December 23, 2012 
< | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | >
December 23, 2012 [00:31:23] *** potatosalad has joined #smartos[00:32:28] *** ryancnelson has quit IRC[00:51:44] *** scubasteve has quit IRC[00:55:44] *** ins0mnia has quit IRC[01:01:52] *** arekinath has quit IRC[01:02:03] *** arekinath has joined #smartos[01:02:06] *** arekinath has joined #smartos[01:05:49] *** vho has joined #smartos[01:32:15] *** deirdres has joined #smartos[01:43:28] *** miine has joined #smartos[01:46:45] *** wolstena has joined #smartos[01:55:08] *** vho has left #smartos[01:59:03] *** potatosalad has quit IRC[02:02:37] *** deirdres has quit IRC[02:23:52] *** potatosalad has joined #smartos[02:23:55] *** potatosalad has left #smartos[02:55:07] *** Vod has quit IRC[03:07:26] *** Vod has joined #smartos[03:09:29] *** ryancnelson has joined #smartos[03:14:20] *** Vod has quit IRC[03:20:46] *** miine__ has joined #smartos[03:22:02] *** tylerflint has joined #smartos[03:23:50] *** miine has quit IRC[03:23:51] *** miine__ is now known as miine[03:26:46] <tylerflint> aside from virtualization, are there any additional advantages to using smartos over openIndiana?[03:28:41] <miine> tylerflint: no rpool needed ...[03:29:16] <tylerflint> miine: cool[03:29:32] <miine> depends on your needs :-)[03:29:39] <tylerflint> miine: would you use smartos for a general purpose server over openIndiana?[03:29:56] * jamesd would not... too much is not in the globalzone...[03:30:10] <miine> tylerflint: you've to look at what you'll need. otherwise there is also omnios.[03:30:57] <miine> I'm using currently oi in production. But I think switching either to my "own" smartos or omnios. currently it's more like that I switch to omnios as it uses IPS...[03:31:57] <tylerflint> miine: agreed. I've mostly been in linux/bsd land. Stumbled across Joyent and started to open my eyes to illumos. That and I'm extremely interested in ZFS and Dtrace.[03:32:04] <miine> but I would also like to have the "no rpool" solution... . as I'm going to make my own distro it might be a "best of" smartos / omnios ... but I have very specific needs[03:32:20] <miine> don't forget zones and crossbow...[03:32:48] <tylerflint> at this point I'm trying to consume as much as I can and trying to understand which distro to use for specific tasks[03:33:33] <miine> but anyway: regardless what you choose - people are very committed to illumos. which is a healthy sign...[03:34:10] <tylerflint> from what I can tell, it seems smartos is mostly a distro for virtualization and hypervised kvm[03:34:20] <miine> oi is for desktop. if you don't need any gui - you can also use smartos or omnios ...[03:34:24] <tylerflint> but I wasn't sure if there were other specific advantages[03:34:49] <jamesd> smartos will probably have the best and fastest support of virtualization and dtrace fixes/features,[03:35:49] <miine> also smartos and omnios should boot via pxe very fine...[03:36:33] <tylerflint> I booted smartos to start poking around. It was pretty awesome.[03:36:48] <miine> but to start with: the distro doesn't matter (Besides the used package format). you'll have enough to learn about the solaris specific features all illumos distros have in common...[03:36:57] <tylerflint> do the other distros start with zones enabled with a global zone?[03:37:13] <jamesd> yes all the distros come with zones, you just have to create them[03:44:00] <tylerflint> so if you were going to run a LAMP stack (I know, you probably wouldn't. But if you were) which of the following approaches would you take:[03:44:21] <tylerflint> ...assuming smartos is already loaded and running...[03:44:54] <tylerflint> 1- hop into the global zone, and install Apache, Mysl, and PHP in the same[03:44:55] <tylerflint> or[03:45:11] <tylerflint> 2- create a zone for apache, a zone for mysql, and a zone for php[03:45:13] <miine> MySQL and Apache into their own zones[03:45:43] <wesolows> (2), no brainer[03:46:18] <miine> if you have PHP tasks which need more security and are "remote calls" into same/another Apache - put those into their own zone too...[03:46:42] <miine> tylerflint: let me guess - never used PHP?[03:46:59] <tylerflint> interesting. I've observed that paradigm shift in the documentation, and I was curious if it was actually being leveraged that way[03:47:32] <miine> PHP is a process bound to apache etc. . so those two will always run in the same zone...[03:47:40] <tylerflint> miine: just searching for a universal example[03:48:23] <miine> think as zones as service and security "containers". and split along those lines accordingly...[03:49:11] <tylerflint> that's quite a different approach to hosting than the linux world. I mean, containers are emerging in linux, and jails have been around in bsd for some time. Though, it's interesting to see illumos people using zones by default[03:49:24] <miine> you can also see zones as virtual machines - just without virtualization overhead...[03:50:11] *** jamesd has quit IRC[03:51:11] <miine> what should use *bsd or linux use else? they don't have zones as solaris... . zones are the most advanced and performant type of virtualization (if you don't need different os)[03:51:14] <tylerflint> I have to say, I haven't been this excited about an operating system in quite awhile[03:51:49] <miine> yep. had this experience 4 years ago :D[03:52:52] <miine> but to forewarn you - there *will* be a phase where you'll experience some downsides. mostly when you'll try to compile stuff...[03:53:33] <tylerflint> haha[03:53:48] <tylerflint> linkers have trouble eh?[03:53:57] <tylerflint> finding libs[03:55:11] <miine> mostly "linuxisms". posix compatible should compile, but some have their own way touting something posix compatible...[03:55:58] <tylerflint> ah[03:56:51] <miine> but most things will compile fine, maybe some patches here and there are needed... but I can speak only for those things I use...[03:57:14] <tylerflint> cool[03:57:29] *** jamesd has joined #smartos[03:57:58] <miine> one thing: those "repositories" for third party packages are "packaged use cases". read: if you'll need LDAP support and they didn't compile it in you'll have to build your own package... .[03:58:17] <miine> so having a working build/package system is essential...[03:59:16] <miine> BUT: you can look at the compiler arguments etc. how they did build their packages as a head start :-)[04:00:27] <miine> my advice: learn how to create a zone and how to use crossbow to connect the zone to the network.[04:00:38] <miine> next: setup a build/package system within a zone.[04:00:55] <miine> then learn about writing SMF service manifests.[04:01:43] <miine> if you'll want to be "complete": learn using a configuration managment system to setup the host / zones ...[04:02:12] <tylerflint> awesome[04:02:16] <tylerflint> you mean like chef?[04:02:31] <miine> i like puppet more :D[04:02:39] <tylerflint> that's cool[04:03:08] <tylerflint> what do you use to build/package custom binaries?[04:03:13] <tylerflint> fpm?[04:03:30] <miine> I use my own IPS packaging system: www.ipsit.de ..[04:04:28] <miine> as I didn't get my head around spec files ... OpenCSW uses their own (garn I think) - but they build the old SVR4 packages...[04:04:52] <tylerflint> cool[04:05:52] <miine> the trick about ipsit is: just compile and install as you would write shell files and make a ZFS snapshot before installing, let ipsit find out what's new (speak installed) and package it ....[04:06:47] <miine> this way it works for python / ruby etc. packages too which have their own system how to install things...[04:12:26] <ryancnelson> you know there's no IPS in smartos, right?[04:12:53] <miine> ryancnelson: exactly the most thing I dislike about smartos ...[04:14:18] <miine> as IPS is a god send. having installation scripts in packages is a "no go"...[04:14:57] *** wolstena has quit IRC[04:16:26] <ryancnelson> joyent took ips *out*. if you dislike it, you may be in the wrong place. IPS assumes that the global zone is keeping track of all the packages in the zones. joyent-branded zones are completely userland, and independent of the global.[04:16:41] <ryancnelson> which is why the global zone doesn't even get installed, it's just in ram[04:17:54] <miine> nope. each zone have their independent IPS stack. but global zones CAN install into non-global zones if needed...[04:18:06] <ryancnelson> not in smartos.[04:18:43] <miine> also: smartos has their own "brand". nothing hinders someone to add an "ips" brand...[04:19:19] <miine> and thats exactly what I'm going to do - or add the non-rpool stuff to omnios ...[04:19:25] <ryancnelson> once you do that, it won't be smartos anymore, so you've forked it.[04:19:43] <miine> of course :-)[04:20:04] <miine> as I'm not joyent I have different needs...[04:22:03] <miine> btw: to "took IPS in" you'll just have to install the pkg stuff somehow. the files don't care which packaging system installed them...[04:22:22] <ryancnelson> then i don't know what you mean by the "non-rpool stuff" … if you want a persistent global zone, you need a root pool. we don't have one, and don't want one, so we run from ramdisk. That lets us avoid dependencies inside the zones, which is why it's done that way.[04:23:50] <miine> what about building my "own global zone" which has everything pre-installed what I need?[04:24:04] <miine> I think thats exactly whats joyent is doing...[04:24:28] <wesolows> the GZ has as little as possible[04:25:44] <miine> I would like to have the thing I need in the global zone. as my own remote managment system, my puppet based configuration managment system (which creates the non-global zones), etc. ...[04:26:01] <miine> the only question is: is it as easy as with distribution constructor on oi?[04:26:13] <ryancnelson> yes. that's the only question.[04:26:24] <ryancnelson> enjoy your crazy holidy[04:27:11] *** ryancnelson has left #smartos[04:27:40] <jamesd> i think joyent would remove / from smartos if they could make it work ;-p[04:30:16] <miine> I will see what "magic" joyent applied to illumos. It's going to be interesting for sure...[04:31:01] <jamesd> checkout the bug reports that have been fixed[04:31:09] *** masked has quit IRC[04:33:18] <wesolows> we hate magic.[04:33:38] <wesolows> magic is lies, and lies always unravel when you most need things to work[04:34:17] <jamesd> any technology that is sufficently advanced looks like magic to mear mortals like us[04:35:33] <nahamu> EveryCity already made an IPS repo for installing in smartos zones.[04:36:06] <nahamu> http://wiki.smartos.org/display/DOC/3rd+Party+Software+Repos[05:59:01] *** tylerflint has left #smartos[06:24:01] *** cjones_ has joined #smartos[06:47:40] *** scubasteve has joined #smartos[06:52:03] *** scubasteve has quit IRC[07:13:40] *** Cpt-Oblivious has joined #smartos[07:23:20] *** jamesd has quit IRC[08:09:57] *** cjones_ has quit IRC[08:16:21] *** darjeeli_ has joined #smartos[08:17:14] *** darjeeling has quit IRC[09:23:21] *** texarcana has quit IRC[09:24:35] *** texarcana has joined #smartos[09:58:33] *** cjones_ has joined #smartos[10:00:08] *** bens1 has joined #smartos[10:03:07] *** cjones_ has quit IRC[10:41:52] *** lundh has joined #smartos[10:42:03] <lundh> hi[10:45:04] <lundh> anyone here that could help me getinto the basics? I'm reading the documentaion but not everything is obvious.[10:47:36] <lundh> For example: My network configuration today is as follows: 1 physical nic: standard interface get its IP from my ISP, i have one vlan (100) that I use internally with a static IP and I have one macvlan (virtual interface with own MAC) that also get the IP through DHCP from my ISP. I would liek my VMs to get their own IP from my ISP like this[10:59:29] *** rodgort` has quit IRC[11:07:37] *** rodgort has joined #smartos[11:47:11] *** cjones_ has joined #smartos[11:51:48] *** cjones_ has quit IRC[11:55:08] *** bens1 has quit IRC[12:13:33] *** Livid has joined #smartos[12:24:29] *** Livid has quit IRC[12:39:19] <lundh> does smartos support intel h77 and realtek 8111f? cant really find any information about it[12:45:25] <MerlinDMC> lundh, http://illumos.org/hcl/[12:46:32] <lundh> no mention of the f-tag there but the base 8111 is listed[12:46:59] <MerlinDMC> to be sure easiest would be to write an iso or usbkey and boot it[12:47:19] <MerlinDMC> just use the noimport/noinstall grub option and see if your hardware is running[12:47:53] <lundh> yeah, I guess I have to do that when I get back home :)[12:48:13] <MerlinDMC> H67N runs fine with a rtl 8168[12:49:29] <lundh> I'm staying positive, the hardware is pretty common and there are open source drivers available[12:50:27] <MerlinDMC> yeah ... but integrating drivers will require a platform rebuild ... all possible but you can't yust compile and "install" them cause of the liveimg[12:52:52] <lundh> I know, I hope someone already have added support :)[12:53:25] <lundh> I'm in deep enough water as it is wothout trying to integrate new drivers :)[12:54:56] <lundh> is it a very bad idea to give the admin-interface a public IP?[12:55:22] <MerlinDMC> depends on what you transmit there[12:55:47] <MerlinDMC> if all data is "secure" there why shouldn't that be a public ip?[12:56:20] <lundh> I pretty much only intend to have sshd-running running there[12:56:30] <lundh> the rest should be separated to a different vlan[13:06:00] *** ins0mnia_ has joined #smartos[13:15:38] <lundh> ok, next question. could I monitor the resource usage (banwidth, memory etc) in a nice way?[13:17:27] <konobi> admin is best on a private network[13:18:18] <konobi> lundh: yes, have a look at the cuddletech blog[13:18:38] <lundh> will do, thanks![13:19:58] <konobi> argh... stupid pulled muscles[13:20:28] <lundh> :([13:35:55] *** cjones_ has joined #smartos[13:40:07] *** cjones_ has quit IRC[13:50:15] *** linuxprofessor_ has joined #smartos[13:53:32] *** linuxprofessor has quit IRC[13:53:33] *** linuxprofessor_ is now known as linuxprofessor[14:41:14] *** saablover has joined #smartos[14:41:42] <saablover> hi all, when I try to create a smartos zone, it times out with this message[14:41:44] <saablover> vmadm create -f /tmp/zone1[14:41:44] <saablover> timed out waiting for /var/svc/provisioning to move for f08b3c76-aa13-4621-830b-e51e6c7037ba[14:41:58] <saablover> this is in virtualbox[14:43:35] <konobi> nopaste your zone1 contents[14:46:33] <konobi> saablover: ?[14:46:44] <saablover> just a second[14:48:04] <saablover> I'm using this example[14:48:09] <saablover> with my own local IP address[14:48:12] <saablover> http://wiki.smartos.org/display/DOC/How+to+create+a+zone+(+OS+virtualized+machine+)+in+SmartOS[14:48:57] <konobi> saablover: the contents of your /tmp/zone1 file will let us know more easily[14:49:17] <saablover> just a second, I rebooted my machine[14:49:23] <konobi> np[14:53:00] <saablover> local network is this: 192.168.2.X/24[15:05:12] <saablover> here is my config:[15:05:13] <saablover> http://pastebin.com/iAEnLMWS[15:07:50] <konobi> you have that dataset uuid imported?[15:07:56] <saablover> yes[15:08:24] <saablover> http://pastebin.com/8mibxv99[15:08:29] <konobi> and sysinfo is showing a NIC Name 'admin' right?[15:09:08] <saablover> http://pastebin.com/dKZSTdWC[15:10:40] <konobi> add "primary": "true" to your nic[15:10:53] <konobi> change dataset_uuid to image_uuid[15:11:18] <konobi> and below image_uuid add "max_physical_memory": 512,[15:11:20] <saablover> in the zone configuration ?[15:11:24] <konobi> yup[15:13:12] <konobi> then give that a try[15:14:08] <saablover> now I have this:[15:14:09] <saablover> http://pastebin.com/Z3KHv5FW[15:15:11] <konobi> give that a try[15:15:23] <saablover> provisioning dataset c830aef4-e1c5-48de-b31d-30920512e18b with brand joyent is not supported[15:17:47] <konobi> yeah, that dataset doesn't exist[15:18:20] <konobi> the one you showed earlier was '84cb7edc-3f22-11e2-8a2a-3f2a7b148699'[15:18:48] <saablover> yeah my fault[15:18:50] <saablover> sorry[15:20:34] <saablover> hmmm it stalls again[15:21:20] <konobi> virtualbox--[15:21:43] <saablover> yeah I will switch to vmware fusion I guess[15:22:02] <konobi> yeah, that's how most folks run smartos when virtualizing it[15:24:33] *** cjones_ has joined #smartos[15:28:59] *** cjones_ has quit IRC[15:30:41] *** saablover has left #smartos[15:40:55] *** scubasteve has joined #smartos[15:42:47] *** ins0mnia_ has quit IRC[15:43:18] *** jamesd has joined #smartos[15:55:09] <konobi> scubasteve: get sorted?[16:03:58] *** dysinger has joined #smartos[16:06:02] *** cjones_ has joined #smartos[16:06:19] *** cjones_ has joined #smartos[16:16:37] <xmerlin> does anyone tried fifo 0.3 beta on the latest smartos?[16:22:27] *** sebasp__ has quit IRC[16:27:46] *** newlix has joined #smartos[16:33:39] *** newlix__ has joined #smartos[16:36:49] *** newlix has quit IRC[16:39:40] *** Vod has joined #smartos[16:57:56] *** newlix__ has quit IRC[16:59:44] *** newlix has joined #smartos[17:00:28] *** newlix__ has joined #smartos[17:01:30] *** newlix_ has quit IRC[17:04:14] *** newlix has quit IRC[17:08:06] *** daniel_wu has quit IRC[17:08:36] *** daniel_wu has joined #smartos[17:08:50] *** newlix_ has joined #smartos[17:10:48] *** newlix__ has quit IRC[17:11:33] *** newlix__ has joined #smartos[17:17:42] *** newlix has joined #smartos[17:20:19] *** newlix_ has quit IRC[17:25:46] <lundh> I'm out of luck if the latest-usb-image cant find my network card right?[17:26:15] <wesolows> maybe. what is it? need its pci id from prtconf -dD and/or prtconf -v[17:26:37] *** tonyarkles has joined #smartos[17:28:27] <lundh> ok, I'll try[17:30:54] <lundh> how do I change keymap?[17:31:04] <lundh> have to get that sorted out first :x[17:33:28] <lundh> wesolows: pci1043,8505[17:34:08] <lundh> it does seem to map to a driver: rge[17:34:09] *** newlix has quit IRC[17:34:53] <konobi> lundh: dladm show-phys -m, is it shown?[17:35:35] <lundh> yes[17:35:59] <konobi> then it's found =0)[17:35:59] <richlowe> konobi: you can do that from the usbkey config too, no?[17:36:01] <richlowe> and also loadkeys.[17:36:14] <lundh> konobi: hehe, then I'll try the installer[17:36:20] <konobi> richlowe: i think so... but can't remember the config option[17:36:32] <konobi> lundh: just note the mac address =0)[17:36:51] <richlowe> probably on the wiki somewhere[17:37:16] <konobi> http://wiki.smartos.org/display/DOC/extra+configuration+options[17:37:30] <lundh> konobi: oki, this could turn out great, or catastrophic :)[17:38:49] <konobi> richlowe: was wondering how much work you thought this might be... https://gitorious.org/avahi/avahi/trees/master/avahi-autoipd[17:39:10] *** tonyarkles has quit IRC[17:39:16] <konobi> um... to port that is[17:39:18] <richlowe> My first reaction is "Isn't avahi the thing the JDS folks did something weird with?"[17:39:34] <konobi> yeah... but plain avahi compiles just fine[17:39:50] <konobi> autoipd is the peice that doesn't[17:39:52] <richlowe> second is: "Didn't carlsonj do this properly already?"[17:40:15] <richlowe> I could swear that was why the link-level conflict stuff got done.[17:40:38] <wesolows> well 1043,8505 is the subsystem vendor and device ID, which is not useful. but if dladm already shows it, this probably isn't your problem anyway.[17:40:53] <konobi> Ou est carlsonj?[17:41:20] <richlowe> konobi: otherwise, I only know enough to say "I think it's possible"[17:41:25] <richlowe> nothing about difficulty or method.[17:41:29] <konobi> kk[17:42:46] <lundh> couldnt read srap... at boot. what does that mean?[17:42:55] <konobi> richlowe: is james an irc lurker?[17:43:04] <konobi> lundh: nothing to worry about, it's just a warning[17:43:06] <richlowe> not that I know of[17:43:16] <konobi> kk[17:43:46] <konobi> i'll drop him an email[17:44:03] <richlowe> well, wait a sec, I'm still searching. :)[17:45:35] <richlowe> at some point, I should run swish-e over this crap or something :\[17:46:19] <konobi> ?[17:46:36] <konobi> which "crap" ?[17:46:37] <lundh> how do I manually configure zfs? I have three block devices connected: 1 ssd and 2 4TB-drives. I want to mirror the 4 TB disks and use the SSD as l2arc[17:47:12] <richlowe> konobi: so, as best as I can still discover, it looks like it may be a case of "If we write it into this spec, surely someone else will do the work"[17:47:23] <konobi> heh[17:48:04] <wesolows> the only way to do that is to create the pool yourself[17:48:13] <lundh> ok[17:48:26] <wesolows> even the sdc pool layout stuff won't do what you want, because that's a bizarre configuration[17:49:00] <lundh> the default for two block devices is a mirror right?[17:49:03] <konobi> or just setup on disk, then do the zpool stuff after[17:49:12] <wesolows> what do you mean by default?[17:49:14] <konobi> *one[17:49:31] <wesolows> yes, that's also an option[17:49:39] <lundh> if I install smartos it will ask me to choose wht disks to use[17:49:47] <konobi> yup[17:49:53] *** dysinger has quit IRC[17:49:59] <lundh> those get combined to a zpool, how do I say how?[17:50:05] <lundh> what are the defaults?[17:50:09] <konobi> just use one for now[17:50:19] <lundh> one drive?[17:50:21] <lundh> k[17:50:21] <konobi> (one of your spindles)[17:50:27] <lundh> yeah[17:50:30] <wesolows> you don't. there's no way to specify topology. so use one, then add the other part of the mirror and the cache device later[17:50:42] <wesolows> (this trick doesn't work if you want raidz for example)[17:50:54] *** dysinger has joined #smartos[17:50:57] <lundh> I want a mirror[17:51:09] <wesolows> so tell it to use one of the spinning disks[17:51:14] <lundh> yes[17:51:46] <lundh> how do I restart the installer?[17:51:55] <konobi> reboot[17:52:04] <lundh> I got it installed the wrong way (asked after the fact...)[17:53:09] <lundh> that only works if the disks are unformatted. mine have the system already[17:53:28] <lundh> since I got my zpool set up but in a bad way[17:53:33] <konobi> you can choose the noimport option from grub and then destroy the zpool[17:56:10] <lundh> I can try that :)[17:56:24] <lundh> this is all very new to me. coming from freebsd[17:57:07] *** ins0mnia_ has joined #smartos[17:58:31] <konobi> well, smartos ain't yer normal distro[17:58:58] <lundh> I have realised that. spent most of the day today and yesterday reading up on it :)[18:17:53] *** sebasp__ has joined #smartos[18:20:02] *** daniel_wu has quit IRC[18:21:30] *** daniel_wu has joined #smartos[18:27:58] <lundh> I teied to add a new disk using the following command:[18:28:11] *** deirdres has joined #smartos[18:28:16] <lundh> zpool attach zones mirror c1t1d0 c1d4d0[18:28:24] <lundh> but it didnt work?[18:28:31] <lundh> too many arguments[18:28:35] <lundh> what am I missing?[18:30:55] <lundh> ah[18:32:56] *** tonyarkles has joined #smartos[18:33:01] <konobi> did you add the single disk via the installer?[18:33:05] <lundh> yes[18:33:15] <lundh> and then the other one after install[18:33:19] *** tonyarkles has quit IRC[18:33:23] <xmerlin> lundh, zpool add zones mirror c1t1d0 c1f4d0[18:33:28] <konobi> so you just need to specify the empty spindle[18:34:32] <xmerlin> lundh, I've assumed you want to add a couple of disks in mirror to your zpool[18:35:12] <lundh> I did a: zpool attach zones c1t1d0 c1t4d0[18:35:49] <lundh> seems like that worked. the mirror ended up being called mirror-0 for som reason though[18:38:15] *** ins0mnia_ has quit IRC[18:40:06] <lundh> did not work, lost the entire array[18:40:39] <konobi> zpool status?[18:41:10] <lundh> ended up in an infinit loop with the installer[18:41:15] <lundh> trying to fix that right now[18:41:29] <konobi> how did you end up like that in the installer?[18:41:50] <konobi> run the full install on a single disk... reboot and THEN add the mirror[18:42:27] <lundh> yeah, I'm on the live UDB trying to destroy all zfs that I have :)[18:45:13] <konobi> remember, smartos doesn't install to disk... it's just setting stuff up[18:45:15] *** scubasteve has quit IRC[18:45:25] <lundh> I know :)[18:47:18] <lundh> how do I destroy a zpool that is bad enugh not to be importable?[18:48:17] <konobi> zpool import -f zones; zpool destroy zones[18:48:30] <lundh> did that, not even -f worked[18:49:25] <konobi> what did it say?[18:49:42] <lundh> one or more devices is currently unavailable[18:50:04] <konobi> and you booted with the noimport=true option?[18:50:51] <lundh> no, but I booted into the noinstall-environment[18:51:13] <konobi> you need the noimport=true option[18:51:54] <lundh> are you sure? I was able to import and destroy fron this environment before[18:52:04] <lundh> this seems like a zfs bug[18:52:04] <konobi> yup[18:52:09] <lundh> ok[18:52:24] <lundh> I'll do as you way :)[18:53:17] <lundh> heh, we were talking about the same thing :) the boot option I used did infact set that flag[18:54:11] <konobi> and import doesn't work[18:54:14] <konobi> huh[18:54:17] <lundh> no[18:54:17] <wesolows> it is possible you have hit OS-1745. While it's possible to examine the pool to determine whether that's the case, it's not really that interesting. Regardless, to fix it, create a pool with all the devices you want, then destroy it.[18:54:43] *** scubasteve has joined #smartos[18:54:50] <konobi> you can also overwrite the first and last megabyte of the disk[18:55:13] <wesolows> yeah. I wish dd took a negative argument to skip, to skip backward from the end[18:56:46] <lundh> would be nice[18:57:15] <lundh> created a new one instead[18:58:14] <xmerlin> ehm ...I've found a smartos bug ;)[18:58:25] <konobi> lies![18:59:27] <xmerlin> I've putted a wrong ip inside a nick ...and vmadm said "timed out waiting for /var/svc/provisioning to move for 46bab787-37d8-4dab-a21a-f8e58b0466ce"[18:59:41] <xmerlin> wrong ip ....xxx.xxx.xxx.264[18:59:43] <konobi> wesolows: i wanted to play with ip4ll, so that i could experiment with kicking up a node based web interface for promptconfig[18:59:46] <xmerlin> bad cut and past[19:00:12] <konobi> xmerlin: add an issue on smartos-live[19:01:01] <xmerlin> ok[19:06:36] <lundh> ok, I think I got the zfs setup done now :)[19:07:10] <lundh> state: online, scan: reslivered 4.01G in 0h0m with 0 errors[19:07:34] <lundh> the config does however say mirror-0 instead of just mirror, any idea why?[19:07:52] <konobi> you can have more than one mirror per zpool[19:08:21] <lundh> doh :)[19:08:37] <lundh> now, how do I get my public key into the host?[19:09:48] <konobi> see the extra config doc link[19:10:02] <lundh> yep[19:14:08] <lundh> I might be blind...[19:14:23] <konobi> authorize_keys[19:14:31] <konobi> *authorized_keys[19:14:36] <lundh> yeah[19:17:56] *** wolstena has joined #smartos[19:19:24] <lundh> nice, got it working :)[19:21:50] *** ins0mnia has joined #smartos[19:29:56] <lundh> any best practices that I shuld follow?[19:30:07] *** bsdguru has joined #smartos[19:30:11] <lundh> where to save configurations etc[19:30:26] <bsdguru> any clues on how to track down: Dec 23 12:09:03 boxlet01 genunix: [ID 702911 kern.notice] bash[17540]: missing privilege "file_dac_search" (euid = 1032, syscall = 59) needed at zfs_zaccess+0x1fc[19:33:17] <konobi> well, /opt is persistent[19:33:39] <lundh> oki :)[19:33:47] <konobi> bsdguru: rbac?[19:34:00] <lundh> vimrc can be placed where? Iäm not really comfortable with the standard vim setup[19:34:25] <wesolows> it's generally wise to torch everything in root's home directory in the GZ[19:34:46] <Licenser> xmerlin to answer your question, not me but it should be working non the less[19:34:56] <lundh> wesolows: huh?[19:34:59] <wesolows> replace it with your own, or just leave it empty since there's very little you ought to be doing in the GZ anyway[19:35:13] <wesolows> lundh: all the dotfiles, including .vimrc[19:35:19] <konobi> lundh: or sshfs![19:35:42] <lundh> wesolows: isnt /root a ramdisk?[19:35:54] <lundh> konobi: thats not a bad idea[19:35:58] <wesolows> how often are you going to reboot?[19:36:14] <lundh> it will happen, it is a home server after all[19:36:30] <wesolows> then create a persistent service that unpacks sane dotfiles in there on bootup[19:36:48] <wesolows> /opt/custom/smf holds manifests for such things (I think that's true on SmartOS too)[19:36:59] <konobi> yup, it is[19:37:28] <lundh> just reate the folder and then the service in there?[19:37:44] <konobi> yup, smf manifest file[19:37:57] <lundh> yeah, I have seen them but nit studied them so far[19:37:59] <konobi> heh, Josh is up at at it early on a sunday again[19:38:00] <wesolows> you need to put the SMF manifest there. that manifest should have a start method that does whatever you want[19:38:02] <lundh> guess I have to do that[19:38:17] <konobi> smfgen is a decent start[19:38:26] <lundh> :)[19:38:30] <wesolows> although, seriously, this isn't that important. the number of things you should be doing in the GZ anyway is negligible.[19:38:38] <konobi> https://github.com/davepacheco/smfgen[19:38:57] <lundh> wesolows: I'll leave it as it is for now[19:39:06] <konobi> yup, i mainly do stuff in zones[19:39:10] <konobi> bsdguru: sorted?[19:39:31] <wesolows> I believe the only essential commands are rm .vimrc, vi /tmp.whatever.json, vmadm create[19:39:43] <lundh> my plan i so use this machine for TimeMachine backups, serving files to my HTPC and web dev. should be fine right?[19:39:54] <wesolows> sigh[19:40:43] <konobi> yup, that's what i do from one of my zones[19:40:56] <lundh> wesolows: sigh? what did I do?[19:41:10] <wesolows> nothing. it's fine.[19:41:22] <lundh> I dont belive you :)[19:41:24] <wesolows> for whatever reason I thought time machine used iscsi.[19:41:31] <lundh> konobi: how do you set up the zone?[19:41:32] <lundh> ah[19:41:41] <konobi> wesolows: nope, there's a few options[19:41:42] <lundh> konobi: especially the file server part[19:41:53] <konobi> lundh: check the wiki[19:42:01] <lundh> oki[19:42:02] <konobi> wesolows: cifs, afp, etc.[19:42:08] <konobi> nfs too iirc[19:42:16] <wesolows> ok. I'm just sad that so many people want SmartOS to be a storage server, which is probably the thing it's worst at. And that happens to be the product we made at our last gig.[19:43:27] <konobi> well, i think it's mainly just that folks have one server, they want to do VM stuff with it... but also want to do storage on zfs[19:43:40] <lundh> wesolows: well, it might be the thing it is worst at but if I want a stable host for VMs and ZFS for the storage then there really is not much of a choise[19:44:04] <wesolows> like I said... it's fine.[19:44:14] <xmerlin> Licenser, ok[19:44:33] <konobi> mayor of cranky solaris town =0)[19:44:47] <xmerlin> I've found also a bug in the percona image ...what's the right place to file it?[19:45:13] *** deirdres has quit IRC[19:46:17] <Licenser> xmerlin actually I'll go test it right now :)[19:49:19] <konobi> wesolows: get the last project out the door all nice 'n smooth?[19:49:28] <xmerlin> innobackupex wants /mysql/bin/xtrabackup but you provide only /mysql/bin/xtrabackup_55[19:49:53] <xmerlin> no I've to go ... bye bye[19:49:53] <konobi> xmerlin: ping mamash[19:49:56] <xmerlin> now[19:50:18] <xmerlin> ok ...when I come back[19:50:20] <xmerlin> bye[19:50:29] *** xmerlin has quit IRC[19:51:14] <bsdguru> konobi: snv_121[19:51:40] <bsdguru> konobi: snv_67 -> snv_121 inside the zones seem to have issues[19:52:01] <konobi> bsdguru: well, it's an rbac privilege[19:52:15] <konobi> so you might need to add it to the user[19:52:24] <bsdguru> # su - joe[19:52:44] <bsdguru> ./etc/mnttab: open: Permission denied[19:52:49] <konobi> usermod -K defaultpriv=basic,file_dac_read,file_dac_search joe[19:52:52] *** scubasteve has quit IRC[19:53:17] <bsdguru> [root@bee ~]# usermod -K defaultpriv=basic,file_dac_read,file_dac_search joe[19:53:17] <bsdguru> UX: usermod: ERROR: joe is not a local user.[19:53:22] <bsdguru> on shared the users are in ldap[19:53:34] <konobi> oh gawd[19:53:46] <bsdguru> yip[19:54:25] <bsdguru> and it's snv_67 --> snv_121 migrations[19:54:26] <bsdguru> I'm pondering rejumping a couple of globals[19:54:28] <lundh> how long should vmadm create < zone.json take?[19:55:30] <konobi> ou=SolarisAuthAttr[19:55:42] <konobi> "The auth_attr table should also be set to be searched in ldap in /etc/nsswitch.conf"[19:55:51] <bsdguru> we only have ou's for group and users in ldap[19:55:57] <bsdguru> it's one of those legacy setups[19:56:04] <konobi> lundh: vmadm create -f zone.json[19:56:24] <konobi> bsdguru: and you're supposed to be the ldap expert![19:56:26] <konobi> =0P[19:56:43] <bsdguru> not the solaris ldap expert[19:57:08] <konobi> yeah, hit up #illumos... i'm well out of my depth[19:57:36] <bsdguru> I've done ldap implementations for other internal clients[19:57:55] <bsdguru> but none of them have this weird ldap setup that shared got[20:02:29] <bsdguru> looking back at the snv_57 -> snv_121 and snv_67 -> snv_121 notes -- this problem was not hit in aug 2011[20:02:52] <konobi> bsdguru: could always drop ben an email[20:04:31] <bsdguru> I have a jira open for him to dig[20:06:29] <konobi> well, if anyone'll know, it'll be him[20:09:55] *** newlix has joined #smartos[20:12:54] *** newlix__ has quit IRC[20:16:05] *** Vod has quit IRC[20:38:49] <AlainODea> Will the global zone attempt STP or similar protocols over its "admin" NIC?[20:38:56] <AlainODea> by default.[20:39:31] <konobi> nope[20:40:00] <AlainODea> konobi: Thank you. I am troubleshooting an inability to get a 10GbE NIC working. It's probably cabling, but I've been assured that it is cabled as documented. The joys of troubleshooting blind :)[20:40:54] <konobi> AlainODea: dladm show-linkprop should give you a status indication[20:41:50] <AlainODea> konobi: Perfect! I'll try that now.[20:45:02] <AlainODea> konobi: lots of output to consider here. The man page is a good reference for this right?[20:45:19] <konobi> yup[20:45:26] <konobi> show-link might also be good[20:46:23] *** Vod has joined #smartos[20:47:04] <AlainODea> konobi: all four GbE NICs show unknown as do both 10GbE NICs[20:47:25] *** rancor has joined #smartos[20:47:49] *** rancor has quit IRC[20:47:58] <konobi> which nic type?[20:49:03] <AlainODea> konobi: 10GbE is Intel(R) Ethernet 10G 2P X520 Adapter[20:49:21] <konobi> try ifconfig plumb -a and then check the statuses again[20:49:52] <AlainODea> konobi: GbE is onboard Dell R720xd, whatever they use[20:50:13] <AlainODea> konobi: I'll try ifconfig plumb -a now now :)[20:51:10] <AlainODea> konobi: I get "ifconfig: -a: bad address"[20:51:25] <konobi> oh, try without the -a[20:51:46] <Alasdairrr> AlainODea: I had an issue with an X520 where if I plumbed it up "post boot" with an SFP+ optic in, it worked fine, but if I booted it with the optic plugged in, there were loads of boot errors and the device didn't show up via dladm[20:51:56] <Alasdairrr> very strange[20:52:03] <Alasdairrr> gave up and used 1Gig for now[20:54:35] <AlainODea> konobi: without -a I get "ifconfig: status: SIOCGLIFFLAGS: plumb: no such interface"[20:54:59] *** jamesd has quit IRC[20:55:00] <konobi> huh[20:56:00] <AlainODea> Alasdairrr: good to know. I may have to go that route. Without 10GbE the possibility of using SmartOS beyond research becomes slim.[20:56:29] <konobi> definitely used 10 and 40 gig links on smartos[20:56:55] <AlainODea> konobi: the NICs show link up 10Gbps in iDRAC[20:57:34] <AlainODea> konobi: iDRAC also shows an OS driver state warning on slot 2 of the 10GbE card[20:58:12] <Alasdairrr> AlainODea: Sorry I wasn't very clear there, if I booted the machine with the optic plugged in already, it wouldn't work, but if i booted with no optic, and pushed the optic in post-boot, the adapter worked absolutely fine. It may have been an issue with the optics we were using (Dell and ProLabs fake-juniper), didn't try with an official Intel SFP+ optic[20:58:29] <Alasdairrr> Also intend to try with a SFP+ direct attach cable at some point[20:59:09] <Alasdairrr> Didn't have much time to diagnose it sadly[20:59:49] <konobi> used it with a dell, intel cards... specific version i'm not sure[20:59:50] <AlainODea> Alasdairr: okay. That's a problem since it would requires hands and feet for reboots. We're using Cisco branded Twinax cables[21:01:12] <AlainODea> konobi: it could be a weird BIOS configuration. The R720xd's have the Lifecycle Controller which has some strange options for OS-specific driver deployment. It's possible it may have been bastardized before it was given to me for research.[21:01:53] <konobi> ugh[21:02:12] <konobi> the r11 and r12 stuff can be a pain[21:02:26] <wesolows> and because it's Dell, anywhere you see "OS" you have to read "Windows"[21:02:40] <konobi> AlainODea: though you can update all the firmware from there[21:03:22] <konobi> the R210s gave me no end of trouble[21:03:34] <AlainODea> konobi: I'll give that a try. Thank you again for your help :)[21:03:40] <konobi> which reminds me that I need to ask bryan for that IPKVM at some point[21:03:49] <konobi> office space time![21:04:21] <konobi> wesolows: how's the sea creature looking?[21:05:58] <wesolows> it's kicking ass and eating other sea creatures for breakfast![21:06:07] <konobi> excellent[21:07:21] <konobi> wesolows: any idea if that's getting hooked up with images?[21:07:44] <wesolows> I think so but I'm not following the details closely enough.[21:07:54] <konobi> epic[21:08:14] <wesolows> need to get imgapi working first, think Trent's working on that[21:08:56] <konobi> i'll have to bug him, since i was going to make a patch to imgadm to move image datasets under a child[21:11:19] <MerlinDMC> "imgapi" ... so a new version of the dsapi?[21:11:27] <konobi> but no doubt everyone's off till the new year[21:11:41] <konobi> MerlinDMC: we couldn't possibly talk about it[21:12:47] <MerlinDMC> i stopped working on imgadm the day rmustacc said smth about a rewrite of it ;)[21:20:24] <konobi> MerlinDMC: you can always add tests[21:21:06] <MerlinDMC> yeah but i have other things on my lists that can get infront of imgadm[21:21:39] <konobi> initial_script support would be nice[21:22:03] <MerlinDMC> i'm on qemu for adding ipv6 address autoconfiguration for guests[21:22:21] <MerlinDMC> but i think i hate ipv6 a little right now ^^[21:22:38] <konobi> who doesn't[21:22:44] <linuxprofessor> wouldn't it be better just to skip ipv4? :-)[21:23:31] <linuxprofessor> but on the other hand, my isp doesn't even offer me ipv6. have to use tunnel.[21:24:12] <konobi> until ISPs start supporting it at the client end, there's little benefit[21:24:54] <linuxprofessor> i don't have to use nat, port forwarding, etc.[21:25:00] <linuxprofessor> me like[21:25:04] <MerlinDMC> we have native ipv6 here ... but i can't use that until we get another gateway inhouse :P[21:25:18] *** tonyarkles has joined #smartos[21:25:51] <konobi> linuxprofessor: until you have to go to ipv4[21:26:13] <linuxprofessor> if everyone else just started using v6 i wouldnt have to[21:26:26] <konobi> catch 22[21:26:33] <linuxprofessor> sad panda[21:27:55] <linuxprofessor> think ill have to install openvms to cheer myself up[21:27:57] <konobi> and ipv4 over ipv6 is still messed[21:30:37] <LeftWing> linuxprofessor: I don't think anybody has ever said that before.[21:33:25] <Zigara> haha ^[21:34:50] <linuxprofessor> trying out charon-axp[21:34:58] <konobi> LeftWing: what you up to these days?[21:35:21] <LeftWing> Trying to remember how FFTs work at the moment.[21:35:28] <linuxprofessor> my fascination for openvms never stops to amaze me[21:36:08] <LeftWing> konobi: How is post-Joyent life treating you?[21:36:15] <konobi> good[21:36:56] <konobi> will need to chase up more business in the new year, but looking good[21:37:11] <LeftWing> Such is consulting. :)[21:39:24] *** Guest58651 is now known as Red_Devil[21:40:35] <konobi> more time to do fun stuff though[21:45:44] <konobi> well... fun non-work stuff ^_^[21:47:21] <linuxprofessor> installing openvms?[21:49:10] <konobi> product-ey idea and shoving xbmc into chrome[21:49:42] <linuxprofessor> doesnt sound as that much fun tbh =)[21:50:40] <LeftWing> konobi: chrome the browser or the OS ?[21:51:07] <konobi> browser[21:51:38] <LeftWing> I am, perhaps, missing the synergies there? :P[21:51:49] <konobi> (working along appjs)[21:54:02] <konobi> LeftWing: think html5 media element[21:54:51] <LeftWing> So just the XBMC backend to play videos, rather than the entire XBMC UI?[21:55:10] <konobi> yup[21:55:21] <konobi> there's now libxbmc =0)[21:55:24] <LeftWing> Ah, now it makes sense. :P[21:56:49] <linuxprofessor> did you figure out a way to solve "timed out waiting for /var/svc/provisioning to move for 7a581406-240c-412c-8970-0ba4d6afd6b4"?[21:56:55] <linuxprofessor> wasnt that discussed earlier?[21:57:32] <konobi> virtualbox one?[21:57:47] <linuxprofessor> running smartos on bare metal[21:58:21] <linuxprofessor> been happening with zones and kvm vms[21:59:00] <lundh> whats is the best way to add a zfs volume to a zone?[22:00:05] <konobi> lundh: see vmadm and delegate_dataset[22:00:16] <lundh> thank you[22:00:19] <konobi> linuxprofessor: what was the json config?[22:01:24] *** Cpt-Oblivious has quit IRC[22:02:44] <lundh> konobi: can only be done at create time?[22:03:08] <linuxprofessor> http://pastebin.com/h02pbFt8[22:03:22] <konobi> lundh: odd[22:03:35] <lundh> according to the docs anyway[22:04:00] <konobi> linuxprofessor: and you have the dataset/image_uuid imported?[22:05:53] <linuxprofessor> yupp[22:07:48] *** Webhostbudd has joined #smartos[22:08:52] <konobi> oh, remove the second nic[22:15:42] <konobi> linuxprofessor: that work?[22:17:18] <lundh> konobi: is delegate_dataset documented anywhere?[22:17:30] <lundh> the manual for vmadm is not that verbose[22:17:45] <konobi> it's in there[22:18:45] <lundh> I dont get it. This property indicates whether we should delegate a ZFS dataset to an OS VM. If true, the VM will get a dataset /data which it will be able to manage.[22:18:49] <lundh> is all I find[22:19:17] <konobi> yup[22:19:19] <wesolows> what else do you want? ;-)[22:19:56] <wesolows> delegation (described in more detail in the zfs docs) allows you to make snapshots, set properties, etc. on that dataset, which you normally cannot do.[22:20:12] <wesolows> it gives you a second dataset, in addition to the root dataset, that you completely control[22:20:34] <lundh> guess I was unclear[22:21:34] <lundh> I have a zfsvolume /zones/media where I want to save all my music. I want that area to be available to zones so that a zone can share it to my mac[22:21:56] <LeftWing> Then you don't want to delegate.[22:22:01] <MerlinDMC> lundh, use filesystems[22:22:02] <konobi> have it in one zone and then share it from there[22:22:03] <wesolows> that's not delegation[22:22:19] <wesolows> and what do you mean by share? HTTP, fine. NFS, not possible.[22:22:30] <lundh> afs using netatalk[22:22:33] <wesolows> ok[22:22:45] * wesolows assumes that works but has no direct knowledge[22:23:00] <LeftWing> It's just a daemon like samba, and I can confirm it works.[22:23:02] <lundh> well, netatalk installed so I would assume that it works[22:23:10] <MerlinDMC> filesystems: [{ type: "lofs", src: "/zones/media", target: "/data" }][22:23:23] <MerlinDMC> if i recall those names correctly -.-[22:23:26] <lundh> MerlinDMC: thank you :)[22:23:49] <LeftWing> I would probably use a different target, but otherwise yes.[22:25:32] <konobi> i just shared a directory from within a zone[22:25:58] <konobi> using cifs[22:26:27] <LeftWing> Personally I use UUCP to get my files.[22:26:40] <lundh> I would be nervous about keeping those files in a vm[22:27:17] <LeftWing> What are you worried will happen to them?[22:27:40] <lundh> destroying the wrong vm would kill them for instance[22:27:49] <LeftWing> So would typing rm in the wrong place[22:28:16] <lundh> I know[22:28:37] <lundh> what is UUCP?[22:28:48] <wesolows> something that no one uses to access their files[22:28:50] <LeftWing> A terrible joke, for which I apologise. :P[22:28:56] <lundh> hehe[22:29:07] <lundh> you had me googleing it for like 2 minuts[22:29:09] <lundh> ![22:29:17] <wesolows> its main function is in fidonet gateways[22:29:46] <lundh> the consensus is to keep the files directly in the the VM?[22:29:54] <konobi> lundh: that's what snapshots are for[22:29:58] <wesolows> that's what I do, and what I would do[22:30:17] <lundh> konobi: snapshot the vm to keep the files safe?[22:30:23] <LeftWing> If you're likely to destroy the wrong VM, though, while I am sad I guess I would advise you to do whatever makes you comfortable. :P[22:30:35] <LeftWing> konobi: Surely the snapshots would go away if you destroy the VM.[22:30:36] <wesolows> the only reason to do otherwise is if you somehow need to access the files from multiple zones, which seems potentially dangerous anyway[22:30:50] <konobi> LeftWing: you'd think so =0)[22:30:59] <lundh> LeftWing: I'll liste to the experts :)[22:31:07] <wesolows> we're all experts :-)[22:31:17] <LeftWing> Well, just make sure you get both bits of advice -- the latter surely being not to delete random VMs. :P[22:31:17] <lundh> I'm not :)[22:31:18] <konobi> actually, i use a /data delegated_dataset and serve from there[22:31:36] <wesolows> konobi: yes that's the approach I've used. it works well.[22:32:11] <wesolows> put it this way, that's how the system was designed to be used. it's obviously not the only possible solution.[22:32:14] <lundh> care to explain or send me to a nice page to understand what that is?[22:32:17] <konobi> yeah, can snapshot it seperately and from within the zone[22:32:26] <LeftWing> I may have done that as well, except that my pool used to be an OI box and my data was already laid out. :P[22:33:01] <wesolows> set delegated_dataset to true, then store all your data in /zones/<uuid>/data in the zone, and share it from there[22:33:30] <wesolows> you can obviously mount that wherever you want, take snapshots, etc. if/when you vmadm destroy, your data will go too, so Don't Do That Then (TM).[22:33:47] <LeftWing> Thanks, Doctor![22:33:50] <konobi> one thing i've been confused about is why we never stuck all the kvm images/disks under a single child[22:34:56] * LeftWing has often wondered that, also.[22:34:59] <lundh> thanks wesolows[22:37:06] <lundh> wesolows: if I snapshot the /zones/<uuid>/data will that snapshot disapear when I destroy the vm?[22:37:50] <wesolows> I assume so[22:37:58] <wesolows> it would be a bug if not[22:38:16] <lundh> dont rely on it, check :)[22:38:22] <wesolows> remember, in the design of the system, the only reason to vmadm destroy is that either the tenant requests it or you haven't paid your bill[22:38:46] <wesolows> letting a tenant make destruction impossible by creating a snapshot would be a bug[22:38:48] <lundh> yes[22:39:10] <lundh> when you lay it out like that... :)[22:40:27] <konobi> wesolows: there's an issue about child snapshots, iirc[22:40:44] <wesolows> zfs destroy -r should do the job no matter what[22:42:21] <lundh> wesolows: I could potentially clone the /data-volume before the destruction and then use it afterwards?[22:42:24] <konobi> oh, that was template snapshots[22:44:35] *** Vod has quit IRC[22:47:18] *** Webhostbudd has quit IRC[22:48:17] <konobi> lundh: yup, clone and promote probably[22:48:22] <konobi> or just move[22:48:49] <lundh> :)[22:52:24] <lundh> I have a stray /zome/<uuid> volume that doesnt have a vm according to vmadm, safe to delete or should I remove it in a special way?[22:52:52] <konobi> is it a img?[22:53:07] <lundh> oh, that could be it[22:53:45] <linuxprofessor> oh, removing the 2nd nic didnt make a difference[22:53:58] <linuxprofessor> still getting the same fault[22:54:47] <konobi> remove the trailing comma too?[22:54:54] <linuxprofessor> yupp[22:55:10] <konobi> odd[22:55:23] <wesolows> if you want to minimise the probability of accidentally deleting something, it's best to stick to the tools and not go manually changing stuff in zfs[22:55:27] <linuxprofessor> been getting those now and then when creating vms[22:55:54] <linuxprofessor> for instance, i cant create smartosplus machine[22:55:58] <linuxprofessor> just standard and base[22:56:08] <linuxprofessor> smartosplus gives the same error message[22:56:15] <konobi> really slow disks or something?[22:56:37] <linuxprofessor> standard sata disks[22:56:39] <linuxprofessor> 7200 rpm[22:57:17] <konobi> over usb 1.1?[22:57:24] <linuxprofessor> hehe, nooo =)[22:57:48] <konobi> can't think of anything that would cause that[22:57:54] <konobi> rebooted?[22:58:18] <linuxprofessor> what is this? windows 95? :-P[22:58:19] <MerlinDMC> linuxprofessor, what platform version are you running on?[22:58:44] <linuxprofessor> MerlinDMC: how do i check that?[22:58:54] <MerlinDMC> sysinfo[22:59:21] <linuxprofessor> 20121213T231116Z[22:59:28] <linuxprofessor> is that it?[23:00:48] <MerlinDMC> that should have the commit https://github.com/joyent/smartos-live/commit/927cfb6b3b79211f62655cf3b55b4e0e4466586e[23:01:07] <MerlinDMC> so i don't know what would cause the waiting[23:01:51] <lundh> the delegate_dataset did not create a /data in the vm. it created a /zones/<uuid>/data in the VM[23:01:54] <konobi> mdata?[23:01:55] <lundh> any idea why?[23:02:01] *** wolfeidau has quit IRC[23:02:14] <konobi> lundh: where is it mounted?[23:02:51] <lundh> from within the vm:[23:02:52] <lundh> zones 20.6G 3.55T 1.30M /zones[23:02:52] <lundh> zones/b9d6e92d-5cf6-454b-ad9e-79984bacece4 14.9M 9.99G 420M /zones/b9d6e92d-5cf6-454b-ad9e-79984bacece4[23:02:55] <lundh> zones/b9d6e92d-5cf6-454b-ad9e-79984bacece4/data 144K 9.99G 144K /zones/b9d6e92d-5cf6-454b-ad9e-79984bacece4/data[23:03:10] <konobi> you can change the mountpoint =0)[23:03:34] <lundh> but why was it created like that at all?[23:03:47] <konobi> incase there's a /data to begin with[23:03:47] <lundh> it shouldnt according to the docs[23:03:57] <lundh> there isnt[23:05:24] * MerlinDMC sleeping ... hard day tomorrow - i hate xmas[23:05:41] <konobi> mkay... too tired to be thinking about this stuff... laters[23:05:48] <linuxprofessor> xmas indeed[23:05:52] <lundh> thansk for all the help![23:14:54] *** Vod has joined #smartos[23:19:31] *** daniel_wu has quit IRC[23:21:18] *** jamesd has joined #smartos[23:22:16] *** daniel_wu has joined #smartos[23:48:27] *** tonyarkles has quit IRC[23:53:15] *** masked has joined #smartos[23:58:02] *** wolfeidau has joined #smartos