April 29, 2009 
< | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | >
April 29, 2009 NOTICE: This channel is no longer actively logged.
[00:00:37] <zooko> Hm. All five of the grub options result in "Inconsistent filesystem structure".[00:01:22] <zooko> crap[00:03:39] <zooko> The internet suggests that this might be a bug in the ZFS implementation in grub.[00:03:41] *** pino42 has joined ##nexenta[00:06:03] *** randybias has joined ##nexenta[00:09:06] <rootard> zooko: that's probably the boot_archive that has a problem[00:09:37] <rootard> if you can boot via a recent install cd then I can give you a few hints on recreating the boot_archive[00:10:36] <rootard> Since you upgraded to NCP1 I would recommend the latest NCP1 boot cd (http://www.nexenta.org/releases/nexenta-core-platform_1.0.1-b85_x86.iso.zip)[00:12:20] <zooko> Thanks![00:12:30] <zooko> I'll ask Brian to burn an NCP1 iso disc.[00:14:06] <zooko> Brian is off to ask if he can borrow a CDRW[00:14:13] <zooko> Or "have", I guess.[00:14:42] <zooko> Is this the workaround? http://sigtar.com/2009/02/02/zfs-error-16-inconsistent-filesystem-structure/[00:22:20] <zooko> I hope you haven't disappeared by the time Brian gets this ISO booted.. :-)[00:22:26] <rootard> no, that is the "zOMG, it's full of stars" approach. Unless you are willing to copy everything off and reinstall it would be less pain (IMHO) to just import the pool and recreate the boot archive[00:22:28] <zooko> But if so I'll try following the recipe.[00:22:38] <zooko> So, I'm pretty lost.[00:22:45] <rootard> zpool import syspool[00:22:48] <zooko> Copying everything off the whole system would be a pain.[00:23:06] <rootard> zfs list | grep rootfs[00:23:28] <zooko> Okay, he doesn't have it booted yet, but what should he do after that?[00:23:34] <zooko> After the zfs list | grep rootfs[00:23:34] <rootard> take the last one (maybe syspool/rootfs-nmu-000) and mount it somewhere:[00:23:49] <zooko> By the way, how did I get into this state?[00:23:51] <rootard> mount -F zfs syspool/rootfs-nmu-000 /mnt[00:24:04] <zooko> It booted successfully, but since then the ZFS fs changed state to such a state that grub can't boot it due to this grub bug?[00:24:08] <zooko> Even though it isn't actually broken?[00:24:12] <rootard> then update the boot archive: bootadm update-archive -vR /mnt[00:24:14] <zooko> I mean the ZFS fs isn't actually inconsistent?[00:24:28] <rootard> I don't think the problem is the zfs filesystem[00:24:32] <zooko> Thanks a lot for the help![00:24:38] <rootard> I think the problem is the boot archive... which is a filesystem on an image[00:24:43] *** koan has quit IRC[00:24:50] *** koan has joined ##nexenta[00:24:58] <zooko> Do you think this is the bug: http://bugs.opensolaris.org/view_bug.do?bug_id=6774616[00:25:13] <rootard> you can ask your friend to scrub the pool from the install cd if you want... just to check the thing over[00:25:45] <zooko> Okay, how do I do that?[00:25:55] <zooko> I'll do that before the fix-the-boot-archive recipe.[00:25:57] <rootard> zpool scrub syspool[00:26:15] <rootard> then watch the output of "zpool status" to wait for it to complete[00:26:38] <rootard> if your pool is nearly full then I think you may have a chance at the linked bug...[00:26:58] <rootard> if the boot-archive does not solve the problem[00:27:14] <rootard> oh... one other thing to do while the image is mounted: update grub[00:28:10] <zooko> How?[00:28:18] <zooko> It has been nearly full in the past.[00:28:19] <rootard> it will looks like: cd /boot/grub;installgrub stage1 stage2 <disk>[00:28:27] <rootard> *look[00:28:40] <rootard> where disk can be found in the zpool status output[00:29:17] <zooko> How long will scrub take?[00:29:27] <rootard> if it isn't full now then you have effectively done the work-around in the bottom of the bug (remove unnwanted datasets, re-copy bootarchive)[00:29:34] <rootard> depends on how much data you have[00:29:37] <zooko> Ok.[00:29:40] <zooko> It is not full now.[00:29:49] <rootard> it will read it all of the disk and check it against checksums[00:29:49] <zooko> 'round 90% or 95% now.[00:29:55] <zooko> Yeah, that'll take a while.[00:30:00] <rootard> That's pretty full ...[00:30:07] <zooko> Hm.[00:30:19] <zooko> Can't really delete anything from there right now, with it off-line and all.[00:30:31] <zooko> Should I ask Brian to find something to delete while he's booted from the ISO?[00:30:33] <rootard> chroot /mnt;apt-get clean[00:30:44] <zooko> I'll suggest that, depending on what the df is.[00:30:57] <rootard> maybe log files if there are a lot hanging around[00:31:32] * rootard wishes dedup were available.[00:31:50] <zooko> Yeah![00:36:13] <zooko> Okay! He's powering it up with the CD-ROM in...[00:36:34] <zooko> poor man's dedup: http://pypi.python.org/pypi/dupfilefind[00:36:40] <zooko> Hm, it didn't boot from the CD-ROM.[00:37:00] <zooko> He's looking at the BIOS settings.[00:38:40] <rootard> zfs dedup will work on a block level so even if parts of files are exactly the same they will be dedup'd[00:38:46] <zooko> Yeah, I know.[00:38:52] <zooko> *very* poor man's dedupe. ;-)[00:39:32] <zooko> whoops, my device is not a DVD reader, just a CD reader. :-P[00:39:38] <rootard> :) there is actually some zdb usage that allows you to list blocks and their checksums in order to get a real handle on which dedup would do for your filesystem[00:40:11] <rootard> I don't remember it but there is a zfs-discuss thread on it[00:41:24] <zooko> Okay, he's going to come back around about 1.5 hours from now with a CDRW. :-)[00:42:14] * rootard nods[00:42:31] <rootard> just keep the above points handy and they should get you through if I am not around[00:42:36] <zooko> Thanks a lot!![00:42:55] <zooko> By the way, one of my mail goals for my Tahoe-LAFS project this summer is to get included into operating systems.[00:42:59] <rootard> No problem :)[00:43:07] <zooko> Debian, Fedora, Ubuntu, Gentoo, and Nexenta. :-)[00:43:28] <zooko> Also, have you seen "Laptop Versus Axe"?"[00:43:35] <rootard> that would be cool... so you are talking about a in-kernel client and the server is still userland?[00:43:36] <zooko> It's a movie I made, on stage at CodeCon last week.[00:43:41] <zooko> All userland.[00:43:52] <rootard> so some kind of FUSE implementation?[00:43:53] <zooko> You communicate with it through HTTP (REST), command-line, etc.[00:44:00] <zooko> There *is* a FUSE implementation, but it doesn't work yet.[00:44:00] *** wurlitzer has joined ##nexenta[00:44:15] <zooko> So therefore it isn't *really* a "filesystem" in the sense of you being able to point your normal apps at it.[00:44:20] <zooko> It's more of a "cloud storage thingie".[00:44:23] <rootard> So you are looking to get packages accepted into these distributions[00:44:37] <zooko> Laptop Versus Axe: http://testgrid.allmydata.org:3567/uri/URI:DIR2-RO:j74uhg25nwdpjpacl6rkat2yhm:kav7ijeft5h7r7rxdp5bgtlt3viv32yabqajkrdykozia5544jqa/wiki.html[00:44:43] *** koan has quit IRC[00:44:50] <zooko> Yes, just a Python userland application, packaged up in the distributions. Exactly.[00:44:50] *** koan has joined ##nexenta[00:44:51] <rootard> you really need tinyurl for these things :)[00:44:57] <zooko> Duly noted.[00:45:38] <rootard> Wow, the laptop won for a moment it looks like[00:45:50] <zooko> http://allmydata.org/trac/tahoe/ticket/217#comment:42[00:45:52] <zooko> :-)[00:48:04] *** wurlitzer has quit IRC[00:48:24] <rootard> lol, I'm immortal now[00:48:28] <zooko> :-)[00:48:42] <zooko> I've already got some folks offering to package it for Debian and Ubuntu.[00:48:48] <zooko> I think someone offered to do so for Gentoo as well.[00:49:00] <zooko> There's a problem with C++ linkage with the Crypto++ library, but other than that I think we're almost ready.[00:49:09] <rootard> If they do a good job with the Debian packages then we should be able to just build and import them as well[00:49:20] <zooko> Yeah![00:49:53] <rootard> Things flow pretty easily (for the most part) from Ubuntu (and sometimes Debian) into Nexenta.[00:54:43] *** koan has quit IRC[00:54:50] *** koan has joined ##nexenta[00:54:55] *** randybias is now known as randybias|away[00:55:19] <zooko> So, I'm busy looking for a job, but I'm also working on Tahoe in my 'spare time', and I have the informal goal for myself of getting Tahoe into Debian by June.[00:59:24] <rootard> awesome. I look forward to being able to apt-get install it.[00:59:43] *** koan has quit IRC[00:59:51] *** koan has joined ##nexenta[00:59:56] <zooko> :-)[01:00:23] <zooko> Actually it is already available in an apt repo, if you want to try it out: http://allmydata.org/trac/tahoe/wiki/DownloadDebianPackages[01:08:40] <rootard> zooko: do you have an incoming ftp space for these repositories?[01:11:01] <zooko> Um, I'm not sure. Why?[01:11:13] <zooko> It's time for me to put dinner on. Back in a bit![01:11:51] <rootard> Well, I can set the autobuilder up to try and produce Nexenta .debs from the published sources. As new sources are available they will be built[01:12:16] <rootard> the only thing I need is an incoming ftp location for the debs to be uploaded into[01:12:53] <rootard> it will even capture build-logs for you so you can easily see what went wrong[01:16:53] *** koan has quit IRC[01:17:01] *** koan has joined ##nexenta[01:21:09] <zooko> Sweet![01:22:18] <zooko> I'll ask Brian Warner, who set up the .deb-building infrastructure, where you can upload yours. :-)[01:36:31] *** zooko has quit IRC[01:43:14] *** pino42 has quit IRC[01:45:58] *** randybias|away is now known as randybias[01:52:52] *** ohol1ks has joined ##nexenta[01:52:56] *** taltamir has joined ##nexenta[01:53:30] *** alfism has quit IRC[01:53:30] *** NCommander has quit IRC[01:53:31] *** asdf has quit IRC[01:53:31] *** nightwish has quit IRC[01:53:31] *** chol has quit IRC[01:53:48] *** alfism has joined ##nexenta[01:53:48] *** asdf has joined ##nexenta[01:53:48] *** NCommander has joined ##nexenta[01:53:48] *** chol has joined ##nexenta[01:53:48] *** nightwish has joined ##nexenta[01:54:28] *** NCommander has quit IRC[01:56:37] *** NCommander has joined ##nexenta[02:06:40] *** alfism has quit IRC[02:08:33] *** chol has quit IRC[02:15:06] *** zooko has joined ##nexenta[02:19:48] *** olsner has quit IRC[02:23:00] *** Capscrew has joined ##nexenta[02:23:44] *** Capscrew has left ##nexenta[02:29:27] *** GHReyes has joined ##nexenta[02:35:58] *** randybias is now known as randybias|away[03:07:08] *** randybias|away is now known as randybias[03:11:52] *** GHReyes has quit IRC[03:26:12] <zooko> Okay, now Brian is trying to fix my server.[03:26:39] <zooko> I told him to do "zpool import syspool"[03:26:52] <zooko> and it says "Cannot import syspool -- pool may be in use by other systems. Use -x to import anyway."[03:27:02] <zooko> rootard: are you around?[03:29:34] <zooko> anyone else know how to rebuild my boot archive?[03:37:08] <zooko> Okay, -x worked...[03:45:40] <zooko> Hm, now we can't figure out what <disk> should be in rootard's instructions.[03:49:03] <zooko> I have ZFS mirror boot disk.[03:49:20] <zooko> I can't figure out how to tell it.. Do I tell installgrub to put it on both of the devices?[03:49:26] <zooko> With two different invocations of installgrub?[03:49:56] *** randybias has quit IRC[03:51:47] *** GHReyes has joined ##nexenta[03:56:05] *** GHReyes has quit IRC[04:03:12] <zooko> Oh no! Same error 16: Inconsistent Filesystem State[04:10:33] *** tsukasa has quit IRC[04:10:50] *** tsukasa has joined ##nexenta[04:16:48] <zooko> Oh it wasn't "-x" it was "-f" for "force".[04:25:48] <zooko> Does anyone know how to tell grub to boot syspool/rootfs and not syspool/rootfs-nmu-004 ?[04:29:02] <zooko> sigh[04:29:04] <zooko> I give up.[04:29:12] <zooko> I'm asking Brian to scp all my data off of that server...[04:32:10] <dtbartle> aot-clone -a perhaps[04:32:12] <dtbartle> er[04:32:14] <dtbartle> apt-clone[04:32:26] <dtbartle> apt-clone -l should tell you the active clone[04:32:30] <dtbartle> (iirc)[04:32:36] <zooko> When I've booted from a rescue CD?[04:32:41] *** randybias has joined ##nexenta[04:32:42] <dtbartle> oh[04:32:50] <zooko> I can't boot.[04:32:57] <zooko> It always says Error 16: Inconsistent Filesystem State[04:33:41] <zooko> I've tried rebuilding my boot archive, although I'm not sure which fs to rebuild the boot archive on,[04:34:44] <zooko> and the bootadm update-archive emits mysterious error/warning messages[04:37:32] *** bobahPhD has joined ##nexenta[04:42:01] <rootard> zooko: there is a grub config option to specify which dataset to boot from[04:42:09] <rootard> there should be an entry already for it[04:42:44] <zooko> Well, yes I think we managed to try booting from syspool/rootfs, but editing the line in the grub stanza that said[04:43:03] <zooko> bootfs syspool/rootfs-nmu-004[04:43:07] <zooko> to say bootfs syspool/rootfs[04:43:22] <rootard> do you have an 003/002/001 ?[04:43:34] <zooko> Also we tried booting syspool/rootfs-nmu-004, after attempting do bootadm update-archive.[04:43:40] <zooko> In both cases we still got Error 16.[04:43:51] <zooko> I'm asking Brian to join irc...[04:43:56] <rootard> k[04:43:59] <zooko> bootadm always emitted warnings/errors.[04:44:20] <zooko> So I wonder if it actually failed to rebuilt the boot archive.[04:44:22] <rootard> like?[04:44:56] *** warner has joined ##nexenta[04:45:05] <zooko> warner: http://codepad.org/seoqBmND[04:45:05] <warner> heya[04:45:12] <zooko> rootard would like to know what were the messages from bootadm.[04:45:23] <zooko> Which might show that bootadm failed to rebuild the boot archive.[04:45:27] <zooko> Maybe we could do it again...[04:45:51] <warner> I'm going to run "bootadm update-archive -vR /mnt" from outside the chroot, right?[04:45:55] <zooko> rootard: I'm very glad you reappeared...[04:46:03] <zooko> Yes, that's what we did last time.[04:46:04] <zooko> Do that again.[04:46:17] <rootard> Sorry I disappeared in the first place... I had to go for a bike ride ;)[04:46:25] <zooko> Good! :-)[04:46:54] <rootard> warner: double checking options... but no you shouldn't run that from inside the chroot[04:46:57] <warner> so /mnt was mounted with "mount -F zfs syspool/rootfs /mnt"[04:47:08] <rootard> ie: /mnt[04:47:25] <zooko> rootard: I'm trying to fix the original version instead of the "upgrade" fs'es which were created by apt-clone.[04:47:25] <warner> and bootadm emitted three "cannot find:" lines and exited with rc=0[04:47:40] <warner> the first line was "cannot find: /mnt/etc/devices/mdi_ib_cache: No such file or directory"[04:47:42] <zooko> That's why we're trying syspool/rootfs instead of what we tried earlier: syspool/rootfs-nmu-004[04:47:51] <zooko> You asedk if we also had a 003/002/001 and the answer is yes.[04:47:54] <warner> the other two were for /mnt/etc/cluster/nodeid and /mnt/platform/i86pc/biosint[04:48:00] <zooko> apt-clone created these as I was trying repeatedly to upgrade the other day.[04:48:28] <zooko> rootard: Brian says he doesn't have an incoming ftp directory yet.[04:48:38] <zooko> I note that Tahoe itself has an FTP interface, so we could just use that. ;-)[04:48:42] <rootard> I don't have any of those files so we can remove them from the list[04:48:46] <warner> heh[04:49:23] <rootard> edit /boot/solaris/filelist.ramdisk[04:49:25] <zooko> warner: do you know how to remove them from the list?[04:49:27] <zooko> Ah.[04:49:27] <rootard> (I think)[04:50:06] <zooko> Wait. in /mnt ?[04:50:12] <rootard> Also, copy yeah[04:50:15] <zooko> Or in the rescue cd?[04:50:16] <rootard> sorry[04:50:23] <rootard> /mnt/boot/solaris/filelist.ramdisk[04:51:25] <zooko> Okay, so probably bootadm *was* successfully creating the boot archive.[04:51:30] <zooko> And we still get Error 16 on boot. :-([04:51:51] <warner> ok, filelist.ramdisk has been edited[04:51:57] <rootard> if it was generating errors then maybe not[04:52:05] <zooko> By the way it is a 33 GB system with 28 GB used.[04:52:07] <warner> shall I run the update-archive again?[04:52:15] <zooko> Brian: if you could copy aside a mere 8 GB of it or so.[04:52:16] <rootard> ok, now re-run bootadm ...[04:52:20] <zooko> Then it would have plenty of free space.[04:52:24] <zooko> Oh, we currently have a zpool scrub going.[04:52:26] <zooko> Should we stop it?[04:52:34] <rootard> no, just leave it[04:52:37] <rootard> it should be fine[04:52:41] <warner> ok, this time it didn't emit anything to stdout/stderr and had rc=0[04:52:58] <warner> "zpool status" reports the scrub is 61% done, eta 11 minutes[04:53:00] <rootard> ok, sounds better/[04:53:11] <rootard> let's do the same thing for the base rootfs as well...[04:53:30] <rootard> umount /mnt;mount -F syspool/rootfs /mnt;bootadm update-archive -vR /mnt[04:53:37] <rootard> oh wait[04:53:39] <zooko> That's the one we just did.[04:53:56] <rootard> ah, I see... then lets do the newer one[04:54:07] <rootard> umount /mnt;mount -F syspool/rootfs-nmu-004 /mnt;bootadm update-archive -vR /mnt[04:54:38] <zooko> Then we also need to do updategrub for both disks in the mirror for both the base and new snapshots? :-)[04:54:44] <rootard> then run installgrub from there... cd /mnt/boot/grub;installgrub stage1 stage2 (disk)[04:54:51] <rootard> and yes, you want to do this for both disks[04:55:04] <rootard> this inly installs the boot sector, everything else is read from the zpool[04:55:08] <warner> should I edit that filelist for -004 too?[04:55:09] *** dtbartle has quit IRC[04:55:13] <rootard> if one drive fails then you can boot from the other[04:55:35] *** dtbartle has joined ##nexenta[04:55:35] <rootard> warner: if it gives any errors then yes...[04:56:02] <warner> running update-archive without editing it first..[04:56:29] <warner> ok, got 5 complaints: etc/cluster/nodeid, etc/devices/mdi_ib_cache, device/retire_store, platform/i86hvm/kernel, platform/i86pc/ucode/AuthenticAMD[04:56:37] <warner> editing the filelist now..[04:58:17] <warner> ok, bootadm exited without output and rc=0[04:58:49] <warner> what should be mounted when I do installgrub?[04:59:01] <rootard> the latest rootfs (-004 I think)[04:59:15] <warner> ok, we've got -004 mounted now[05:00:01] <rootard> ok, once you installgrub for both drives we have one more thing to check[05:00:02] <warner> let's see, the installgrub command should be "installgrub stage1 stage2 /dev/rdsk/c1t0d0d0", I think[05:00:34] <rootard> well, d0d0 looks bad but I think you have the right idea[05:00:36] <warner> it should use the same c1t0d0s0 that's emitted in "zpool status", right?[05:00:54] <warner> zpool status says: syspool, mirror, c1t0d0s0, c1t1d0s0[05:00:59] <rootard> I believe so[05:01:04] <rootard> if not it will complain[05:01:19] <rootard> I can never remember between s0/s2/p0 ...[05:01:53] <warner> ok, that seems happy: "stage1 written to partition 0, sector 0" .. also stage2 written[05:02:01] <warner> now, same thing for c1t1d0s0?[05:02:31] <rootard> yes[05:02:54] <warner> done[05:03:05] <warner> (scrub is 73% complete)[05:03:11] <rootard> ok good.[05:03:14] <rootard> umount /mnt[05:03:31] <rootard> zfs upgrade syspool/rootfs-nmu-004[05:03:38] <zooko> So we didn't do updategrub on the syspool/rootfs[05:03:46] <rootard> that's right[05:03:57] <rootard> you only want to install the latest grub[05:04:02] <zooko> Yay! It is going to get faster![05:04:07] <zooko> I assume that's what "upgrade" means. ;-)[05:04:11] <warner> doh[05:04:17] <zooko> WHAT DID YOU DO[05:04:20] <zooko> ;-)[05:04:27] <warner> "syspool/rootfs-nmu-004: can not be upgraded; the pool version needs to first be upgraded to version 9"[05:04:44] <rootard> I have no idea what the upgrade means except that it increments a number that will save you frustration later.[05:04:59] <zooko> :-)[05:05:06] <rootard> ok, we don't want to upgrade the pool version until we know rootfs-nmu-004 works[05:05:17] <rootard> otherwise you _won't_ be able to boot rootfs[05:05:27] <warner> (scrub is 79% done)[05:05:38] <rootard> ok, I guess we can be safe and wait for it[05:05:56] <warner> 84% done[05:05:56] <rootard> unless you are in a hurry?[05:06:00] <warner> nope[05:06:17] <rootard> ok good. I'm hungry from my ride :)[05:06:25] <warner> 87%[05:06:40] <zooko> :-)[05:06:49] <zooko> Hm, maybe I should get a snack too...[05:06:54] <warner> (I'm in colo now, but I'm also working on my own box, trying to get a serial console running in a Xen dom0)[05:08:06] <warner> 90.28% :)[05:08:07] <zooko> Cool. :-)[05:08:11] <zooko> Serial console in dom0.[05:09:09] <rootard> yay for serial consoles...[05:09:19] <warner> 94.39%[05:09:25] <rootard> the best invention since toast[05:13:58] <zooko> I don't suppose the fact that it can't be upgraded wouldn't be responsible for the Error 16...[05:14:08] <warner> 99.13%[05:14:55] <warner> scrub is complete[05:15:03] <zooko> Everything is A-okay?[05:15:19] <warner> "0 errors"[05:15:26] <warner> so now what?[05:15:55] <zooko> I think rootard is getting food, but I suggest to reboot and choose the nmu 004 option from the grub menu.[05:16:01] <warner> ok[05:16:04] <rootard> back[05:16:28] <rootard> ok now time to init 6[05:19:14] <warner> rebooting[05:19:29] <warner> doh! 16: Inconsistent filesystem structure[05:19:34] <rootard> !?![05:19:35] <warner> when I booted from the nmu-004 option[05:19:49] <rootard> how about rootfs?[05:20:01] <warner> let me try that[05:20:17] <zooko> :-([05:20:35] <warner> that means editing the grub boot stanza to say "bootfs syspool/rootfs", right? instead of "bootfs syspool/rooftfs-nmu-004" ?[05:20:41] <rootard> right[05:20:56] <warner> same error[05:21:06] <rootard> hum[05:21:27] <warner> is there any way to be sure that we updated the right grub thing? a version number or something?[05:21:32] * zooko cries into his yogurt.[05:22:04] <rootard> I'm pretty sure you got it right.[05:22:10] <rootard> ok, I have another idea then[05:22:27] <rootard> if you can boot from the cd again we can try to rollback rootfs[05:22:33] <warner> ok, one minute[05:22:58] <rootard> there are probably several snapshots to look at[05:23:14] <rootard> also, are there other datasets? (rootfs-nmu-00[123]) ?[05:23:41] <zooko> For what it is worth, I did several apt-clone -s dist-ugprades that failed[05:23:43] <zooko> then one that succeeded[05:23:55] <zooko> then it asked if I wanted to "activate" the new snapshot or to reboot first, and I chose to reboot first.[05:23:59] <zooko> Then it didn't come back, and here we are.[05:24:05] <rootard> what made the last one succeed?[05:24:12] <warner> cd is booting now (it takes a while)[05:24:46] <warner> (there's another guy here in colo, and each time I walk around the rack to reboot the machine, we have the next 15 seconds of our conversation)[05:25:03] <zooko> :-)[05:25:13] <zooko> rootard: I wrote about it on this channel at the time.[05:25:21] <zooko> Part of it was I had to put a working ksh interpreter into place.[05:25:30] <zooko> Next I apt-get removed some non-essential packages that were failing to upgrade.[05:25:36] <rootard> ah, it's starting to come back[05:25:37] <zooko> I don't remember exactly what made it succeed.[05:25:44] <zooko> Got a channel log to remind me of my own words?[05:25:57] <zooko> I have one! It's on nooxie -- the server that Brian is resuscitating.[05:26:01] <rootard> http://echelog.matzon.dk/?nexenta[05:26:08] <warner> ok, I've got a shell again[05:26:11] <warner> what am I doing?[05:26:23] <rootard> zfs list -r syspool/rootfs -t snapshot[05:26:33] <rootard> or maybe -t snapshot needs to be before -r[05:27:16] <warner> yeah, it wanted -t first[05:27:31] <rootard> what exists there?[05:27:32] <warner> that lists syspool/rootfs@default and syspool/rootfs@nmu-00{1,2,3,4}[05:27:48] <warner> @default says 214M used, the others are much less[05:28:00] <zooko> [18:41:46] <zooko> Okay, I did sudo dpkg -i ksh_93q-2_solaris-i386.deb[05:28:07] <warner> @default says 682M "REFER", the others are all 28.4G or 28.3G[05:28:19] <rootard> hmm, lets have some fun[05:28:54] <rootard> zfs clone syspool/rootfs@default syspoo/rootfs-default[05:29:10] <rootard> s/syspoo/syspool/ ...[05:29:10] <warner> ok[05:29:24] <rootard> ok, that should be bootable :)[05:29:24] *** alfism has joined ##nexenta[05:29:33] <warner> done[05:30:14] <rootard> ok, let's see if we can remove the excess clones as well using apt-clone[05:30:27] <rootard> apt-clone -r rootfs-nmu-001[05:30:29] <rootard> apt-clone -r rootfs-nmu-002[05:30:30] <rootard> apt-clone -r rootfs-nmu-003[05:31:11] <warner> ok[05:31:25] <rootard> zooko: it's good if you put your data onto a different dataset than syspool.. just fyi[05:31:26] <warner> wait, is this from the install cd prompt? it has apt-clone?[05:31:40] <rootard> I think it should work.[05:31:47] <warner> or should we be mounting/chrooting somewhere first?[05:31:49] <zooko> rootard: well, I don't really understand what a dataset is...[05:32:00] <rootard> ok, we'll talk once the system is configured[05:32:03] <rootard> and running[05:32:08] <zooko> This is just what I got from the default (??) nexenta, either GNU/OpenSolaris or Elatte beta 1 or something...[05:32:13] <warner> I'll run those apt-clone commands now..[05:32:39] <zooko> Ok.[05:32:46] <warner> "This will destroy clone 'syspool/rootfs-nmu-001'. Proceed?"[05:32:50] <rootard> yes[05:33:06] <warner> "apt-clone.DeviceNotfound: Unable to find root device"[05:33:17] <rootard> ugh, ok[05:33:36] <rootard> let's try mounting and then bind mounting a few things to create a sane chroot[05:34:00] <rootard> mount -F zfs syspool/rootfs-num-004 /mnt;[05:34:17] <rootard> mount -F lofs /dev /mnt/dev[05:35:09] <warner> ok[05:35:11] *** anilg has joined ##nexenta[05:35:16] <rootard> replace dev with: devices,proc,etc/mnttab,tmp[05:35:27] <rootard> (rinse, lather, repeat)[05:35:46] <rootard> then chroot /mnt[05:35:54] <rootard> and try the apt-clone commands again[05:36:33] <warner> devices, proc, what?[05:36:48] <warner> really /etc/mnttab ?[05:37:17] <rootard> yeah[05:37:36] <warner> ok, mounted[05:38:00] <warner> running apt-clone -r ..[05:38:20] <warner> ok. "Upgrade changes for clone 'syspool/rootfs-nmu-001' now rolled back/destroyed"[05:38:28] <zooko> cool[05:39:13] <warner> now what[05:39:23] <rootard> you did that for 2/3 as well?[05:39:27] <warner> nope[05:39:29] <warner> doing that now[05:39:46] <warner> removing 1/2/3, preserving 4[05:39:50] <rootard> right[05:40:01] <warner> done[05:41:05] <warner> what now?[05:41:06] <rootard> awesome. ok, now when we reboot we should try to use rootfs-default ... if _that_ doesn't work then we are going to have to do more voodoo[05:41:12] <warner> ok[05:41:50] <rootard> I'll be very surprised it it doesn't work though.[05:41:53] <rootard> *if[05:42:04] <warner> what should the "bootfs" line say?[05:42:12] <zooko> Why is it going to work *this* time?[05:42:13] <rootard> syspool/rootfs-default[05:42:29] <zooko> Oh, because now we have a clone named syspool/rootfs-default instead of syspool/rootfs?[05:42:35] <rootard> because we are essentially booting from the initially installed environment[05:42:40] <rootard> right[05:43:24] <zooko> By the way, does anyone know how to tell the Python distutils to use a compiled named "gcc-3" instead of one name "gcc"?[05:43:29] <zooko> I guess this is the wrong channel for that question.[05:43:59] <warner> it said "Booting Nexenta" for a few seconds, then the screen went blank, then it rebooted back to the BIOS[05:44:23] <warner> I got there by choosing the nmu-004 option, then editing the "bootfs" line before starting it[05:44:23] <zooko> :-([05:44:25] <rootard> ok, put the options "-kdv" onto the kernel line[05:44:35] <warner> is that on the kernel$ line?[05:45:01] <rootard> when prompted type in :c<enter>[05:45:06] <warner> after the -B $ZFS-BOOTFS ?[05:45:18] <rootard> yeah, just after -B $ZFS-BOOTFS or whatever it is[05:45:49] <warner> ok, continuing...[05:45:57] <warner> panic: cannot mount root path /ramdisk:a[05:46:30] <rootard> hmm[05:47:14] <rootard> ok, two options zooko: we keep going down this path or we go for broke and move forward[05:47:41] <rootard> my first idea is to upgrade the zpool[05:47:52] <warner> zooko: I'm about 5 minutes away from getting you 30G of space on my colo box[05:47:53] <rootard> my second idea is to upgrade to NCP2[05:48:16] <rootard> that can be a workable solution as well...[05:48:33] *** anilg has quit IRC[05:49:56] <zooko> How about, whichever you think is the most likely to succeed with the least work? :-)[05:50:04] <zooko> I have a fallback plan which is to copy all my data off...[05:50:11] <zooko> So I'm willing to try either of those.[05:50:13] <zooko> Or both.[05:50:34] <zooko> warner: would you please join ##darn-broken-box[05:56:05] *** anilg has joined ##nexenta[06:01:40] *** fserve has joined ##nexenta[06:03:29] <anilg> rootard: still around?[06:04:21] <rootard> for a bit. helping zooko in private[06:06:27] *** randybia has joined ##nexenta[06:23:09] *** randybias has quit IRC[06:39:29] *** randybias has joined ##nexenta[06:44:40] *** randybia has quit IRC[06:58:44] *** anilg has left ##nexenta[06:59:48] *** Capscrew has joined ##nexenta[07:04:32] *** corwin has joined ##nexenta[07:04:45] *** Capscrew has left ##nexenta[07:05:07] <corwin> whats the best way to back up a zfs pool from one machine to another over the net?[07:05:23] <rootard> zfs send[07:05:40] <corwin> all the scripts I've tried useing send and receive are terrible broke[07:05:40] <rootard> you can pipe the output through ssh is needed[07:06:15] <rootard> if you are doing anything that will fill up a gig-link (like between two thumpers on a local lan) you can pipe through nc instead for speed gains (but no security)[07:06:25] <corwin> I prefer ssh[07:06:46] <rootard> I do too... until I am moving multiple TB of data ... :)[07:06:50] <corwin> haha[07:07:02] *** randybias is now known as randybias|away[07:07:07] <corwin> any good docs? What I run into is either outdated or broke[07:07:24] <rootard> the manpage (if it works)[07:07:48] *** lesterc has joined ##nexenta[07:07:49] <corwin> in Nextena.... bash: man: command not found[07:08:57] <corwin> my receiving machine is on a dynamic IP..... So it need to do a "pull"[07:09:04] <rootard> maybe "apt-get install manpages"[07:09:06] <rootard> or http://docs.sun.com/app/docs/doc/819-2240/zfs-1m[07:09:45] <rootard> ssh remote-host zfs send -R syspool... | zfs recv ...[07:09:46] <corwin> wow, that works (apt-get instal manpages). Thanks![07:10:31] *** fserve has quit IRC[07:10:43] <rootard> np[07:18:53] *** ohol1ks is now known as chol[07:25:05] <corwin> Any idea why I can ste up passwordless ssh logon from a linux machien to nexenta but not nexenta to nexenta?[07:26:31] <rootard> how are you implementing the passwordless login? keys? krb5? ...[07:26:43] <rootard> my guess is credentials aren't being forwarded[07:27:26] <corwin> keys, coping id_rsa.pub into authorized_keys[07:27:47] *** lesterc has quit IRC[07:31:03] <corwin> ok, I admit retardation.[07:31:10] <corwin> I had the wrong IP in hosts[07:32:03] <rootard> oh noes! not /etc/hosts![07:33:20] <corwin> I should learn that when I get really pissed off, it's hosts[07:33:51] <rootard> when I get really pissed off it's a PHB[07:34:12] <corwin> haha. That's one advantage to being self employed with cool clients[07:35:33] <corwin> How's life in PA?[07:36:40] <rootard> warming up. I've been getting back on bike lately... after my ride today I thouroughly fit the description, "mud man"[07:37:28] *** eXeC001er has joined ##nexenta[07:38:17] <rootard> I'm impressed you remembered PA though :)[07:38:51] <eXeC001er> I need to edit the image of NCP2RC1.[07:39:21] <eXeC001er> I nedd[07:39:32] <corwin> haha. What Kind of bike? I just got a Harley Sportser in trade[07:39:59] <rootard> Trek 6700 ;)[07:40:00] <eXeC001er> I need to add PV kernel.[07:40:18] <corwin> You're the guy who always helps me when I have nexenta trouble, the least I could do is rememebr where you moved to ;)[07:40:52] <rootard> :-D[07:41:10] <corwin> cool bike. I could use a little peddle pushing myself[07:41:41] <rootard> I find it's pretty appropriate for the landscape around here[07:41:50] <eXeC001er> I pytasya edit using isomaster.[07:42:00] <corwin> I've been taking resveratrol (1.5g/day) and that's been taking some off my waist.[07:42:46] <rootard> I like to keep away from anything with a fancy name myself.[07:42:58] <eXeC001er> I tried to edit the image using isomaster, but failed. Image don't booting.[07:43:49] <rootard> eXeC001er: are you following justin thomas's instructions or going it on your own?[07:44:15] *** baitisj has joined ##nexenta[07:44:56] <rootard> though I certainly wouldn't mind losing some of my extra gut-volume[07:45:02] <corwin> it has a simple name -> red wine antioxidents[07:45:24] <corwin> or just good stuff from red wine[07:46:08] <corwin> the waist issue is a side effect, it increases the lifespan of lab animals an average of 60%[07:46:25] <eXeC001er> I want to try to make the setup mode PV.[07:46:25] <rootard> aha, I didn't realize that helped with the waist.[07:46:57] <rootard> eXeC001er: ok, then copy the boot_archive off of the CD and the you can loopback mount it to modify it[07:47:07] <eXeC001er> To do this I need an image with the kernel_PV.[07:47:17] <corwin> it improved my vision withing minutes of taking it.. It has interesting (and really good) side effects[07:49:11] <rootard> I've had a very long policy of staying away from alcohol but I've heard that very small amounts of red wine can be very beneficial[07:49:39] <eXeC001er> No, I need to edit "iso".[07:50:06] <rootard> eXeC001er: to just install via PV you don't actually.[07:50:21] <rootard> you just need to modify the boot_archive and use the right kernel in your xen config[07:50:55] <rootard> the rest of the CD should be the same thus the iso itself doesn't need to be altered[07:51:56] <eXeC001er> no, i use pygrub for boot NCP.[07:52:29] <rootard> pygrub still looks at the host filesystem, not the grub on the ISO though... I think?[07:53:20] <eXeC001er> One moment.[07:53:44] <rootard> I could be wrong...[07:54:20] <eXeC001er> http://bderzhavets.wordpress.com/2009/03/20/pygrub-install-opensolaris-200811-at-xen-unstable-dom0-with-2629-rc7-pvops-enabled-kernel/[07:54:25] <rootard> anyway if you do modify the ISO and don't want to use nexenta-builder then you'll need to become a little more familiar with some form of iso building tool.[07:56:56] *** nightwish has quit IRC[07:57:53] *** kisielk_home has joined ##nexenta[08:01:01] <kisielk_home> does the current version of nexenta let you have a raid-z root?[08:01:13] <kisielk_home> I'm setting up a new fileserver, and trying to decide between opensolaris and nexenta...[08:02:02] <rootard> nope, grub can not boot from a raidz pool[08:03:56] <kisielk_home> but I can do a mirrored rpool for the root[08:04:04] <kisielk_home> on the start of each disk, no?[08:04:11] <kisielk_home> and raidz on the rest[08:05:03] <kisielk_home> hm, I guess that doesn't work so well for ZFS[08:05:05] <rootard> if you are careful about it ... yes but it's not supported by the installer afaik[08:05:15] <rootard> zfs likes to have whole disks[08:05:15] *** Subvertir has joined ##nexenta[08:05:17] <kisielk_home> I guess I will need a fourth disk of some sort then[08:06:03] <Subvertir> Hey guys, is there currently any planning/development going on for a "deskop" version of Nexenta? e.g., something more comparable to what's in OpenSolaris 08.11 or Ubuntu[08:06:32] <kisielk_home> it just seems silly to use a 500GB disk for my root[08:06:47] <rootard> stormos is attempting to cover the desktop and it is being developed on by andy_js (not in the chan right now)[08:07:08] <Subvertir> rootard, thank you![08:07:09] <rootard> kisielk_home: you can still create other datasets on there and not keep everything in your root dataset[08:07:30] <rootard> Subvertir: np[08:07:42] <kisielk_home> rootard: yeah, but the primary purpose of this machine is a 3 disk raidz storage[08:07:46] <kisielk_home> not sure what I would use the rest of the other disk for[08:08:12] <rootard> well, you could boot off a single disk for your root pool and then "set copies=2" to keep on-disk redundancy[08:08:34] <rootard> as long as the entire disk doesn't fail you are ok... this would cover blocks going bad[08:09:05] *** MrSteveman1 has joined ##nexenta[08:09:58] <Subvertir> Are there license issues preventing direct ports of GPL wifi drivers to OpenSolaris?[08:10:03] *** fserve has joined ##nexenta[08:11:29] <rootard> Subvertir: good question. I'm not sure it would be easy to directly port the drivers but if they _could_ be ported then there may be some issues with GPL code linking against CDDL[08:12:28] <kisielk_home> oh wonderful, one of these disks is bad[08:12:35] <kisielk_home> I guess I won't be having my 3 disk raidz just yet :)[08:12:38] *** synan has joined ##nexenta[08:13:27] <rootard> bummer[08:15:07] <kisielk_home> I'll have to get more disk tomorrow[08:16:11] <Subvertir> it'd be very cool if I could install solaris and macos onto the same zfs-managed drives (and boot from both, or run them in a VM) - that'd be the perfect setup for my two disk laptop[08:16:44] <kisielk_home> Subvertir: you can install solaris on the second drive, and run it in a VM from OS X[08:17:20] <Subvertir> Are you sure? Parallels and VirtualBox don't seem to allow me to run virtual machines from physical disks[08:17:29] <Subvertir> I know you could do it in older versions of VMWare[08:17:35] <kisielk_home> really?[08:17:57] <Subvertir> Yeah, I could be wrong, I didn't look *that* hard, but I didn't see any obvious option[08:18:09] <Subvertir> it used to be part of the vmware device creation wizard but iirc it isn't even there anymore[08:18:27] <kisielk_home> its on the versions of VMWare Workstation I have[08:19:00] <Subvertir> I'll double check, it'd be very nice if I could do that[08:19:06] <rootard> VM's are so slow though[08:19:29] <kisielk_home> I haven't noticed[08:19:38] <kisielk_home> I've got 4 VM's running on this workstation[08:19:46] <kisielk_home> plenty fast enough for me :)[08:19:48] <Subvertir> yeah honestly parallels is pretty decent, even 3D accel is reasonable[08:19:50] * rootard nods[08:20:25] <rootard> they drive me crazy unless I'm on super-new/fast/expensive hardware... ie: not what I personally own[08:20:44] <kisielk_home> ah ok, well, this one is a quad with 8GB of RAM[08:20:56] <kisielk_home> so a few VM's is no problem[08:21:05] <Subvertir> yeah the ram is really important, im on a 2.2ghz c2duo with 8gb ram[08:21:20] <kisielk_home> totally[08:21:29] <rootard> yeah, otherwise you get a swapfest[08:21:30] <kisielk_home> I also run Photoshop and a bunch of other heavy graphics apps here[08:21:36] <kisielk_home> and the RAM helps in a big way[08:21:44] <kisielk_home> I typically allocate about 512MB to each VM[08:21:58] <kisielk_home> those I use for software dev[08:21:59] <Subvertir> I just run one VM at a time and usually give 2gb[08:21:59] <Bartman007> they don't want to field calls from mac users who nuked their internal drive, so using physical disks isn't an option.[08:22:12] <Bartman007> (unless it is managed via bootcamp)[08:22:31] <Subvertir> that's incredibly weak lol[08:23:11] <rootard> I wonder if you could specify /dev/... ask a disk[08:23:39] <rootard> probably not ...[08:23:54] <rootard> I know VB allows you to view an iSCSI device as a disk[08:24:17] <rootard> it's not in the GUI but avialable via cli[08:24:17] <kisielk_home> if only SSD's were cheaper[08:24:24] <Subvertir> I'll just run time machine and nuke macos[08:24:54] <kisielk_home> you can just install it in a disk image file[08:25:02] <kisielk_home> it doesn't really hurt performance in my experience[08:25:18] <Subvertir> I wanna run it natively (or at least have the option)[08:25:39] <Subvertir> it doesn't take too long to restore macos from a backup so it's all good[08:26:01] <Subvertir> (I already do have it installed in a disk image though hehe)[08:28:04] <kisielk_home> so the question is how many disks to buy tomorrow ..[08:28:34] <rootard> how many can your system handle?[08:28:42] <kisielk_home> 6[08:28:55] <rootard> that sounds like a good number :)[08:29:00] <kisielk_home> haha[08:29:05] <kisielk_home> I have 2 (working) currently[08:29:14] <kisielk_home> I could get another pair of 500GB disks[08:29:22] <kisielk_home> for around $120[08:30:17] <rootard> so you want to have two mirrors?[08:30:34] <kisielk_home> well, I was thinking of doing raidz[08:30:36] <kisielk_home> and a root disk[08:30:57] <kisielk_home> originally my plan was just a single mirror[08:31:17] <kisielk_home> I guess I could stick with that for now[08:31:22] <kisielk_home> and just add more mirrors as I need them[08:31:30] <kisielk_home> disks will only get cheaper after all[08:31:48] *** master_of_master has joined ##nexenta[08:31:53] *** taltamir has quit IRC[08:32:14] <kisielk_home> If I create one mirror now[08:32:40] <kisielk_home> I could add another one later, and then remove the first one at a future time[08:32:40] <rootard> or if you are willing to go with a single root drive you could get an SSD for root, another disk and have your 3-disk raidz[08:33:07] <kisielk_home> I think a mirror would probably be enough for now[08:33:26] <kisielk_home> the nice thing about being able to grow and shrink the storage pools is that it's flexible[08:34:18] <kisielk_home> so I guess the next choice is[08:34:28] <kisielk_home> straight on opensolaris, or nexenta ? :)[08:35:10] <rootard> being in the ##nexenta chat room, would you really expect me to say anything other than Nexenta?[08:35:47] <rootard> :)[08:36:20] <rootard> btw: you can't remove a device from a pool unless there is something to replace it[08:36:46] <Subvertir> I'm surprised there isn't a lot more interest in nexenta tbh[08:36:52] <kisielk_home> hm, but if you grow the pool to have enough redundant space[08:36:56] <kisielk_home> can you not remove some devices?[08:36:57] <rootard> also, you can't add a disk to a raidz group... you would need to add another raidz group (or re-create the pool[08:37:11] <kisielk_home> like, if I have a mirror of 500GB disks, then later I add a mirror of 1TB disks to the same pool[08:37:14] <rootard> you can replace devices with larger ones[08:37:15] <zooko> I'm surprised that Sun hasn't blessed Nexenta with money and marketing and so on.[08:37:18] <rootard> and you can add devices[08:37:21] <kisielk_home> can I not tell ZFS to remove the 500GB disks completely?[08:37:29] <Subvertir> it seems like a really cool project, my understanding is that the solaris standard for backwards compatibility is "forever" so you can have stable 3rd party drivers etc[08:37:48] <zooko> They keep *saying* that they want to appeal to Linux users, but they aren't actually *doing* it by actually implementing apt, or by just using the implementation that already exists: Nexenta.[08:37:52] <kisielk_home> or will I have to replace them with something?[08:38:08] <rootard> Subvertir: it's not "forever" it's something like 3 releases[08:38:11] <Subvertir> zooko, I agree - being able to bootstrap off of the ubuntu/debian repos is huge[08:38:42] <Subvertir> rootard, but there's a serious commitment to maintaining a relatively stable ABI unlike linux where it can/does change every minor release[08:39:18] <rootard> yes, but it's certainly not forever :)[08:39:48] <Subvertir> right ;)[08:39:52] <Subvertir> close enough![08:40:41] * synan still misses the odd-number-means-testing kernel difference[08:41:01] <synan> I'm wanting to try nexenta but can't spare a whole disk for it..[08:43:17] *** master_o1_master has quit IRC[08:45:26] <kisielk_home> hm[08:45:35] <kisielk_home> so it seems even with mirror, I probably want a separate disk for the root pool[08:47:06] <Subvertir> I tried ZFS with FreeBSD a while back[08:47:43] <rootard> Subvertir: I've heard it's not so great[08:48:13] <Subvertir> it was alright, there's some pretty straightforward tuning if you have <2gb of RAM (which I did in that machine), but I didn't have any trouble running it[08:48:24] <Subvertir> it seemed fine for desktop usage[08:48:52] *** andy_js has joined ##nexenta[08:49:22] <rootard> andy_js: Subvertir was asking about the existence of stormos[08:49:38] <Subvertir> yeah it looks really cool[08:50:31] <Subvertir> I was thinking about running kde4 (I haven't ran it at all yet), but XFCE + compiz would probably be really fast[08:51:12] <Subvertir> does flash work with firefox 3?[08:52:14] <Subvertir> http://labs.adobe.com/technologies/flashplayer10/releasenotes.html - is anyone even using it?[08:55:56] <andy_js> Subvertir: flash is on my very long todo list[08:56:27] <Subvertir> cool, I'll try to get it working[08:58:51] <andy_js> I don't think it'll be too hard to do[09:01:00] <kisielk_home> so based on past experiences[09:01:10] <kisielk_home> is upgrading between releases of nexenta relatively painless?[09:02:04] <rootard> kisielk_home: upgrading from NCP1 -> NCP2 is supposed to work. If it doesn't then file a bug[09:02:33] <rootard> upgrading from pervious installs to NCP1 is a PITA sometimes and currently the only reason I am awake at this hour[09:03:05] <kisielk_home> well, I'll be starting with NCP2 RC1 I guess[09:08:54] <andy_js> Subvertir: in an hour or so I should have a stormos-desktop meta-package which you can install on NCP2[09:09:31] <andy_js> It comes with XFWM's compositing turned on by default, so you get things like drop shadows on windows[09:09:48] <andy_js> Who needs compiz anyway?[09:11:52] <Subvertir> haha awesome ;) I appreciate the heads up, I'll definitely try it out[09:12:43] <Subvertir> I really like http://roscidus.com/desktop/[09:13:08] <Subvertir> I'm thinking it might be pretty sweet with compiz ;)[09:14:58] *** gbs_ has joined ##nexenta[09:15:07] *** fserve has quit IRC[09:21:01] *** gbs_ has quit IRC[09:22:05] *** Subvertir has quit IRC[09:31:56] *** Subvertir has joined ##nexenta[09:49:32] *** corwin has quit IRC[09:49:33] *** corwin has joined ##nexenta[09:50:15] *** corwin has quit IRC[09:50:17] *** corwin has joined ##nexenta[09:52:26] *** zooko has left ##nexenta[09:58:14] *** corwin has quit IRC[09:58:16] *** corwin has joined ##nexenta[09:58:32] *** CookieEater has joined ##nexenta[09:58:55] *** CookieEater has left ##nexenta[10:10:52] *** CookieEater has joined ##nexenta[10:11:04] *** corwin has quit IRC[10:11:06] *** corwin has joined ##nexenta[10:11:40] *** corwin has joined ##nexenta[10:14:51] *** CookieEater has quit IRC[10:15:14] *** anilg has joined ##nexenta[10:15:22] *** corwin has quit IRC[10:15:25] *** corwin has joined ##nexenta[10:34:22] <andy_js> Subvertir: the "stormos-desktop" meta-package is now in the StormOS repo, feel free to try it out[10:41:52] <anilg> andy_js: the python-dev issue was resolved?[10:42:18] <andy_js> yep, by using aptitude instead of apt-get :p[10:46:23] *** warner has quit IRC[10:59:42] *** andy_js has quit IRC[10:59:59] *** andy_js has joined ##nexenta[11:12:56] *** nightwish has joined ##nexenta[11:57:20] *** kisielk_home has quit IRC[12:06:27] *** anilg1 has joined ##nexenta[12:26:19] *** anilg has quit IRC[12:44:52] *** anilg1 has quit IRC[12:45:33] *** anilg has joined ##nexenta[13:15:01] *** corwin has quit IRC[13:15:03] *** corwin has joined ##nexenta[13:15:32] *** corwin is now known as Guest71922[13:21:14] *** synan has quit IRC[13:22:13] *** mib_y70n8v has joined ##nexenta[13:28:18] *** mib_y70n8v has quit IRC[13:42:43] *** eXeC001er has quit IRC[13:43:05] <koan> andy_js: nice screenshots on stormos.org[13:45:58] <andy_js> koan: thanks[13:47:25] <andy_js> koan: are you going to try it out?[13:49:40] <koan> I'm planning to[14:13:50] <andy_js> cool[14:13:55] *** proteusguy has quit IRC[14:14:25] <andy_js> anilg: if its not too much trouble, could I get the nexenta banner with a white background?[14:14:40] *** proteusguy has joined ##nexenta[14:16:07] <koan> I see that Nexenta is sponsoring the StormOS project, in what way is that?[14:43:05] *** proteusguy_ has joined ##nexenta[14:43:32] *** proteusguy has quit IRC[14:56:16] *** Subvertir has quit IRC[14:58:28] *** eXeC001er has joined ##nexenta[15:21:14] <anilg> andy_js: done.[15:21:25] <andy_js> thanks[15:23:47] <anilg> ~package inkscape[15:23:48] <nexybot> anilg: http://builder.tajinc.org/?f=package_status&view_method=name&package_name=inkscape[15:24:21] <anilg> that's cool![15:25:37] *** randybias|away is now known as randybias[15:41:28] *** pino42 has joined ##nexenta[15:43:06] *** randybias is now known as randybias|away[15:43:12] *** randybias|away is now known as randybias[15:49:48] *** anilg has quit IRC[16:01:11] *** randybias is now known as randybias|away[16:09:33] *** randybias|away is now known as randybias[16:22:50] *** randybias has quit IRC[16:37:31] *** proteusguy_ has quit IRC[16:38:04] *** proteusguy has joined ##nexenta[16:49:00] *** Subvertir has joined ##nexenta[17:09:06] *** bobahPhD has quit IRC[17:10:48] *** randybias has joined ##nexenta[17:12:19] *** Subvertir has quit IRC[17:24:03] *** eXeC001er has quit IRC[17:50:33] *** randybias is now known as randybias|away[17:59:30] *** randybias|away is now known as randybias[18:07:59] *** ekix has joined ##nexenta[18:15:59] *** randybias is now known as randybias|away[18:39:36] *** anilg has joined ##nexenta[18:58:17] *** randybias|away is now known as randybias[19:04:26] *** Subvertir has joined ##nexenta[19:29:30] <andy_js> ~package update-manager[19:29:30] <nexybot> andy_js: http://builder.tajinc.org/?f=package_status&view_method=name&package_name=update-manager[19:31:44] <andy_js> so what does this actually mean?[19:34:42] *** anilg1 has joined ##nexenta[19:34:57] <anilg1> That the package needs to be ported[19:35:23] <andy_js> how do you tell?[19:35:45] <andy_js> ah, I see[19:35:57] <andy_js> Nexenta/Hardy == Ported[19:36:01] <anilg1> Right[19:36:31] *** xinkeT has quit IRC[19:37:12] *** anilg1 has quit IRC[19:40:45] *** xinkeT has joined ##nexenta[19:53:59] *** alfism has quit IRC[20:02:07] *** anilg has quit IRC[20:21:06] *** sabor has quit IRC[20:22:57] *** sabor has joined ##nexenta[20:50:01] *** eXeC001er has joined ##nexenta[20:54:42] *** alfism has joined ##nexenta[20:55:27] *** olsner has joined ##nexenta[21:05:47] *** randybias is now known as randybias|away[21:19:54] <andy_js> I think I've found those native ext3/ntfs drivers for solaris:[21:20:01] <andy_js> http://blogs.sun.com/pradhap/entry/mount_ntfs_ext2_ext3_in[21:26:28] *** xolaris has joined ##nexenta[21:27:51] <xolaris> I'm working with NCP2. Thinking of upgrading openssl to the latest patch. Not available from apt-get[21:28:34] <xolaris> To install from source means removing several packages including a SUNW meta package.[21:28:47] <xolaris> Not sure if this is a good idea... Any thoughts?[21:31:21] <rootard> xolaris: I think there may be some Sun patches in openssl... if you were to include those as well then it may be ok to replace the sunw package on your system[21:34:34] <xolaris> Thanks, rootard. I'll leave it alone for now. What's the nexenta policy of upgrades to the sunw packages?[21:37:51] <rootard> upgrades usually come as the upstream bits are updated.[21:38:51] <rootard> the sunw* packages are built internally by nexenta and are based on stabilization builds and critical patches[21:40:36] <rootard> andy_js: have you tried installing the packages mentioned in that article?[21:41:00] <xolaris> ok, thanks. I'm coming from the FreeBSD world and ports. So I'm trying to figure a strategy for keeping up-to-date.[21:44:53] <rootard> I can understand that. I would like to see less sunw* depends in general but there are some issues with just yanking out certain packages and replacing them with vanilla builds.[21:46:08] <rootard> if openssl is of large concern to you then you could probably take on the Debian/Ubuntu openssl package and maintain a patch for Nexenta.[21:47:20] <andy_js> rootard: No I've not tried the drivers yet.[21:49:01] *** rootard is now known as rootard_away[21:51:10] <xolaris> rootard: yes, I'd like to help, but I'm still trying to get my head around nexenta/debian packaging...[21:52:35] *** xolaris has left ##nexenta[21:54:15] *** eXeC001er has quit IRC[21:56:22] *** theatro has joined ##nexenta[21:57:36] *** Subvertir has quit IRC[22:12:02] *** zooko has joined ##nexenta[22:38:22] *** andy_js has quit IRC[22:38:57] *** andy_js has joined ##nexenta[22:41:30] *** fasfas3 has joined ##nexenta[22:44:16] *** TALzz has joined ##nexenta[22:46:15] <fasfas3> has /usr/lib64 been removed from the system ?[23:05:59] *** baitisj1 has joined ##nexenta[23:14:33] *** randybias|away is now known as randybias[23:21:15] *** fasfas3 has quit IRC[23:22:48] *** zooko has quit IRC[23:23:44] *** theatro has left ##nexenta[23:33:26] *** TALzz has left ##nexenta[23:45:03] *** randybias is now known as randybias|away[23:45:39] <andy_js> fasfas3: I think /usr/lib64 is actually /usr/lib/amd64[23:46:08] <andy_js> or at least it was on another opensolaris based distro I used :p[23:48:29] *** alfism has quit IRC[23:51:24] *** Subvertir has joined ##nexenta[23:56:26] <andy_js> bye everyone[23:56:35] *** andy_js has quit IRC[23:57:27] *** randybias|away is now known as randybias