedness [23:27:21] *** medar has quit IRC [23:27:24] <cmihai> phs2_, do you wish to ask me something? [23:27:33] *** charlesnw has left #opensolaris [23:29:40] *** fdo has joined #opensolaris [23:30:59] *** noobuntu has quit IRC [23:31:21] *** fdo has left #opensolaris [23:31:26] *** timsf has quit IRC [23:32:48] *** noobuntu has joined #opensolaris [23:33:20] *** charlesnw has joined #opensolaris [23:34:06] <wesolows> stevel: you should save such perfect storms for opensolaris-discuss [23:34:48] *** yarihm has joined #OpenSolaris [23:34:52] <e^ipi> osol-discuss is already a storm of retardedness [23:34:54] *** yippi_ has quit IRC [23:34:57] <cmihai> Nah, indiana-discuss [23:35:09] <wesolows> we need autokick for that word [23:35:24] <e^ipi> it's like the mos eisley cantina... a greater hive of scum and villany doesn't exist in this quadrant of the galaxy [23:37:19] <sommerfeld> whoever said that clearly hadn't visited jabba's palace. [23:37:32] <cmihai> or my room [23:39:32] *** jcsmith has joined #opensolaris [23:40:49] *** mikefut has left #opensolaris [23:43:10] *** cmihai_ has joined #OpenSolaris [23:43:23] *** cmihai_ has quit IRC [23:43:27] *** mikefut has joined #opensolaris [23:44:29] *** cmihai_ has joined #OpenSolaris [23:44:44] *** cmihai_ has left #OpenSolaris [23:44:59] *** cmihai has quit IRC [23:45:15] *** cmihai has joined #OpenSolaris [23:46:27] *** locy has quit IRC [23:47:55] *** postwait has quit IRC [23:55:56] * holcomb pours a 40 for ben kenobi [23:57:12] *** mw46 has quit IRC [23:57:16] <l1s> uh, our politic is dump [23:58:18] <l1s> they just made the system to pay universitys that made inventions in the 1980's and are usefull anywhere on the planet [23:58:34] <l1s> all other universitys do not get any money [23:58:47] <TrogL> l1s:thanks [23:59:42] <l1s> for normal the state should pay money to any university that works on inventions used all over the planet...