July 12, 2012 
< | 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 | >
July 12, 2012 [00:07:56] *** knweiss has quit IRC[00:13:58] *** CnC has quit IRC[00:18:06] *** chorrell has quit IRC[00:54:18] *** deirdres has joined #smartos[00:55:06] *** PiotrSikora has quit IRC[00:57:11] *** PiotrSikora has joined #smartos[01:03:16] <deirdres> jeez, I go away for a few hours and come back to majol trolling. Demerits all round![01:34:49] <amublance_wookie> What are the SF office hours again?[01:40:04] <wesolows> it's not SF, it's virtual. The hours are in the topic, 1700-1800 UTC on Thursdays[01:41:43] <rmustacc> Which is 10-11am if TZ=US/Pacific.[01:47:09] *** darjeeling has quit IRC[02:29:30] *** hij1nx has joined #smartos[02:38:21] *** darjeeling has joined #smartos[02:39:09] *** hij1nx has quit IRC[02:47:36] *** enmand has joined #smartos[03:01:15] *** dap has quit IRC[03:10:58] *** CarlosC has joined #smartos[03:21:32] <amublance_wookie> Ok - makes sense.[03:22:08] *** CarlosC has quit IRC[03:22:10] <amublance_wookie> Thanks![03:23:20] *** CarlosC has joined #smartos[03:29:55] *** deirdres has quit IRC[03:30:27] *** CarlosC has quit IRC[03:38:02] *** jdavis has quit IRC[04:40:16] *** enmand has quit IRC[04:52:06] *** darjeeling has quit IRC[04:54:22] *** denizr has quit IRC[05:08:04] *** jamesd_ has quit IRC[05:09:22] *** jamesd_ has joined #smartos[06:42:45] *** darjeeling has joined #smartos[06:48:05] *** jamesd_ has quit IRC[06:48:59] *** jamesd_ has joined #smartos[07:37:24] *** jamesd_ has quit IRC[07:37:54] *** jamesd__ has joined #smartos[07:40:49] <MerlinDMC> hello[07:41:19] <rmustacc> Hello[07:41:19] <MerlinDMC> some OS/kernel/dtrace guy around ... or someone who builds smartos releases for joyent?[07:41:29] *** deirdres has joined #smartos[07:41:32] <rmustacc> Uh oh. Maybe I shouldn't have said anything.[07:41:46] <rmustacc> MerlinDMC: I can and try and help a little bit.[07:41:56] <richlowe> ditto, though I'm not smartos-ily smart.[07:42:08] <MerlinDMC> ;) ... just a little "bug" I saw on my builds and nahamu could verify that[07:42:43] *** Licenser has quit IRC[07:43:14] <MerlinDMC> [root@00-25-90-74-3b-ad ~]# intrstat[07:43:15] <MerlinDMC> intrstat: invalid program: "/usr/lib/dtrace/tcp.d", line 105: syntax error near "tcph_t"[07:43:21] <rmustacc> Hmm[07:43:29] <rmustacc> Try running that with -x nolibs[07:43:42] <rmustacc> And see if that solves the problem.[07:43:45] <MerlinDMC> this happens on custom builds from master branch sice about 5. of july - on official releases it'S all fine[07:44:06] <MerlinDMC> [root@00-25-90-74-3b-ad ~]# intrstat -x nolibs[07:44:06] <MerlinDMC> intrstat: invalid program: "/usr/lib/dtrace/tcp.d", line 105: syntax error near "tcph_t"[07:44:28] <rmustacc> Oh, sorry, yeah, intrstat probably doesn't take that as an argument, that's a dtrace argument.[07:44:44] <MerlinDMC> maybe that's a problem with the build zone seup that differs somehow ... or it's some commit that broke smth[07:44:59] <MerlinDMC> oh .. k ... let me change that[07:45:28] <rmustacc> Did you update your system but not recreate the fake-subset?[07:45:50] <MerlinDMC> i destroyed the zone ... and created a new one[07:46:00] <rmustacc> Okay, so that rules out that.[07:46:03] <MerlinDMC> so everything should be up to date[07:46:20] <rmustacc> This is happening in the GZ of a freshly booted system that you made right?[07:46:21] <richlowe> is there more to that error message?[07:46:39] <MerlinDMC> that's everything that gets to the console[07:47:10] <MerlinDMC> that's in the GZ of a freshly booted system ... on real hardware and inside vmware/parallels[07:47:34] <richlowe> rmustacc: I have suspicions[07:47:57] <MerlinDMC> should i upload the platform tgz or the complete iso?[07:48:15] <rmustacc> richlowe: I'll take suspicions.[07:48:25] <rmustacc> That might trigger something I've tried to block out of my memory.[07:48:38] <richlowe> MerlinDMC: can you run mdb -k, and ::print -T tcph_t[07:48:51] <richlowe> which should work, showing us the type exists[07:48:56] <richlowe> if it does, run ::objects -v, and pastebin the result[07:49:16] <richlowe> rmustacc: I have this vague memory that you guys pulled a change of mine that I didn't expect, and can do this to you if you're not careful[07:49:39] <richlowe> rmustacc: if it's not that, I'm out of ideas though, CTF _mismatches_ tend to give nonsensical type errors, not syntax errors.[07:49:56] <MerlinDMC> https://gist.github.com/511b52a90ca2b3e7ef19[07:50:15] <richlowe> Yup, it's probably the change of mine[07:50:18] <richlowe> do a clean illumos build.[07:50:30] <richlowe> smartos defaults to incrementals, I don't know how to force it[07:50:32] <richlowe> rmustacc?[07:50:45] <rmustacc> Let me remember.[07:50:51] <richlowe> rmustacc: fwiw, the change you guys pulled I'm in the process of integrating properly, which should save you this.[07:50:55] <rmustacc> I'd just manually do a clobber and remove the proto area probably.[07:51:05] <MerlinDMC> richlowe, i was really sure to have killed everything ... i did two builds in that zone both with this result[07:51:31] <MerlinDMC> but i'll rebuild everything and see what goes[07:52:26] <richlowe> MerlinDMC: what I'm using to judge it is that the CTF label of your objects is changing[07:52:46] <richlowe> unix and genunix from july 5th, everything else from june 29th[07:52:58] <MerlinDMC> i saw that ... yep[07:53:03] <richlowe> that's pretty much directly pessimal to what I think has happened[07:53:27] <MerlinDMC> i did remove the complete smartos-live directory and now i'll fetch everything fresh and rebuild[07:54:09] <MerlinDMC> just wanted to make sure it's not a commit before a release is build officially ;)[07:58:14] <MerlinDMC> k ... build running ... breakfast for me[07:58:21] <richlowe> MerlinDMC: basically, don't allow the VERSION string to change between builds, for right now[07:58:29] <richlowe> the joyent_* bit[07:59:26] <MerlinDMC> richlowe, i'll check that again and if it's broken with all the same versions i'll whine and poke you guys again ;)[07:59:59] <richlowe> Well, I can't say it won't be. But I can say the versions that you have will cause this to happen.[08:00:30] <richlowe> and when I get the complete set of fixes into illumos, I'll go bug e^ipi to pull them to smartos[08:00:43] <richlowe> Though given this stuff escaped early, perhaps the other bit can too[08:01:04] <richlowe> e^ipi: Hey, you have an old version of my dtrace ctf parent checking in smartos, but you don't have the "don't merge genunix" changes that need to go with it.[08:01:22] <MerlinDMC> yeap ... we'll just X-out one thing and see if it's ok after that ... it's not much work for me to report that smth is not working[08:01:36] <richlowe> e^ipi: you should either ditch the parent checking stuff, and get it back via illumos, or take the "don't merge" stuff.[08:01:38] <MerlinDMC> just bad that i'm not that much into the system to get more info and fix it myself[08:03:59] *** deirdres has quit IRC[08:29:49] *** bens1 has joined #smartos[08:47:53] *** darjeeling has quit IRC[08:48:26] *** darjeeling has joined #smartos[08:56:30] *** Licenser has joined #smartos[09:03:21] <Licenser> morning everyone :)[09:03:32] <MerlinDMC> morning[09:03:42] <Licenser> how is life for all of you?[09:04:01] <MerlinDMC> <- broke the build[09:04:28] <wesolows> excellent; node-contract is now somewhat functional which means I can go to sleep[09:04:57] <Licenser> MerlinDMC I'm not even commenting on that ;)[09:05:02] <Licenser> night wesolows :)[09:22:53] *** jamesd__ has quit IRC[09:23:29] *** jamesd__ has joined #smartos[09:35:37] *** jamesd__ has quit IRC[11:00:09] *** d[^_^]b has joined #smartos[11:05:28] *** darjeeling has quit IRC[11:47:34] <MerlinDMC> rmustacc + richlowe, after rebuilding everything from scratch it seems to work. so i need to clean everything up before rebuilding.[11:48:24] <Licenser> dun dun dun[11:49:04] <MerlinDMC> and just for the log: building smrtos using a zone created from sdc:sdc:base:1.7.0 fails ... maybe useful for nahamu ;)[11:49:31] <Licenser> there was a discussion about that this morning I tried to convice the guy to clean my flat instead of doing the build - din't worked, I really wonder why[11:51:01] <aszeszo> hi all, is SDC product using gcc-built illumos bits?[12:24:49] *** sjorge has quit IRC[12:28:30] *** sjorge has joined #smartos[12:54:52] *** Worsoe has joined #smartos[13:24:11] *** enmand has joined #smartos[13:28:37] *** darjeeling has joined #smartos[13:32:41] *** enmand has quit IRC[13:36:08] *** enmand has joined #smartos[13:39:53] *** darjeeling has quit IRC[13:41:50] *** bens1 has quit IRC[13:52:49] *** Licenser has quit IRC[14:02:21] *** sjorge has quit IRC[14:28:16] *** ryancnelson has joined #smartos[14:34:13] *** ira has joined #smartos[14:48:34] *** hij1nx has joined #smartos[14:53:51] <nahamu> MerlinDMC: thanks. :)[14:58:26] <MerlinDMC> nahamu, you need to clean the proto directory before rebuilding smartos to fix that dtrace error[14:58:35] <JT-EC> Is init 5 a normal way to shut a smartos host down?[14:58:54] <MerlinDMC> JT-EC, i'm lazy and do a "halt" ... or push the button[14:58:57] <nahamu> gmake clean does clean out the proto dir... (which I always do... I've been doing "gmake clean update world live")[14:59:36] <nahamu> but this time I just cd'd into projects/illumos and ran "git clean -fdx"[14:59:45] <Alasdairrr> Any JoyentCloud peeps around?[14:59:47] <nahamu> we'll see if that clears things up.[14:59:49] <MerlinDMC> nahamu, maybe there are some other directories not cleared and illumos object files remain[15:00:38] <nahamu> I'm guessing based on richlowe's comment that it's the illumos dir... we'll see how this goes.[15:02:49] *** hij1nx has quit IRC[15:18:16] <jperkin> Alasdairrr: it's highly unlikely that I'll have necessary privs to help, but what's up? :)[15:19:15] <ryancnelson> hey, jperkin … i'm over in #joyent[15:19:39] <jperkin> ah ok[15:30:38] <nahamu> oops. after that git clean -fdx... need to re-run configure.[15:31:37] *** neophenix has joined #smartos[15:38:02] *** ira has quit IRC[15:40:13] *** CarlosC has joined #smartos[15:47:01] *** ryancnelson has quit IRC[15:48:22] *** jamesd__ has joined #smartos[15:49:39] *** jamesd__ has quit IRC[15:50:16] *** jamesd__ has joined #smartos[15:54:04] *** jamesd__ has quit IRC[15:54:27] *** jamesd_ has joined #smartos[15:55:57] *** Worsoe has quit IRC[16:12:31] *** jamesd_ has quit IRC[16:13:23] *** jamesd_ has joined #smartos[16:13:36] *** jperkin has quit IRC[16:16:41] *** akole has quit IRC[16:22:44] *** ira has joined #smartos[16:34:48] *** jperkin has joined #smartos[16:35:57] <rmustacc> MerlinDMC: Right, well, we never mentioned that it did work.[16:36:07] <rmustacc> aszeszo: SDC 6.5 is not built with gcc.[16:36:31] <aszeszo> rmustacc: thanks[16:36:41] <rmustacc> aszeszo: SmartOS and everything onwards is.[16:36:47] <aszeszo> is there an easy way to tell if illumos was built with studio or with gcc?[16:37:42] <rmustacc> Mmm, maybe. Are you seeing something where it makes a difference?[16:39:48] <aszeszo> we currently have few production smartos machines here, we built smartos images ourselves and used studio to build illumos bits[16:40:37] <aszeszo> were wondering if gcc built illumos was reliable enough and is being used by joyent/SDC[16:42:46] <rmustacc> We think gcc4.4.4 is reliable.[16:43:17] <rmustacc> And it is the future, given that you can't get a proper version of Studio without a support contract.[16:44:45] <jperkin> and we use gcc4.6/4.7 in pkgsrc[16:45:04] <rmustacc> Right, it's a slightly different issue with which gcc to use in the kernel.[16:45:41] <CarlosC> rmustacc how long as SmartOS been using gcc4.4.4? about a year right?[16:46:01] <rmustacc> Since SmartOS has been released, iirc.[16:46:23] <rmustacc> We've been using the illumos gcc 4.x branch long before it got integrated into illumos.[16:46:35] <rmustacc> And engineering uses it internally for everything.[16:49:08] *** kfr has joined #smartos[16:50:04] <CarlosC> yeah...it's been pretty solid for our uses...do you know if the next version of SDC will be using it?[16:51:24] <rmustacc> CarlosC: Oh, without a doubt.[16:51:40] <rmustacc> Everything you see in SmartOS is the same platform and build process (modulo some extra bits) for SDC.[16:52:48] <aszeszo> rmustacc: but you don't feel gcc444 build it is ready to use it in the public cloud/SDS?[16:53:27] <aszeszo> (right now)[16:55:51] <rmustacc> aszeszo: Well, it's a matter of risk. When we originally forked SDC 6.5 off of our development we weren't going to change the compiler. Since then we gained a lot of confidence with gcc 4.4, but I don't think there's much value in us backporting all the changes necessary for illumos to build/function properly to it.[16:56:48] <rmustacc> It's just not worth the engineering effort to backport and test everything for that kind of substantial change.[16:58:33] *** mamash has left #smartos[16:59:14] *** mamash has joined #smartos[16:59:16] <rmustacc> If I could magically get a gcc 4.4.4 build of SDC 6.5 I'd have no qualms or questions about deploying it.[16:59:46] <rmustacc> aszeszo: Does that answer your question?[17:01:03] *** Licenser has joined #smartos[17:01:36] <aszeszo> yes, it all makes sense now[17:01:39] <aszeszo> thanks rmustacc![17:03:00] *** jdavis has joined #smartos[17:05:14] <kfr> I'm going to be ordering 4 of http://www.supermicro.com/products/system/1u/1026/SYS-1026T-URF4_.cfm?UIO=Y[17:06:47] <kfr> once I get kvm/smartos running on it, how can I add it to the hcl?[17:07:16] <rmustacc> All the components should already be on the illumos hcl.[17:07:29] <kfr> great![17:07:45] <rmustacc> But you can check on http://illumos.org/hcl[17:08:04] <rmustacc> At least, the peripherals/chipsets, which are the more important pieces.[17:08:58] <kfr> we are a small shop[17:09:04] <kfr> so I'm only ordering 4[17:10:02] *** mamash has left #smartos[17:10:10] *** mamash has joined #smartos[17:13:55] *** mamash has left #smartos[17:14:01] *** mamash has joined #smartos[17:26:09] *** dap has joined #smartos[17:28:14] *** relling has joined #smartos[17:30:12] *** jamesd__ has joined #smartos[17:30:47] <Licenser> dun dun dun[17:30:48] *** jamesd_ has quit IRC[17:31:34] *** denizr has joined #smartos[17:41:25] *** hij1nx has joined #smartos[18:28:55] <jelmd> rmustacc: hcl still contains broadcom 5761 - thingy doesn't work on S11 and I doubt, that it works on illumos based OS, since the driver is even much older, than recent ones ...[18:30:54] *** deirdres has joined #smartos[18:31:00] <killfill> http://www.ebay.com/itm/1U-Server-X8DTU-F-2x-Intel-Xeon-L5640-6-Six-Core-96GB-RAM-4-Bay-CSE-815TQ-R700-/150854302043?pt=Motherboards&hash=item231f9df95b#ht_5735wt_1057 <-- this should work with SDc right?..[18:39:32] <rmustacc> jelmd: Then remove it from the IPS manifest.[18:41:08] <jelmd> no problem 4me, but if ppl looking on it and think it'll work, buy it and get upset ...[18:42:20] *** pijewski has joined #smartos[18:42:39] *** denizr1 has joined #smartos[18:42:48] *** denizr has quit IRC[18:43:39] <nahamu> jelmd: so in theory a webrev for illumos would be in order, no?[18:43:40] <rmustacc> jelmd: I'd be curious who added the PCI ID that says it worked.[18:44:10] <jelmd> the same one, who added 5761E ;-)[18:44:55] <jelmd> nahamu: have not that much time :([18:46:09] <nahamu> bug report?[18:46:38] *** nicksellen has joined #smartos[18:47:15] <nicksellen> I'm having trouble with vmadm stop on a zone, it just doesn't stop, zoneadm halt works as expected - http://paste.ec/?97a00c84525fb27c#B9NJGkEqU+av/izMi46I+UkXUAX+b9YB9PrX7zMOHR0=[18:51:43] <jelmd> nahamu: hmm, thought, I had it already hammered in, but can't find it. Maybe it was a not so redmine day and I forgot to do it again ...[18:55:00] <nicksellen> fault found - a dodgy activemq package inside the zone failing to stop[18:55:30] <nicksellen> it's a bit disconcerting seeing "Succesfully completed stop" though when it hasn't[18:56:24] *** jdavis has quit IRC[18:57:55] <rmustacc> killfill: Seems okay, you'll want to double check the storage controller.[18:58:33] <rmustacc> nicksellen: Can you file a bug on that behavior and include the buggy activemq configuration so we can try and reproduce it.[18:58:53] *** denizr1 has quit IRC[18:59:11] *** denizr has joined #smartos[18:59:28] <Alasdairrr> The problem will be SMF services with a -1 timeout on the stop method[19:00:13] <nicksellen> <exec_method name='stop' type='method' exec='/ec/var/activemq/bin/activemq-admin stop' timeout_seconds='60'/>[19:00:22] <nicksellen> it's got a timeout and a proper activemq stop method[19:00:49] <Alasdairrr> oh weird[19:00:50] <wesolows> good news, everyone! it's office hours[19:01:02] <Alasdairrr> \o/[19:01:21] <Alasdairrr> oh wow nicksellen, that's quite a bug[19:01:45] <Alasdairrr> smf ftw[19:01:49] <nicksellen> rmustacc: where do the bugs go?[19:02:13] <rmustacc> nicksellen: https://github.com/joyent/smartos-live/issues[19:04:15] <nahamu> e^ipi: I've signed my own CLA too, so definitely feel free to do stuff with my pull request.[19:05:31] <nahamu> rmustacc: looks like MerlinDMC put in a pull request for minor cleanup of a manpage... should be a no-brainer to pull in.[19:06:50] <rmustacc> #97?[19:07:00] <nahamu> yeah[19:07:45] <rmustacc> Sure, I can take care of that today.[19:08:59] <nahamu> plus it will let you close off issue #96[19:10:08] <nahamu> I was curious if #94 is likely to be fixed in the next build[19:12:55] <rmustacc> I don't know. Depends on where RobG is on it.[19:13:51] <nahamu> and where it is in his list of priorities, I'd imagine. I'd like IPv6 too, but clearly Rob can only give us so many lovely presents at a time...[19:14:31] <rmustacc> I'd assume it probably won't be in today's release.[19:14:35] *** hij1nx has quit IRC[19:15:46] <nahamu> fair enough.[19:27:20] <rmustacc> That simple pull request is handled.[19:27:31] <nahamu> indeed. thanks![19:29:12] *** hij1nx has joined #smartos[19:29:18] <nahamu> I'm sure MerlinDMC will be pleased as well.[19:33:12] <nahamu> rmustacc: your comment on my usb image hack about normal users sparks a question:[19:33:27] <nahamu> do you guys do builds as non-root users?[19:34:00] *** tafryn has joined #smartos[19:34:30] <rmustacc> People do, yes.[19:34:44] <rmustacc> But they do set up the appropriate pfexec issues.[19:35:47] *** Niamkik has joined #smartos[19:36:43] <nahamu> do you mean they specifically run "pfexec ./configure" (or whichever commands need elevated privs)?[19:37:19] <rmustacc> nahamu: Specifically commands that need elevated privileges currently use pfexec in the scripts.[19:37:25] <rmustacc> Things are not run that way.[19:38:04] <nahamu> ah, I think I see that now ("git grep pfexec" in smartos-live)[19:38:48] <rmustacc> Ideally, you want to drive the set that has that to zero.[19:39:34] <nahamu> right.[19:40:31] <nahamu> I'd imagine that the current method of generating/updating a USB image requires pfexec, though I guess e^ipi is the one who would know for sure.[19:40:42] <rmustacc> Probably, yes.[19:41:02] <nahamu> so it's probably defensible in this case. :)[19:45:53] <nicksellen> rmustacc: https://github.com/joyent/smartos-live/issues/99[19:46:51] <nicksellen> I'm not actually sure what the problem is (vmadm not doing whatever zoneadm does, smf not handling it better, the activemq manifest… etc)[19:47:08] *** gflarity has joined #smartos[19:52:29] *** jdavis has joined #smartos[19:53:40] *** Licenser has quit IRC[20:01:22] *** knweiss has joined #smartos[20:07:08] <nahamu> I guess office hours were pretty quiet today. Thanks rmustacc![20:07:52] <rmustacc> Sure, no problem.[20:14:36] *** jamesd__ has quit IRC[20:31:47] *** hij1nx has quit IRC[20:31:58] *** banjiewen|away is now known as banjiewen[20:34:20] *** hij1nx has joined #smartos[20:35:09] *** dis has quit IRC[20:36:15] *** kfr has quit IRC[20:37:13] *** dis has joined #smartos[20:42:07] *** dis has quit IRC[20:42:30] *** dis has joined #smartos[20:47:49] *** hij1nx has quit IRC[21:02:05] *** bradleymeck has joined #smartos[21:02:22] <bradleymeck> is there a good ipfilters tutorial somewhere?[21:05:47] <nahamu> MerlinDMC (and whoever else): cleaning out the illumos directory before a rebuild resulted in a build where intrstat worked again.[21:06:36] <nahamu> so clearly a from-scratch build would be fine. what's the simple way to tell the illumos build system to clean up?[21:07:26] <richlowe> remove the 'i' from NIGHTLY_OPTIONS in illumos.sh[21:07:32] <richlowe> or bldenv in and 'dmake clobber' at the top level[21:07:49] <richlowe> fwiw, the fix for this problem is going to land in illumos the moment I have more time, and I told both Robert and John what their options were for pulling it early.[21:08:27] <richlowe> nahamu: as I said to MerlinDMC, the important thing is that VERSION does not change between incremental builds[21:08:49] <richlowe> nahamu: I'm not sure how smartos arranges the value of VERSION.[21:09:16] *** denizr has quit IRC[21:09:28] *** denizr has joined #smartos[21:10:30] *** denizr1 has joined #smartos[21:12:29] <nahamu> so the NIGHTLY OPTIONS in smartos-live/configure could have the 'i' removed if that was deemed a good idea.[21:12:51] *** denizr has quit IRC[21:13:45] *** denizr1 has quit IRC[21:14:15] <richlowe> it'd make all your builds slower[21:14:46] <richlowe> it's going to be better for John to either pull the proper fix when I push it, pull it ahead of time from the diff, or backout the half of it he already has.[21:15:00] <nahamu> got it.[21:43:06] *** michaeldexter has joined #smartos[21:44:53] *** enmand has quit IRC[21:44:55] <MerlinDMC> helas[21:45:44] <michaeldexter> Is ISO-based installation documented anywhere? The installer tells me "could not open zones" and reboots.[21:46:06] <MerlinDMC> michaeldexter, did you specify any disks?[21:46:17] <MerlinDMC> or just hit enter at disk selection? :)[21:46:17] *** nicksellen has quit IRC[21:46:31] <michaeldexter> One disk, just hit enter.[21:47:08] <MerlinDMC> specify that disk ... there is no default so just hitting enter will not create the zpool -> installer barks and a reboot will give you the installer again[21:47:52] <michaeldexter> No commands were suggested, just what might be a disk name[21:48:10] <MerlinDMC> there should be a list of disks on that screen[21:48:27] <MerlinDMC> c0...[21:48:30] <michaeldexter> disks yes, commands not so sure (should be back up shortly)[21:48:56] <rmustacc> It's a bug that it accepted empty input that should be rectified at some point.[21:49:14] <michaeldexter> "Valid choices are c1t0d0" (blinking prompt _ )[21:49:22] <MerlinDMC> rmustacc, there was a fix :P[21:49:27] <rmustacc> I know.[21:49:31] <MerlinDMC> michaeldexter, enter that name ...[21:49:36] <rmustacc> It's on e^ipi's plate to pull it in, iirc.[21:49:46] <MerlinDMC> no extra commands[21:50:24] <michaeldexter> Better...[21:52:09] <MerlinDMC> rmustacc, the comment reg building smartos on base:1.7.0 was mostly for nahamu and others doing their builds had a free box to try it[21:52:33] <MerlinDMC> but had no time to look further into it[21:53:06] <michaeldexter> MerlinDMC: seems to start but soon reboots. no os found… where can I find the appropriate documentation?[21:53:22] <MerlinDMC> michaeldexter, boot from the cd or usb stick[21:53:25] <richlowe> there is no OS placed on the disk[21:53:27] <richlowe> leave the CD in[21:53:30] <MerlinDMC> the os does not get installed on disk[21:53:52] <michaeldexter> interesting[21:54:38] <MerlinDMC> michaeldexter, http://www.youtube.com/watch?v=ieGWbo94geE (and yes i like that video ^^)[21:58:31] *** Niamkik has quit IRC[22:04:41] <michaeldexter> MerlinDMC: I am booted. Will I need the CD in for every boot?[22:04:52] <MerlinDMC> michaeldexter, yep[22:05:09] <MerlinDMC> because you need the kernel ... and that is on the cd[22:05:23] <michaeldexter> cp … /boot? :)[22:05:48] <richlowe> You could put the kernel and archive on the disk, but that's sorta not the point.[22:05:52] <MerlinDMC> hmm ... there is some way to do that ... but you really should not[22:06:02] <michaeldexter> Interesting[22:06:15] <MerlinDMC> just get a $5 usbkey with 1 gig and use that ;)[22:06:18] <michaeldexter> Let's see what the next 10 minutes of the video has to say....[22:06:21] <michaeldexter> Let's see what the next 10 minutes of the video has to say….[22:06:50] <michaeldexter> That is a different matter… and a smart one.[22:07:06] <michaeldexter> Not clear in your documentation. Anyhow. Thanks.[22:08:41] *** bradleymeck_ has joined #smartos[22:08:42] <MerlinDMC> is that really written down nowhere?[22:10:36] <MerlinDMC> michaeldexter, here it is documented somehow: http://wiki.smartos.org/display/DOC/Getting+Started+with+SmartOS[22:10:37] *** bradleymeck has quit IRC[22:10:38] *** bradleymeck_ is now known as bradleymeck[22:16:37] <michaeldexter> Thanks. Old habits die hard![22:17:34] *** bradleymeck has left #smartos[22:26:12] <MerlinDMC> yep ... but you'll love upgrading the os ;)[22:28:11] *** sjorge has joined #smartos[22:36:39] *** hij1nx has joined #smartos[22:37:04] * MerlinDMC out to get some sleep[22:38:14] <nahamu> g'night MerlinDMC[22:41:35] *** hij1nx has quit IRC[22:51:10] *** knweiss has quit IRC[23:14:41] *** gflarity has left #smartos[23:16:27] *** jamesd has joined #smartos[23:16:38] *** hij1nx has joined #smartos[23:27:35] *** jmeredith has joined #smartos[23:28:52] *** CarlosC has quit IRC[23:35:42] <jmeredith> anybody who knows about smartos zones and disk quotas?[23:36:28] <jmeredith> I filled up a disk, and it won't let me remove anything[23:36:29] <jmeredith> [root@mymachine /var/db/riak/leveldb/0]# rm 000003.log[23:36:29] <jmeredith> rm: cannot remove `000003.log': Disc quota exceeded[23:36:42] <jmeredith> help![23:37:06] <rmustacc> The quickest way out of that is to do something like truncate on the file.[23:37:17] <rmustacc> Now, to understand why this happened, there are a few questions.[23:37:23] <rmustacc> Do you have any snapshots?[23:37:30] <jmeredith> not that I'm aware of[23:37:42] <jmeredith> (it's a hosted machine on joyent)[23:37:47] <rmustacc> Okay.[23:37:57] <jmeredith> let me find some larger files and truncate them[23:38:35] <jmeredith> ok, now I can remove again[23:38:37] <rmustacc> The problem is that ZFS is a copy on write filesystem and so to remove files it takes disk space to remove before it can then remove the other space.[23:39:00] <jmeredith> makes sense now you remind me[23:39:03] <jmeredith> thx for the help[23:39:32] <rmustacc> Sure, no problem.[23:40:11] <rmustacc> If you have snapshots, then removing files doesn't necessairly end up eliminating all the space you're using if it's in that snapshot.[23:40:28] <jmeredith> that also makes sense[23:40:39] <jmeredith> I'd just never had a remove operation when I was overquota fail on me[23:40:57] <jmeredith> didn't think of truncating[23:41:17] *** Licenser has joined #smartos[23:41:34] <rmustacc> You're probably not on a copy-on-write filesystem previously, but yeah, it's pretty bad when that happens.[23:42:11] <jmeredith> alright, I'm off to the races - thanks again[23:54:14] *** banjiewen is now known as banjiewen|away