Switch to DuckDuckGo Search
   March 11, 2013  
< | 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 | >

Toggle Join/Part | bottom
[00:01:21] *** fsteinel has quit IRC
[00:24:04] *** jelmd has quit IRC
[00:30:24] *** jelmd has joined #smartos
[00:35:41] *** Cpt-Oblivious has quit IRC
[00:46:03] *** enmand has joined #smartos
[01:04:39] *** enmand has quit IRC
[01:04:45] *** denizr has joined #smartos
[01:10:21] *** potatosalad has joined #smartos
[01:22:36] *** echelog-1 has joined #smartos
[01:37:33] *** matthewpucc_ has joined #smartos
[01:37:41] *** jefferai has joined #smartos
[01:38:37] *** id_ has joined #smartos
[01:39:06] *** nahamu_ has joined #smartos
[01:42:32] *** ikwildrp1pper has joined #smartos
[01:43:09] *** ivan\_ has joined #smartos
[01:44:18] *** nahamu has quit IRC
[01:44:21] *** nefilim has quit IRC
[01:44:22] *** texarcana has quit IRC
[01:44:27] *** ferai has quit IRC
[01:44:27] *** ivan\ has quit IRC
[01:44:30] *** id has quit IRC
[01:44:35] *** ikwildrpepper has quit IRC
[01:44:37] *** ClausA has quit IRC
[01:44:39] *** matthewpucc has quit IRC
[01:45:28] *** ivan\_ is now known as ivan\
[01:47:40] *** jgrafton has quit IRC
[01:51:31] *** ClausA has joined #smartos
[01:53:09] <trentster> Is there a way to get virtual machine individual uptime via vmadm or from the GZ?
[01:59:53] *** denizr has quit IRC
[02:18:29] <elijah-mbp> trentster: i imagine there's a kstat for that.
[02:18:35] <elijah-mbp> let me go poke in and look.
[02:21:20] <trentster> elijah-mbp: thanks
[02:22:05] <elijah-mbp> depending on what you want to do - there is a crtime for each zone that records the offset from boot time of the machine. (e.g., the gz has a crtime of 0)
[02:22:33] <elijah-mbp> if its a VM, you'd probably just want to know the timestamp when that copy of qemu started - yes?
[02:22:52] <trentster> elijah-mbp: a bit off topic but would you know how to get basic cpu and network stats for each vm via kstat?
[02:23:03] <elijah-mbp> that's probably marginally more precise than the crtime of hte wrapping zone is
[02:23:05] <elijah-mbp> trentster:
[02:23:07] <elijah-mbp> yes
[02:23:36] <elijah-mbp> it's a flood of data :)
[02:23:42] <trentster> elijah-mbp: The timestamp when that copy of qemu started would suffice , yes ;-)
[02:24:38] <elijah-mbp> you can do kstat -p and grep it out - if you know the zone boot ID of your zone - e.g. if you know that you are zone ID 3 from zoneadm list -vc - then you will be grepping for ^z3:
[02:25:19] <trentster> elijah-mbp: yeah but doesent the zone id change every time you power off a vm and back on again?
[02:25:46] <elijah-mbp> trentster: yes, and it's a huge PITA for scripting. :)
[02:26:05] <elijah-mbp> usu you wanna immediately map the UUID back to the zN so you can use it for scripts. first thing.
[02:26:07] <trentster> elijah-mbp: yikes, yeah, I can believe it
[02:26:33] <elijah-mbp> so on one of my GZs - i just did kstat -p | grep z4_net
[02:26:42] <elijah-mbp> and it spews the list of net things you might want to know back at you
[02:26:45] <trentster> so some powerful script Fu is needed then...
[02:26:54] <elijah-mbp> from there, you can look at things like ipackets, opackets
[02:27:16] <elijah-mbp> honestly if you don't NEED the data in the kernel's native form - you want to use dladm create-flow for network.
[02:27:32] <elijah-mbp> for network in particular.
[02:31:58] *** enmand has joined #smartos
[02:32:19] <elijah-mbp> kstat -p and kstat -l are good places to dig in
[02:32:41] <elijah-mbp> because once you know what those do and have some minimal understanding of the object hierarchy for kstats - you can grep around and find the bits you need pretty quickly.
[02:37:14] *** elijah-mbp is now known as eli-away
[02:47:17] *** enmand has quit IRC
[03:17:21] <nahamu_> what about ctime of the current VM log file?
[03:18:05] <nahamu_> ( /zones/<uuid>/root/tmp/vm.log )
[03:20:10] *** rbrown_ has joined #smartos
[03:21:27] <rbrown_> how different is the smartos kernel compared to the Oracle Solaris one?
[03:29:39] <jesse_> I'd say marginally, without any real knowledge of differing line counts
[03:33:04] *** potatosalad has quit IRC
[03:35:27] <nahamu_> rbrown_: depending on what you're trying to accomplish, either "not very" or "critically"
[03:39:56] <jesse_> when I was compiling openjdk, I was wondering when will it be right time for illumos to shed sunos uname
[03:40:34] <jesse_> (and os name, obviously)
[03:40:58] <jesse_> it'll break everything, but it'll be easy to fix at first
[03:43:42] <jesse_> there are already two kernels, sunos(solaris) and illumos, and several distributions, solaris, openindiana, omnios, smartos, etc.
[03:44:37] *** jelmd has quit IRC
[03:44:59] <jesse_> (and, actually, packaging system on top, as pkgsrc can be used in any of the distributions)
[03:45:26] <rbrown_> I think it would break way too many utilities
[03:45:36] <jesse_> at first, yes
[03:45:46] <jesse_> it'd break everything, basically
[03:46:24] <jesse_> but, for example, in the openjdk I had to just do #ifdef solaris && gcc -> illumos running with pkgsrc
[03:46:47] <jesse_> (or to be more specific, assume smartos location of pkgsrc, /opt/local
[03:47:10] *** serengeti has joined #smartos
[03:47:35] <jesse_> (ofcourse there were bits where solaris && gcc was the right limitation, for now, as illumos and oracle solaris are close enough)
[03:47:58] <jesse_> and then there were the X11 font paths
[03:48:07] <jesse_> that was just ...insane
[03:48:22] <jesse_> the build considered 'linux' a single distribution
[03:48:53] <jesse_> so it had all the possible locations of X11 fonts in any linux distro they've encountered
[03:49:10] <jesse_> (it wasn't even 50 lines long!=))
[03:51:01] *** jelmd has joined #smartos
[04:00:50] *** serengeti has left #smartos
[04:06:42] *** abhishekvasisht has joined #smartos
[04:13:52] *** nikolam has quit IRC
[04:25:40] *** darjeeling has joined #smartos
[04:43:16] *** tonyarkles has joined #smartos
[05:13:33] *** emocakes has joined #smartos
[05:15:14] *** emocakes has quit IRC
[05:27:57] *** jelmd has quit IRC
[05:33:13] *** ivan\ has quit IRC
[05:36:04] *** tonyarkles has quit IRC
[05:46:34] *** dizko has quit IRC
[05:46:46] *** sachinsharma has joined #smartos
[05:52:30] *** tonyarkles has joined #smartos
[05:55:51] *** abhishekvasisht has quit IRC
[05:56:28] *** abhishekvasisht has joined #smartos
[06:04:03] <rmustacc> trentster, elijah-mbp: The uptime of a qemu process does not correlate to the uptime of a guest.
[06:04:16] <rmustacc> All it takes is someone typing reboot and the data will be wrong.
[06:04:59] <rmustacc> As a guest rebooting via 'reboot' does not cause qemu to exit.
[06:07:36] <trentster> rmustacc: thanks, so what would be the correct indicator?
[06:10:30] <rmustacc> Ask the guest.
[06:10:55] <rmustacc> That's one of the problems of hardware virtualization.
[06:22:18] *** tonyarkles has quit IRC
[06:29:57] *** ivan\ has joined #smartos
[06:33:11] *** abhishekvasisht has quit IRC
[07:16:15] *** abhishekvasisht has joined #smartos
[07:17:42] *** abhishekvasisht has quit IRC
[07:19:33] *** abhishekvasisht has joined #smartos
[07:19:59] *** abhishekvasisht has quit IRC
[07:30:13] *** JT-EC_ has joined #smartos
[07:30:55] *** JT-EC has quit IRC
[07:34:11] *** gutleib has joined #smartos
[07:44:07] *** dizko has joined #smartos
[07:49:01] *** emocakes has joined #smartos
[07:51:49] *** darjeeling has quit IRC
[07:58:22] *** darjeeling has joined #smartos
[07:59:39] *** emocakes has quit IRC
[08:01:30] *** ipalreadytaken has joined #smartos
[08:02:47] *** densone has joined #smartos
[08:05:28] *** gutleib has quit IRC
[08:10:27] *** emocakes has joined #smartos
[08:21:57] *** texarcana has joined #smartos
[08:22:30] *** ikwildrp1pper is now known as ikwildrpepper
[08:28:51] *** emocakes has quit IRC
[08:30:41] *** alucardX has joined #smartos
[08:34:02] <alucardX> morning
[08:35:10] <Licenser> morning alucardX
[08:44:37] *** hasker has quit IRC
[08:46:23] *** bens1 has joined #smartos
[08:58:04] *** ipalreadytaken has quit IRC
[08:59:13] *** chh has joined #smartos
[09:01:25] *** emocakes has joined #smartos
[09:02:10] *** bens1 has quit IRC
[09:05:36] *** alcir has joined #smartos
[09:07:30] *** Azbruh has quit IRC
[09:09:08] *** Azbruh has joined #smartos
[09:16:50] *** KahnAu has joined #smartos
[09:19:14] *** densone has quit IRC
[09:20:31] <KahnAu> How to best edit /etc/system in the global zone?
[09:24:41] *** andoriyu_ has joined #smartos
[09:31:37] *** rc10 has joined #smartos
[09:32:03] <rc10> hi, is there Aoe client avaialble for SmartOS ? I have exported drives in linux box and want to discover in smartos
[09:32:19] <rc10> how to do ? Please help
[09:40:46] *** kschiess has joined #smartos
[09:40:57] <Alasdairr> I'm not aware of one. There is an iSCSI initiator though.
[09:47:27] <rc10> just wondering.. why aoe isnt devleoped in solaris ?
[09:47:37] *** des2 has quit IRC
[09:49:53] <Alasdairr> rc10: We've got good iSCSI support, iSCSI gives reasonable performance with good nics and a good network
[09:50:12] <Alasdairr> And then there's fibre channel support
[09:50:33] <Alasdairr> iSCSI is more flexible than AoE
[09:50:42] <rc10> but Aoe is at layer 2 and faster than aoe ?
[09:51:02] <rc10> no routing involved in aoe
[09:51:36] <rc10> http://www.ict.co.th/download/EDProductDescription.pdf
[09:53:02] <Alasdairr> rc10: If there was the demand for it I'm sure it would happen
[09:53:18] *** des2 has joined #smartos
[09:53:19] <Alasdairr> But I haven't personally seen much demand for it
[09:53:42] <Alasdairr> People who want to do storage over ethernet seem to be fairly happy with iSCSI, and those who want performance go for FCoE or FC
[09:54:11] <rc10> ok, can you pls point to Fcoe or FC docs for setting up ?
[09:54:41] <Alasdairr> FC needs special FC adapters and FC switches
[09:55:26] <rc10> yah, but setting up in solaris ? is there any doc
[09:55:48] <Alasdairr> http://hub.opensolaris.org/bin/view/Project+fcoe/
[09:56:34] <rc10> thnx
[09:57:02] <Alasdairr> A lot of the links may be broken
[09:57:19] <rc10> ok
[09:58:00] <Alasdairr> A lot of the technologies involved belong to illumos, which as an open source project doesn't have full time documentation writers
[09:58:01] <aszeszo> i remember seeing AoE driver for soalris
[09:58:13] <aszeszo> *Solaris
[09:58:26] <Alasdairr> http://download.alyseo.com/pub/partners/Coraid/Drivers/Solaris/aoe-guide.html
[09:58:27] <Alasdairr> that?
[09:58:57] <aszeszo> one I saw was provided directly by CORAID
[09:59:21] <Alasdairr> yeah
[10:01:08] *** KermitTheFragger has joined #smartos
[10:03:50] *** emocakes has quit IRC
[10:08:38] *** ipalreadytaken has joined #smartos
[10:09:12] *** chh has quit IRC
[10:10:39] *** chh has joined #smartos
[10:12:59] *** ipalreadytaken has quit IRC
[10:22:14] *** andoriyu_ has quit IRC
[10:23:32] *** JT-EC_ is now known as JT-EC
[10:30:11] *** stetho has joined #smartos
[10:35:27] *** papertigers has quit IRC
[10:37:53] *** papertigers has joined #smartos
[10:38:42] *** mamash has joined #smartos
[10:53:43] *** bens1 has joined #smartos
[11:05:42] *** ktkNA is now known as ktk
[11:06:05] *** defa has joined #smartos
[11:12:06] *** wramthun has quit IRC
[11:12:12] *** darjeeli_ has joined #smartos
[11:15:24] *** darjeeling has quit IRC
[11:16:18] *** defa has quit IRC
[11:16:27] *** defa has joined #smartos
[11:17:46] *** leecallen has quit IRC
[11:17:46] *** leecallen35 has joined #smartos
[11:19:49] *** KahnAu has left #smartos
[11:20:16] *** emocakes has joined #smartos
[11:20:23] *** wramthun has joined #smartos
[11:26:00] <defa> Does anyone know if it is possible to use smarmontools to monitor hard drive temperature? I compiled the package but reading fails.
[11:29:35] *** chh has quit IRC
[11:40:36] *** chh has joined #smartos
[11:54:31] *** emocakes has quit IRC
[12:14:21] *** bluezenix1 has joined #smartos
[12:23:22] *** robinsmidsrod has quit IRC
[12:24:09] *** nikolam has joined #smartos
[12:30:56] *** robinsmidsrod has joined #smartos
[12:33:33] *** sachinsharma has quit IRC
[12:36:39] *** rc10 has quit IRC
[12:43:07] *** noahmehl has joined #smartos
[12:50:27] *** rc10 has joined #smartos
[12:53:47] *** ira has joined #smartos
[13:08:26] *** rc101 has joined #smartos
[13:08:27] *** rc10 has quit IRC
[13:12:31] *** ascii` has joined #smartos
[13:16:13] <ascii`> Hello, I'm planning to migrate a XCP (Xen) cluster to smartos. I documented myself a little and like the thin hypervisor concept but especially the ability to use ZFS. At the moment these two Xeon machines have a local storage (4 SAS drives) and a PCI Express 120GB SSD disk used as read/write cache using bcache.
[13:18:06] <ascii`> Is it possible to use Smart OS with a local disk pool using ZFS L2ARC? And make the configuration permanent?
[13:30:04] <linuxprof> yes
[13:32:09] <nahamu_> zpool configuration always persists. adding log (write cache / SLOG) or cache (read cache / L2ARC) devices is a pool configuration change.
[13:33:17] *** nahamu_ is now known as nahamu
[13:35:36] *** jelmd has joined #smartos
[13:37:41] <ascii`> thanks linuxprof and nahamu_
[13:38:05] <ascii`> changes are written to some xml files on the usb key right?
[13:38:22] <ascii`> sorry for such newbie questions but this will be my first smartos deployment
[13:45:35] <linuxprof> nothing is written to the USB key
[13:45:43] <linuxprof> it's all on disk
[13:45:56] <linuxprof> the zfs configuration is stored withing zfs, so to speak
[13:46:27] <linuxprof> and /usbkey is really /zones/usbkey, so it's also stored on disk
[13:47:10] <linuxprof> ok, /usbkey is zones/usbkey since it's not mounted under /zones. to be precise. :)
[13:52:11] *** ira has quit IRC
[13:52:26] <ascii`> nice, this means that when i upgrade smartos changing the usb stick or by the live upgrade mechanism described on the wiki i don't have to migrate any configuration
[13:53:00] <linuxprof> correct
[13:53:03] <jperkin> yup, that's the point of this design
[13:54:41] *** denizr has joined #smartos
[13:54:46] *** marsell has joined #smartos
[13:55:39] <ascii`> i think that this will be a great improvement compared to our current debian+custom bcache kernel+mdadm setup, i'll need to figure out on how to replace some scripts and logic that runs on hw but could guess that the answer is to virtualize it
[14:13:15] *** enmand has joined #smartos
[14:15:58] *** kamilr has joined #smartos
[14:16:01] <kamilr> Hi there
[14:16:24] <kamilr> Why i can't boot centos from foyent datasets in ide hdd model ?
[14:16:32] <kamilr> i got kernel panic every time
[14:17:29] *** nikolam has quit IRC
[14:17:51] <kamilr> i got "VFS: cannot open root device vda1 or unknown-block (0,0)"
[14:18:17] <nahamu> kamilr: it's expecting the virtio driver
[14:18:43] <nahamu> looks like it has "/dev/vda" hardcoded somewhere
[14:18:48] <kamilr> i see
[14:19:06] <nahamu> in theory if all instances of /dev/vda were replaced with UUID based identification you could use ide emulation... but why would you want to?
[14:19:10] <yofuh> adding root=/dev/hda1 should help. but i would not expect it to give any advantage
[14:19:17] <nahamu> virtio will give better performance.
[14:19:32] <kamilr> i know i need it to compare it to virtio
[14:19:40] <nahamu> ah
[14:19:51] <kamilr> i know that virtio is better but i have to do some tests
[14:20:23] <nahamu> I would boot it virtio, see if you can figure out where all the hardcoding is (look at kernel command line and at /etc/fstab) and fix it up, then switch it to ide
[14:20:56] <nahamu> Or, depending on what you're trying to do, perhaps add a *second* drive to the VM that's the ide one and do the tests on the non-boot drive
[14:21:13] <kamilr> i did it
[14:21:23] <kamilr> i have added secod hdd in ide
[14:21:33] <kamilr> but it is also in fstab
[14:21:39] <kamilr> so i think i have to chande it
[14:21:56] <linuxprof> you could install it from scratch, not using a dataset
[14:22:04] <nahamu> or use a livecd
[14:22:14] <linuxprof> yupp, that too
[14:26:40] <pringlescan> I read the ZFS portion of the OpenSolaris bible, and I'm confused, are there features that didn't make it into SmartOS? Like the automatic failover when a spare is specified? I believe someone mentioned that wasn't implemented.
[14:27:46] <nahamu> pringlescan: did you add a spare to your pool and pull a different drive and nothing happened?
[14:28:18] *** enmand has quit IRC
[14:28:22] <nahamu> "zpool add zones spare <disk spec>"
[14:28:26] <pringlescan> nahamu, no I thought I was told the disadvantage to HBA in ZFS and a hardware RAID is that it didn't happen like I'd expect
[14:28:55] <pringlescan> nahamu, yeah, that's the line I saw in the book, which made it sound like I chose the behavior and the only human intervention that's required is to bring the failed disk back online and then remove the spare from the pool
[14:29:00] *** MerlinDMC has quit IRC
[14:30:06] *** emocakes has joined #smartos
[14:30:09] <nahamu> I think my line there was wrong.
[14:30:30] <nahamu> nevermind
[14:31:04] <nahamu> I would expect spares to still work the same way as they did in the OpenSolaris days.
[14:31:18] <pringlescan> your line was right as per book
[14:31:59] *** MerlinDMC has joined #smartos
[14:32:04] *** rc101 has quit IRC
[14:34:58] <pringlescan> Also, is L2ARC with SSDs and HDDs persistent across reboots?
[14:35:31] <nahamu> yes
[14:37:10] *** emocakes has quit IRC
[14:48:09] *** enmand has joined #smartos
[14:55:02] *** neophenix has joined #smartos
[14:55:54] *** stetho has quit IRC
[14:56:25] <pringlescan> if I add SSDs and HDDs will it do an L2ARC on it's own (If there's more than one of each drive type, 5x the same SSD 5x the same HDD)
[14:57:45] <nahamu> what do you mean by "it"?
[14:58:25] <nahamu> If you want a complicated pool layout you're much better off creating the zones pool yourself manually
[14:59:23] <pringlescan> so I can just boot up into no setup mode and make a zpool named zones and that'll make it work with SmartOS on the next boot? and I shouldn't make more than one zone, because it's not "supported", and I'm too new to go with an unsupported config? just making sure
[15:01:48] <yofuh> jperkin: any opinion about using pkgsrc on solaris 10?
[15:03:28] <jperkin> yofuh: go for it, most bits should be fine (at one point I had pkgsrc mostly working on SunOS 4.x), likely lagging parts will be sun studio and older gcc support, as there simply aren't many users left.
[15:04:10] <jperkin> and sparc, of course
[15:04:45] <jperkin> I have a sol9 VM I use for testing basic bootstrap functionality from time to time
[15:05:35] <pringlescan> hm, glad I re-read my server manual, one riser has 2x 8x PCI-E and the one marked for internal storage had 4x PCI-E ...
[15:06:45] <yofuh> jperkin: thx, customer just have a x89 s10 box here
[15:13:13] *** darjeeling has joined #smartos
[15:16:52] *** darjeeli_ has quit IRC
[15:19:18] *** papertigers has quit IRC
[15:20:25] *** papertigers has joined #smartos
[15:26:32] *** denizr has quit IRC
[15:27:12] <wesolows> L2ARC is not persistent. There is active work to make it so.
[15:27:53] <wesolows> automatic sparing on disk failure is known to work in SDC 7 using the same platform code that is in SmartOS; I know no reason it should not work.
[15:28:22] *** darjeeling has quit IRC
[15:29:15] <wesolows> automated creation of L2ARC with 2 or more distinct SSD types is an SDC only feature. You can use disklayout and mkzpool manually on SmartOS as nahamu described, but the setup script cannot distinguish storage device types.
[15:29:35] <wesolows> SmartOS setup script pool creation is just about the dumbest thing on earth.
[15:30:57] *** Azbruh has quit IRC
[15:30:58] *** texarcana has quit IRC
[15:30:59] *** Lumb_ has quit IRC
[15:31:00] *** Webhostbudd has quit IRC
[15:31:06] *** Teknix has quit IRC
[15:31:07] *** d[^_^]b has quit IRC
[15:31:08] *** scarcry has quit IRC
[15:31:17] *** hungryblank has quit IRC
[15:31:23] *** rodgort has quit IRC
[15:31:25] *** linuxprof has quit IRC
[15:31:30] *** halldorh has quit IRC
[15:31:30] *** Lumb has joined #smartos
[15:31:30] *** Azbruh has joined #smartos
[15:31:31] *** linuxprof has joined #smartos
[15:31:31] *** d[^_^]b has joined #smartos
[15:31:32] *** Webhostbudd has joined #smartos
[15:31:33] *** hungryblank has joined #smartos
[15:31:35] *** halldorh has joined #smartos
[15:31:35] *** rodgort has joined #smartos
[15:31:59] *** Teknix has joined #smartos
[15:36:05] *** scarcry has joined #smartos
[15:36:30] <pringlescan> wesolows, thanks, now, when we say "persistent" do we mean the data in there is invalidated and has to warm up on reboot, or, does it mean the configuration is gone on reboot?
[15:36:35] <pringlescan> in regards to l2arc
[15:36:44] *** denizr has joined #smartos
[15:41:02] *** mamash has left #smartos
[15:42:52] <nahamu> pringlescan: warm up on reboot
[15:43:51] <nahamu> https://www.illumos.org/issues/3525 <- illumos issue tracking making L2ARC contents persist across reboot.
[15:44:11] <pringlescan> oh I don't care about that, I can just do reboots at night and make a warmup script
[15:44:22] <pringlescan> that seems reasonable for the interim
[15:44:33] <wesolows> the data. the configuration persists.
[15:44:50] *** potatosalad has joined #smartos
[15:44:54] <wesolows> "warming up" often takes days. sometimes weeks.
[15:45:00] <wesolows> JFYI.
[15:45:22] <pringlescan> i have enough ram that I don't think it's an issue? but I do remember reading that.
[15:46:19] *** tonyarkles has joined #smartos
[15:46:27] <wesolows> there is a limit to how fast data can be read in, and how fast it can be written out. And I believe the writing out part is further limited by code.
[15:46:44] <wesolows> especially since you're presumably doing random reads off disk, which are themselves slow.
[15:47:12] <wesolows> it's probably not as bad now as it was 3 or 4 years ago, but it's probably still going to take quite a while to warm up after rebooting.
[15:47:14] <pringlescan> I guess I could make my pool entirely on the SSDS then, when the L2ARC is persistent, copy that over to my HDDs then re-add the SSDs as cache to that zpool?
[15:47:25] <wesolows> a pool on all SSDs has no L2ARC.
[15:47:29] <wesolows> nor does it need any.
[15:47:48] <pringlescan> I'd like to two SSD -> HDD, but I have enough space for now to work just on my SSDs
[15:47:57] <wesolows> and no, that's not how it works. L2ARC is not another copy of the pool; it's in its own format. They're not interchangeable.
[15:48:11] *** rbrown__ has joined #smartos
[15:48:38] <pringlescan> I meant, I won't run out of space on my SSDs for the next 6-18 months, and I'm hoping persistence is added by then.
[15:48:39] <wesolows> I'd suggest just putting something together and seeing how it works. There's no need to spend weeks figuring this out.
[15:49:09] <nahamu> pringlescan: you might not even need L2ARC for your workload.
[15:49:38] *** ipalreadytaken has joined #smartos
[15:49:43] <pringlescan> I could always just use my SSDs and add more when I run out of space and use spinning disks for snapshots and other backups… that might be fine.
[15:50:55] <pringlescan> I just want to set it and forget it for a while and not have to performance tune until I have my application to a point where I can do capacity planning.
[15:51:17] *** rbrown_ has quit IRC
[15:51:29] *** phus has quit IRC
[15:51:48] *** phus has joined #smartos
[15:52:06] <wesolows> if ease of achieving high random I/O performance is more important than cost, an all ssd pool is a no-brainer
[15:52:41] <wesolows> these tradeoffs are pretty well understood, and aren't even specific to ZFS.
[15:56:12] *** vsomes_ has joined #smartos
[15:56:19] *** rc10 has joined #smartos
[15:56:28] <pringlescan> wesolows, that summed it up well and reminded me I already did ample research and decided on that a while ago, i just got nervous with two pools not being explicitly supported; but it doesn't really matter for my purposes
[16:06:32] *** szaydel has joined #smartos
[16:08:43] *** Teknix has quit IRC
[16:08:59] *** Teknix has joined #smartos
[16:14:45] *** oozbooz has joined #smartos
[16:16:54] *** nefilim has joined #smartos
[16:17:56] *** szaydel has quit IRC
[16:20:10] *** szaydel has joined #smartos
[16:20:46] *** psanford has joined #smartos
[16:25:27] *** enmand has quit IRC
[16:27:53] *** hasker has joined #smartos
[16:30:08] *** enmand has joined #smartos
[16:30:42] *** jelmd has quit IRC
[16:41:20] *** szaydel has quit IRC
[16:42:51] *** szaydel has joined #smartos
[16:45:15] *** alucardX has quit IRC
[16:48:10] *** andoriyu_ has joined #smartos
[16:53:09] *** CarlosC has joined #smartos
[16:54:12] *** bluezenix has joined #smartos
[16:56:01] *** darjeeling has joined #smartos
[16:56:28] <wesolows> nothing is "explicitly supported" as SmartOS is not a product and we do not provide support of any kind for it. But everything in software does assume that there is a single pool and I doubt the case where there are more gets any significant testing.
[16:56:58] <wesolows> if I were you I would throw out the spinning disks, make a single pool on my ssds, and get back to work.
[16:58:12] *** bluezenix1 has quit IRC
[17:05:58] *** szaydel has quit IRC
[17:08:36] *** dap has joined #smartos
[17:08:47] *** rc10 has quit IRC
[17:14:52] <yofuh> there is (almost) no problem with running smartos with two pools and zones on both
[17:15:08] <yofuh> vmadm has a zpool propertie
[17:15:37] <yofuh> it is not ully implemented, thus some things (like destroying snapshots when deleting the zone) does not work
[17:16:17] <yofuh> and you have manially copy ove any image to the second pool, imgadm will not do it for you
[17:18:18] <rmustacc> The new imgadm I think has a flag to try and be helpful to you for that.
[17:19:21] <yofuh> i didn't had a look on the new so far
[17:19:31] <wesolows> the best thing it could really do to be helpful is to delete all but one pool.
[17:22:11] <nahamu> wesolows: and pop up a message telling you which of your hardware components should be replaced with better alternatives? :-P
[17:23:45] <yofuh> if your goal is to make the os useful only to you or people who have exaclty the same usecase, that's clearly the direction to go
[17:25:52] <nahamu> well, they're certainly only testing their own use cases.
[17:26:30] <nahamu> (unless people have added test cases to the test suite for other use cases)
[17:29:23] *** ryancnelson has joined #smartos
[17:29:25] *** bluezenix has quit IRC
[17:32:12] *** bluezenix has joined #smartos
[17:32:18] *** bbarker has joined #smartos
[17:34:41] *** andoriyu_ has quit IRC
[17:35:03] *** chh has quit IRC
[17:35:23] *** jaakkos has quit IRC
[17:36:17] *** Cpt-Oblivious has joined #smartos
[17:36:19] *** sachinsharma has joined #smartos
[17:37:12] *** jim80net has joined #smartos
[17:40:02] *** kschiess has quit IRC
[17:42:11] *** jaakkos has joined #smartos
[17:44:22] <pringlescan> wesolows, can I add more capacity to my pool later without degrading performance when compared to having all disks present at the initial pool creation?
[17:44:38] *** szaydel has joined #smartos
[17:45:44] <ryancnelson> sure, you can… modulo "how it was set up to begin with"
[17:46:25] <ryancnelson> it's straightforward enough to backup-restore, too...
[17:52:27] *** alcir has quit IRC
[17:54:06] *** szaydel has quit IRC
[17:55:21] *** sachinsharma has quit IRC
[17:56:33] *** ipalreadytaken has quit IRC
[18:00:55] *** andoriyu has joined #smartos
[18:04:16] *** ipalreadytaken has joined #smartos
[18:10:38] *** abnormal has joined #smartos
[18:11:56] *** andoriyu has quit IRC
[18:19:22] *** kamilr has quit IRC
[18:21:47] *** ryancnelson has quit IRC
[18:36:05] *** mamash has joined #smartos
[18:37:19] *** denizr has quit IRC
[18:40:37] *** wramthun has quit IRC
[18:42:35] *** wramthun has joined #smartos
[18:43:40] <wesolows> you can add more vdevs. you cannot extend a raidz or raidz2 vdev.
[18:45:48] <wesolows> yofuh: yes, we're interested in doing one thing and doing it well. The pooled storage model, especially with the flexibility of slogs and l2arc, makes it very rare for 2 pools to be sensible, and the purpose of our system is to be a hypervisor, not a hypervisor and a backup server.
[18:45:51] *** tru_tru has quit IRC
[18:46:06] <wesolows> It's your system, of course, so do whatever, but that's not how it's being designed and built.
[18:47:50] *** ryancnelson has joined #smartos
[18:51:07] *** bluezenix has quit IRC
[18:53:46] *** tru_tru has joined #smartos
[18:54:36] *** noahmehl_ has joined #smartos
[18:56:33] *** noahmehl has quit IRC
[18:56:34] *** noahmehl_ is now known as noahmehl
[18:58:18] *** eli-away_ has joined #smartos
[19:00:03] *** eli-away has quit IRC
[19:09:13] <e^ipi> yofuh: if you think the feature is incomplete, file specific bugs
[19:09:24] <e^ipi> it's not something we test.
[19:09:48] <yofuh> e^ipi: i guess i did
[19:09:53] *** denizr has joined #smartos
[19:14:19] *** szaydel has joined #smartos
[19:20:55] *** chh has joined #smartos
[19:21:32] *** chh_ has joined #smartos
[19:21:44] <yofuh> e^ipi: yes, ther is one of them https://github.com/joyent/smartos-live/issues/140 which appears to has been fixed, however, i found an other one wich i havn't reported jet (++todolist)
[19:23:14] <Licenser> out of curiosity, is it purpose or accident that kstate requires a tialing : now?
[19:25:19] <rmustacc> Licenser: Hmm?
[19:25:31] *** chh has quit IRC
[19:25:49] <Licenser> rmustacc in the latest smartos 'kstat -p zfs:0:arcstats' does no longer work it has to be 'kstat -p zfs:0:arcstats:' now
[19:26:00] <Licenser> not a big deal just curiose if it's a regression or a feature :)
[19:28:01] *** denizr has quit IRC
[19:28:31] *** denizr has joined #smartos
[19:29:40] *** KermitTheFragger has quit IRC
[19:31:39] <rmustacc> Probably regression.
[19:31:45] <rmustacc> Someone rewrote kstat in C in illumos.
[19:31:50] <rmustacc> I would file an illumos bug.
[19:32:53] <Licenser> :)
[19:38:54] *** quasi has quit IRC
[19:39:44] <brendang> I always use kstat -pn arcstats
[19:42:04] <Licenser> brendang but you know what you're doing ;) I don't!
[19:42:37] <Licenser> it proppably isn't a bit issue but I guess there might be more then my script out there that could break
[19:57:56] *** noahmehl has quit IRC
[20:02:25] *** bluezenix has joined #smartos
[20:03:24] *** khushildep has joined #smartos
[20:04:47] *** CarlosC has quit IRC
[20:05:37] *** mobidev has joined #smartos
[20:06:42] *** natefoo has quit IRC
[20:10:36] *** ryancnelson has quit IRC
[20:11:53] *** axonpoet has joined #smartos
[20:17:50] *** tonyarkles has quit IRC
[20:24:24] *** bens1 has quit IRC
[20:26:16] <ascii`> umh i have a failed vm i can't delete
[20:26:21] <ascii`> state failed
[20:26:57] <ascii`> and i can't stop it too
[20:27:54] *** khushildep has quit IRC
[20:29:13] <linuxprof> ascii`: what happens when you try to delete it?
[20:29:21] <ascii`> it hangs
[20:29:24] <yofuh> ascii`: you might run into https://github.com/joyent/smartos-live/issues/167 which can be wworkarrounden by stopping the zone (see zoneadm not vmadm) it it is not already stopped and remove the transition atribute from the zones config (zonename.zml in .etc.zones)
[20:29:43] <ascii`> now i have rebooted
[20:30:04] <ascii`> i was installing fifo
[20:30:13] <ascii`> vmadm create -f setupfifo.json
[20:30:37] <ascii`> but i never got the confirmation of creation okay
[20:30:59] *** wolfeida_ has joined #smartos
[20:31:01] *** tonyarkles has joined #smartos
[20:32:22] <yofuh> if my typo rate keeps increassing this way, i'll need assistence soon
[20:32:23] *** ipalreadytaken has quit IRC
[20:32:34] *** jelmd has joined #smartos
[20:32:35] <yofuh> let's hope it is the colding
[20:33:26] <ascii`> yofuh, while reading we have a great correction algorithm :)
[20:33:55] <ascii`> actually we don't read at all, just guess, as far i remember from cognitive psychology
[20:33:59] <yofuh> the shell is not doing that good, it will become more and more dangerous to give me root access
[20:34:39] *** wolfeidau has quit IRC
[20:36:23] <ascii`> yofuh, i remember a tool that was asking for the hostname of the server connected to, seems of little use but my colleague told me it saved him more than few times
[20:39:52] <ascii`> http://nopaste.dk/p22721
[20:40:23] <ascii`> strange, it has now rebooted, i still see it as failed (in the json the autostart attribute was active)
[20:40:38] <ascii`> and deleting it hangs again
[20:40:56] <ascii`> yofuh, i'm going to try your suggestion
[20:41:16] <ascii`> zoneadm list
[20:41:22] <ascii`> i only see global
[20:41:48] <pringlescan> is this the recommended way to run a node.js app using SMF? https://github.com/bahamas10/node-smf
[20:42:57] *** emocakes has joined #smartos
[20:43:34] <pringlescan> ah, I found it in the joyent wiki. I'm going to make a google custom site search with illumos, smartos and joyent
[20:43:48] <ascii`> okay, zoneadm list -c, i see the ID of the failed vm
[20:44:29] *** porkbelt has quit IRC
[20:45:04] *** porkbelt has joined #smartos
[20:45:25] <yofuh> ascii`: i would expect a zone to be deletable if in state failed, if not go with the deletion of the state attribute, however, i think it's a bug you might report
[20:45:25] <yofuh> pringlescan: i would not assume that a node.js app is any different than other applications, just create a manifest, it's up to you if you want to use a method script or start the application directly
[20:45:41] <ascii`> yofuh, zoneadm -z 0b2bc82f-2c48-4acd-b6fe-75bd4e6b5e32 halt hangs too
[20:45:55] <ascii`> is this the expected behavior?
[20:46:10] <ascii`> my version is joyent_20130222T000747Z
[20:47:00] *** vsomes_ has quit IRC
[20:47:08] <opeth_> hey. can /usbkey/config be reloaded without rebooting the image?
[20:47:15] *** th has left #smartos
[20:47:22] *** kamilr has joined #smartos
[20:47:24] <opeth_> perhaps by refreshing the smartdc-config service?
[20:47:40] <ascii`> for sure if needed i'll report, i just want to make sure that it's not a direct consequence of my newbness with smartos
[20:47:45] *** mobidev has quit IRC
[20:49:36] *** ahaydock has joined #smartos
[20:49:44] <nahamu> what are others doing to calculate the new hash when changing the root password in the GZ?
[20:49:44] <yofuh> ascii`: if the zone is in installed state (zoneadm list) zoneadm halt schould not hang, but it is not usefull to call it anyway since there is nothing to stop
[20:50:22] *** emocakes has quit IRC
[20:51:48] <ascii`> zoneadm list doesn't show it but vmadm list does
[20:51:49] <nahamu> hmmm, modifying the PAM stack on my linux machine to use sha256 instead of sha512 seems to generate $5$ hashes... good enough for me.
[20:52:07] <ascii`> zoneadm list -c shows it
[20:52:28] <yofuh> mamash: you could just use the common "passwd" tool
[20:52:55] <yofuh> ascii`: in what state ist it? (might run zneadm list -cv)
[20:53:43] <ascii`> - 0b2bc82f-2c48-4acd-b6fe-75bd4e6b5e32 installed /zones/0b2bc82f-2c48-4acd-b6fe-75bd4e6b5e32 joyent excl
[20:53:48] *** andoriyu has joined #smartos
[20:53:52] <ascii`> it says installed
[20:54:08] <yofuh> so it is stopped
[20:54:36] <ascii`> zoneadm -z 0b2bc82f-2c48-4acd-b6fe-75bd4e6b5e32 delete should work right?
[20:54:50] <yofuh> not really what you should try
[20:55:02] <ascii`> uninstall pardon
[20:55:08] <yofuh> zoneadm will not delete the zoneconfig not will it destroy the zfs
[20:55:15] <yofuh> so don't do that
[20:55:50] <yofuh> i would recommend to edit the zones config, remove the state attribute and may be call vmadm delete after
[20:56:42] <yofuh> once the state attribute is gone, vmadm will display the zone as stopped, it has no magic source of information, it just uses zoneadm and stores states by the evol force of adding attributes to the zones config
[20:58:09] <ascii`> this was the installation json http://nopaste.dk/p22723
[20:58:23] <ascii`> taken from the fifo wiki
[20:58:35] <ascii`> issues started from here
[20:59:04] <ascii`> perhaps i took a bad snapshot? it was the advised one on smartos website
[20:59:50] *** abnormal has quit IRC
[21:00:25] <ascii`> http://nopaste.dk/p22725 this is a cpuinfo
[21:00:57] *** tonyarkles has quit IRC
[21:01:38] <ascii`> Intel(R) Xeon(R) CPU L5420 @ 2.50GHz CPU 0
[21:01:38] <ascii`> Intel(R) Xeon(R) CPU L5420 @ 2.50GHz CPU 1
[21:01:42] <yofuh> ascii`: the config doesn't look suspicius, i would try to install it again and run "zlogin -e'#' -C $zonename" to see what's going on
[21:01:54] <Licenser> ascii` usually what you experienced should not happen when I remember right there are even fifo installations with your release somewhere
[21:01:57] <ascii`> http://ark.intel.com/products/33929/Intel-Xeon-Processor-L5420-12M-Cache-2_50-GHz-1333-MHz-FSB
[21:02:03] <ascii`> yes it's 64-bit
[21:02:25] <yofuh> it would not boot smartos if it would not
[21:03:55] <ascii`> i tough that perhaps userland was x86 compatible, so the cpu seems not the problem
[21:05:52] *** opeth_ has quit IRC
[21:06:56] *** opeth_ has joined #smartos
[21:07:17] *** Azbruh has quit IRC
[21:08:19] <ascii`> changed ip and vmadm create -f setupfifo.json hangs again, no disk activity
[21:08:46] <ascii`> i feel a little unlucky, zlogin -e'#' -C ea62331b-04db-45d5-a959-90d10bd7aca4 hangs too
[21:08:47] *** Azbruh has joined #smartos
[21:08:48] <Licenser> ascii` have you created any other VM's before?
[21:08:52] <ascii`> is there a way to debug this?
[21:09:01] <ascii`> Licenser, clean install from this afternoon
[21:09:09] <Licenser> you can look into the /var/log/vm/* I think
[21:10:34] <ascii`> Licenser, tailing it
[21:10:45] <ascii`> it hangs on /usr/sbin/zoneadm -u 79829319-ff9a-4101-8e4a-ecae7f43f54e boot -X
[21:11:03] <ascii`> changed ip again and launched vmadm create -f setupfifo.json
[21:12:00] <Licenser> very odd
[21:12:11] <ascii`> iostat is quiet
[21:12:11] <Licenser> the dataset is installed right?
[21:12:34] *** Wraithhh has left #smartos
[21:12:47] <ascii`> i have been asked for it during installation, and specified two disks space separated
[21:13:10] <Licenser> no I mean with imgadm
[21:13:34] <ascii`> sorry it's my first day with this technology, let me check
[21:14:05] <Licenser> no worries :) we all had our first day with this technology (except for rmustacc, brendang and e^ipi perhaps)
[21:14:07] <ascii`> fdea06b0-3f24-11e2-ac50-0b645575ce9d base64 1.8.4 smartos 2012-12-05T21:59:37Z
[21:15:07] <ascii`> so seems yes, the command was successful (it downloaded a base image from joyent as my understanding)
[21:15:19] <Licenser> yup exactly
[21:15:25] <Licenser> it's like a template for a VM
[21:16:16] <ascii`> if the usb image was somehow corrupted i would noticed it well before reching this point i guess
[21:17:04] *** noahmehl has joined #smartos
[21:17:24] <Licenser> propably yes
[21:18:17] <ascii`> okay got some info from the tail
[21:18:23] <ascii`> {"name":"vmadm","hostname":"00-24-81-b4-49-04","pid":3674,"action":"create","vm":"79829319-ff9a-4101-8e4a-ecae7f43f54e","level":40,"msg":"Marking VM 79829319-ff9a-4101-8e4a-ecae7f43f54e as a \"failure\" because we hit waitForProvisioning() timeout.","time":"2013-03-11T20:15:17.341Z","v":0}
[21:18:30] *** opeth_ has quit IRC
[21:18:53] *** wramthun has quit IRC
[21:18:59] <ascii`> last issued command by smartos itself was /usr/sbin/zoneadm -u 79829319-ff9a-4101-8e4a-ecae7f43f54e halt -X
[21:19:12] <Licenser> but that seems all prett sane
[21:19:12] <ascii`> the command on the console is still stuck
[21:19:25] <ascii`> eg: didn't exited
[21:19:31] <Licenser> and your version is recent
[21:19:31] *** vsomes_ has joined #smartos
[21:20:17] *** opeth_ has joined #smartos
[21:20:27] <ascii`> i can upload the whole log if it's useful
[21:20:35] *** wramthun has joined #smartos
[21:20:57] <Licenser> I don't think I could make much out of it :(
[21:22:23] *** wramthun has quit IRC
[21:22:57] <Licenser> hmmm ascii` one thing, have you adjusted the ip addresses to your network?
[21:23:04] <ascii`> well perhaps it's useful for the bug report http://nopaste.dk/p22733
[21:23:39] *** wramthun has joined #smartos
[21:23:52] <ascii`> yes i'm on the, configured it as a /24 but there is a virtual 254 gateway for every /24
[21:24:07] <Licenser> okay just making sure :)
[21:24:23] <ascii`> :)
[21:30:05] *** ahaydock has quit IRC
[21:38:16] *** enmand_ has joined #smartos
[21:38:52] *** Niamkik_ has joined #smartos
[21:39:23] *** chh_ has quit IRC
[21:39:59] *** chh has joined #smartos
[21:41:29] *** wesolows has quit IRC
[21:43:10] *** wesolows has joined #smartos
[21:44:12] *** chh has quit IRC
[21:47:08] *** chh has joined #smartos
[21:47:15] *** enmand has quit IRC
[21:47:23] *** Niamkik has quit IRC
[21:49:25] *** des2 has quit IRC
[21:53:41] *** chh has quit IRC
[22:13:43] *** ryancnelson has joined #smartos
[22:17:56] *** potatosalad1 has joined #smartos
[22:19:00] *** potatosalad has quit IRC
[22:19:29] *** bens1 has joined #smartos
[22:19:37] *** chh has joined #smartos
[22:20:12] *** alpharender has joined #smartos
[22:24:40] *** alpharender has left #smartos
[22:25:19] *** alpharender has joined #smartos
[22:27:47] *** chh has quit IRC
[22:27:48] <alpharender> Hi. I am trying to install smartos. I am at a question that now asks me to "select disks for storage pool, space seperated" it lists the disks as Valid choices are c0t500xxx…….. I have 3 jbod units of 49 disks…. I assume if I make ONE mistake typing the drive IDs; this is impractical. There is no automatic drive-pool selection mechanism?
[22:28:12] <alpharender> tab does not work
[22:28:21] *** potatosalad1 has quit IRC
[22:29:40] <rmustacc> There isn't currently in SmartOS.
[22:29:54] <rmustacc> There is in Joyent SDC.
[22:30:28] <alpharender> heh
[22:30:33] <rmustacc> You could always just go ahead and create the zpool yourself and import it.
[22:30:45] <rmustacc> All the tools to do it automatically are in SmartOS.
[22:30:52] <rmustacc> No one has just gone back and wired up the intial installer there.
[22:31:06] <alpharender> i see
[22:32:10] *** mamash has left #smartos
[22:32:33] *** wesolows has quit IRC
[22:32:55] *** tonyarkles has joined #smartos
[22:33:44] *** wesolows has joined #smartos
[22:33:57] *** Cpt_Oblivious has joined #smartos
[22:45:16] *** axonpoet has quit IRC
[22:45:52] <ryancnelson> so:
[22:45:56] <ryancnelson> boot "
[22:46:02] <ryancnelson> (sorry)
[22:46:56] <ryancnelson> boot "no-import" ... create your zones zpool , then download http://ryan.net/misc/zones.postsetup.zfs.gz ... then gzcat that | zfs recv zones
[22:47:10] <ryancnelson> ... you'll need to maybe use -F
[22:47:15] *** denizr has quit IRC
[22:47:23] <ryancnelson> ... that's the stub of the zfs filesystems, post-setup
[22:47:39] <jesse_> ryancnelson, one can't no longer run the installed after creating the zones pool?
[22:47:46] <ryancnelson> there's a config file in zones/usbkey, the password is "root"
[22:47:56] <ryancnelson> jesse_: , that's terrible english :)
[22:48:02] <ryancnelson> what?
[22:48:16] <jesse_> agreed=)
[22:48:16] <ryancnelson> oh. "installER"
[22:48:41] *** bluezenix has quit IRC
[22:48:43] <jesse_> create 'zones' in no-import/recovery boot
[22:48:49] <jesse_> then run installer
[22:49:01] <ryancnelson> you probably can, yes.
[22:49:06] <jesse_> select one (doesn't matter which) device and install
[22:49:17] <ryancnelson> probably just create the zones zpool in no-import mode, then reboot
[22:49:25] <jesse_> when it comes to creating the pool, the install script sees there is a pool named zones already
[22:49:33] <jesse_> and doesn't run zfs create
[22:50:31] <jesse_> I've done that with several boxes with existing pools, just renamed the pool zones in recovery boot (and made sure it didn't have fs called opt/var/etc.
[22:51:18] <ryancnelson> i just skipped it all, and saved my post-setup zfs send -R of zones
[22:51:41] <jesse_> might be better if the 'request devices' part would just tell the user 'found zones pool, using it'
[22:51:58] <ryancnelson> true that
[22:52:38] <jesse_> smartos installs seem to be one-off kind for many people
[22:52:45] <ryancnelson> or: if the installer ran inside "gnu screen", so we could flip to another window, generate a list with format and sed , and copy it to a buffer
[22:52:59] <jesse_> and they want to configure exotic zpool layouts
[22:53:22] *** des2 has joined #smartos
[22:53:42] <ryancnelson> nothing wrong with that.... but if you know what you want, you should know enough to make that happen, no?
[22:53:44] <ryancnelson> :)
[22:54:14] <jesse_> well, I just wrote how to make it happen ;)
[22:54:41] *** chh has joined #smartos
[22:55:18] <ascii`> stupid question, how to run the installer again?
[22:55:32] <jesse_> just reboot with the default selection?
[22:55:50] <jesse_> or, if you've alredy installed a pool, you'll need to delete a file
[22:55:59] <ryancnelson> you mean, "how do i destroy my disks and start over?"
[22:56:26] <jesse_> rm /zones/.system_pool
[22:56:38] <ascii`> ok
[22:56:39] <jesse_> and then zfs destroy the filesystems
[22:57:01] <jesse_> (or, just wipe the pool and create an empty zones pool, and see above=))
[22:57:31] *** bens1 has quit IRC
[22:58:36] *** deirdres has joined #smartos
[22:59:23] *** chh has quit IRC
[23:01:48] *** ChrisPartridge has joined #smartos
[23:04:49] *** ira has joined #smartos
[23:05:45] *** wolfeida_ has quit IRC
[23:07:35] *** neophenix has quit IRC
[23:10:17] <ascii`> jesse_, asked because the question was somehow connected :)
[23:17:46] *** leecallen has joined #smartos
[23:17:47] *** leecallen35 has quit IRC
[23:20:23] *** Cpt_Oblivious has joined #smartos
[23:23:47] *** Cpt-Oblivious has quit IRC
[23:24:18] *** Cpt_Oblivious is now known as Cpt-Oblivious
[23:26:04] *** ktk is now known as ktkNA
[23:26:33] *** wolfeidau has joined #smartos
[23:35:00] *** szaydel has quit IRC
[23:35:30] *** defa has quit IRC
[23:35:40] *** szaydel has joined #smartos
[23:39:06] *** alpharender_ has joined #smartos
[23:39:43] *** alpharender has quit IRC
[23:40:23] *** alpharender_ is now known as alpharender
[23:45:11] *** wesolows has quit IRC
[23:45:51] *** wesolows has joined #smartos
[23:50:12] *** gkyildirim has joined #smartos
[23:51:02] *** ira has quit IRC
[23:52:54] *** nikolam has joined #smartos
[23:55:11] *** chh has joined #smartos
[23:57:52] *** gkyildirim has left #smartos
[23:58:57] *** denizr has joined #smartos
[23:59:37] *** chh has quit IRC

   March 11, 2013  
< | 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 | >