September 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 | >
September 11, 2013 [00:00:27] *** neophenix has quit IRC[00:01:40] *** ryancnelson has quit IRC[00:02:43] *** ryancnelson has joined #smartos[00:03:25] *** gkyildirim has left #smartos[00:04:20] *** deirdres has quit IRC[00:04:38] *** deirdres has joined #smartos[00:16:30] *** backjlack has quit IRC[00:24:00] *** noahmehl has quit IRC[00:24:11] *** pringlescan has joined #smartos[00:32:12] *** backjlack has joined #smartos[00:36:10] *** mary5030 has quit IRC[00:42:13] *** tbatchelli_m has quit IRC[00:44:28] *** oddsignals has joined #smartos[00:44:54] *** sbellows has joined #smartos[00:47:08] *** deirdres_ has joined #smartos[00:47:48] <sbellows> hello. i'm trying to setup an internal image server for smartos following the directions found here: https://http://wiki.smartos.org/display/DOC/Managing+Images. I'm running into problems. First, /var/db/imgadm/ is empty. No sources.list. no imgcache.json. I created a sources.list, added my local server's url to it, and ran 'imgadm update'. Still no imgcache.json. The 'imgadm avail' command still returns only joyent images. a[00:47:48] <sbellows> ideas what I am doing wrong?[00:48:41] <bixu> Try trussing imgadm update to see what it's doing an if there are any weird errors?[00:50:38] *** deirdres_ has joined #smartos[00:50:45] *** deirdres has quit IRC[00:50:46] *** deirdres_ is now known as deirdres[00:50:47] <ryancnelson> ... when you run "imgadm avail" , is it actually hitting your server?[00:50:59] <ryancnelson> (seen in your access logs?)[00:51:11] <sbellows> no. no connection attempt to my server at all.[00:51:19] <ryancnelson> imgadm update ?[00:52:47] <sbellows> no output at all to 'imgadm update'. return code of 0. still nothing in /var/db/imgadm/ (other than the sources.list i created there).[00:53:10] *** oddsignals has quit IRC[00:53:11] <ryancnelson> what's in your /var/db/imgadm/sources.list ?[00:53:17] <ryancnelson> can you use "curl" to hit that url?[00:54:23] *** axonpoet has quit IRC[00:54:47] *** jim80net_ has joined #smartos[00:57:08] *** jim80net has quit IRC[00:57:16] <sbellows> i can. and it returns the image manifests i created. (via curl)[00:57:26] <sbellows> imgadm doesn't see it though.[00:57:38] <sbellows> [pax] 00-25-90-6c-6d-d2 /var/imgadm:# imgadm sources[00:57:38] <sbellows> https://images.joyent.com[00:59:00] *** pringlescan has quit IRC[00:59:20] <ryancnelson> so, man page for imgadm says you use "imgadm sources -a <URL>" to add that[00:59:33] <sbellows> ahhh. helps to read the man page. imgadm version 2 no longer uses sources.list or /var/db/imgadm/[00:59:54] *** jim80net_ has quit IRC[01:00:08] <sbellows> should have tried the man page before the wiki. thanks for the help![01:07:06] <ryancnelson> i tweaked that wiki page to be (mostly?) correct. thanks for the report[01:26:49] *** wolfeidau has joined #smartos[01:27:09] *** axonpoet has joined #smartos[01:31:34] *** ira has quit IRC[01:45:05] *** bixu has quit IRC[01:45:33] *** axonpoet has quit IRC[01:49:51] *** ismell has joined #smartos[01:49:53] <ismell> ello[01:50:47] *** Aredridel has joined #smartos[01:53:08] *** julianduque has joined #smartos[01:53:55] *** trentster has quit IRC[01:55:57] *** pringlescan has joined #smartos[01:56:29] <sbellows> i'm now wondering if that entire wiki page is out of date[01:57:38] <sbellows> i'm thinking the imgadm v2 tool now uses IMGAPI instead of DSAPI and that the newer protocol uses different syntax.[01:58:54] *** saltsa has quit IRC[02:00:11] *** trentster has joined #smartos[02:00:40] *** saltsa has joined #smartos[02:03:21] *** wolfeida_ has joined #smartos[02:03:58] *** wolfeidau has quit IRC[02:05:35] <sbellows> for instance, imgadm appears to want a 'ping' page at the root of the server. it will fail to add the source without it. the wiki page makes no mention of this.[02:07:16] *** cjones__ has quit IRC[02:10:52] *** chorrell has quit IRC[02:11:17] *** julianduque has quit IRC[02:16:04] *** axonpoet has joined #smartos[02:18:02] *** bixu has joined #smartos[02:31:41] *** vizant0 has quit IRC[02:42:50] *** jim80net has joined #smartos[03:06:16] *** deirdres has quit IRC[03:07:57] *** pringlescan has quit IRC[03:08:27] *** bixu has quit IRC[03:08:59] *** pringlescan has joined #smartos[03:13:42] *** FatDarrel has quit IRC[03:21:09] *** potatosalad has joined #smartos[03:26:34] *** chorrell has joined #smartos[03:34:04] *** potatosalad has quit IRC[03:51:45] *** vaskas has joined #smartos[03:51:50] *** ilovezfs has quit IRC[03:56:16] <vaskas> Hi guys. I'm new to Solaris and I'm currently trying SmartOS out with our Node and Rails apps. The pattern I noticed about memory consumption in a SmartMachine is that it bites into Swap heavily before it maxes out the RAM. Is this a valid observation?[03:56:18] *** julianduque has joined #smartos[03:56:21] *** ilovezfs has joined #smartos[04:03:05] *** wolstena has quit IRC[04:04:33] *** pringlescan has quit IRC[04:06:52] *** Aredridel has quit IRC[04:07:58] *** pringlescan has joined #smartos[04:19:23] <rmustacc> vaskas: Well there are two different things.[04:19:51] <rmustacc> The first is what you probably think of as 'swap' being a partition on disk traditionally where things go when they get swapped out.[04:20:26] <vaskas> i do :)[04:20:32] <rmustacc> Where as with what you're looking at it more generally refers to the maximum amount of allocatable virtual memory.[04:20:42] <rmustacc> So every time memory is allocated a swap reservation is made.[04:20:52] <rmustacc> That doesn't mean that anything is going out to that swap partition at all.[04:21:20] <vaskas> right, but what's the point of making the swap reservation upfront?[04:21:31] <rmustacc> Because every memory allocation makes a swap reservation there is no equivalent of an oom killer that will kill arbitrary processes.[04:23:02] <rmustacc> If you don't make that reservation up front and you run out then you devolve into the situation of well, who do you kill to make room?[04:24:14] <richlowe> your sysadmin.[04:24:53] <rmustacc> vaskas: That make a bit of sense?[04:27:17] *** sbellows has quit IRC[04:28:22] *** ryancnelson has quit IRC[04:29:35] <vaskas> rmustacc yes.. but who decides what to "oom kill" then?[04:30:03] <rmustacc> vaskas: Well, there is no oom killer on us.[04:30:21] <rmustacc> Either you can get the reservation when you allocate memory or you get ENOMEM.[04:31:22] <rmustacc> So there is no arbitrary kill signal dropped on you.[04:32:31] *** nefilim has joined #smartos[04:34:02] <vaskas> rmustacc so what if I'm out of RAM and want to start a new process?[04:34:32] <rmustacc> You cannot.[04:34:39] <rmustacc> If you've exceeded the total swap.[04:34:42] <vaskas> ouch[04:35:03] <rmustacc> Generally though the amount of swap a zone gets is larger than the amount of RAM.[04:35:46] <rmustacc> So generally speaking you'll have the memory for that process.[04:36:25] <rmustacc> But you're also going to be penalized for being over your memory cap and page ins will be throttled until sufficient data has been paged out.[04:37:30] *** deirdres has joined #smartos[04:37:59] <vaskas> right, my zone has as much swap as RAM (2048) so it looks like swap doesn't make sense at all this way[04:39:12] <vaskas> there surely must be some positive implication of this design[04:41:07] <rmustacc> It's simply two different knobs you can tune based on your own needs.[04:41:25] <rmustacc> Personally if I'm running a service I don't want my processes killed in those situations.[04:41:27] *** mary5030 has joined #smartos[04:41:30] *** axonpoet has quit IRC[04:43:33] <rmustacc> If you're looking to understand the original motivation behind the VM design then I can point you to the original USENIX papers from back in the day about this.[04:48:47] <vaskas> rmustacc that would be great, thanks[04:50:41] *** chorrell has quit IRC[04:50:44] <rmustacc> This is old, but a lot of it is still fairly relevant: http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.7.3530&rep=rep1&type=pdf[04:51:09] <rmustacc> Often the problem with the alternative approach is entirely what do you kill.[04:51:32] <rmustacc> I started an ssh session and I accidentally killed your database.[04:52:01] <vaskas> sounds pretty bad[04:52:51] <rmustacc> I'm not sure why the Linux kernel opted for that direction instead.[04:53:58] <wesolows> very simple, really. none of the software in the GNU/Linux ecosystem checks return values, so something's going to explode regardless.[04:54:27] <wesolows> the idea is that at least with the oom killer there's some chance it won't be your production database[04:55:14] <wesolows> And of course we can't possibly not overcommit, because *handwave* *handwave* *NIH* *handwave*, so naturally the oom killer is the best choice![04:56:24] *** axonpoet has joined #smartos[04:56:30] <richlowe> I believe they have a whole intermixed set of tunables now[04:56:38] <richlowe> though if I know that, it's because of brendang[04:56:41] <wesolows> yeah[04:57:00] <wesolows> the oom killer grew a lot of tunables, so that people could try to control it. It can even be disabled now, as can overcommit.[04:57:19] <wesolows> They may be like the Americans: they'll do the right thing after exhausting the alternatives.[04:57:25] <wesolows> But NEVER by default.[04:58:24] *** pringlescan has quit IRC[05:06:58] *** jim80net has quit IRC[05:11:43] *** tbatchelli_m has joined #smartos[05:14:33] *** mary5030 has quit IRC[05:17:40] *** leecallen35 has joined #smartos[05:19:50] *** leecallen has quit IRC[05:22:30] *** potatosalad has joined #smartos[05:24:30] *** jim80net has joined #smartos[05:28:57] *** patrickmslattery has left #smartos[05:31:10] *** potatosalad has quit IRC[05:31:23] *** pringlescan has joined #smartos[05:42:27] *** tbatchelli_m has quit IRC[05:59:46] *** potatosalad has joined #smartos[06:04:50] *** sa5bke has quit IRC[06:06:54] <vaskas> I'm curious whether there will ever be a SmartOS EC2 AMI. I liked the nodejs SmartMachine - it's nicely equipped and ready for Node development. Wouldn't SmartMachine cloud images benefit Joyent?[06:18:07] *** potatosalad has quit IRC[06:18:26] *** pringlescan1 has joined #smartos[06:20:21] *** pringlescan has quit IRC[06:20:41] <wesolows> It wouldn't be a SmartMachine if it ran as a Xen guest.[06:22:00] <wesolows> As for what benefits us, I obviously don't speak for the company but we make money selling instances in our own Public Cloud. It's not clear to me that having people running SmartOS in EC2 helps us in any way, although of course we don't (and wouldn't) do anything to stop people from making AMIs of it.[06:22:24] <wesolows> It's an open source OS; you're free to do that if you like. Go wild.[06:29:47] *** pringlescan1 has quit IRC[06:32:37] *** pringlescan has joined #smartos[06:33:09] *** szaydel has quit IRC[06:34:52] *** axonpoet has quit IRC[06:41:27] *** nefilim has quit IRC[06:41:31] *** yruss972_ has joined #smartos[06:41:44] *** deirdres has quit IRC[06:49:12] *** yruss972_ has quit IRC[06:49:17] *** yruss972 has joined #smartos[06:49:57] *** yruss972 has quit IRC[06:50:09] *** yruss972 has joined #smartos[06:51:38] *** yruss972 has quit IRC[06:51:42] *** yruss972_ has joined #smartos[06:53:56] *** yruss972_ has quit IRC[06:54:38] *** yruss972 has joined #smartos[06:55:35] *** yruss972 has quit IRC[06:55:39] *** yruss972_ has joined #smartos[06:56:33] *** yruss972 has joined #smartos[06:56:48] *** yruss972_ has quit IRC[06:56:55] *** yofuh has quit IRC[06:57:12] *** yofuh has joined #smartos[06:59:57] *** yruss972_ has joined #smartos[07:01:38] *** yruss97__ has joined #smartos[07:02:35] *** yruss972_ has quit IRC[07:02:53] *** yruss972 has quit IRC[07:03:54] *** yruss97__ has quit IRC[07:05:41] *** yruss972 has joined #smartos[07:08:27] *** deirdres has joined #smartos[07:15:53] *** daleg has quit IRC[07:16:22] *** daleg has joined #smartos[07:30:31] *** trentster has quit IRC[07:31:09] *** trentster has joined #smartos[07:34:38] *** nikolam has joined #smartos[07:55:07] *** trentster has quit IRC[07:59:44] *** trentster has joined #smartos[08:06:53] *** pringlescan has quit IRC[08:07:10] *** chakatz has joined #smartos[08:16:17] *** mamash has joined #smartos[08:37:46] *** pringlescan has joined #smartos[08:45:18] *** pringlescan has quit IRC[08:48:21] *** deirdres has quit IRC[08:49:43] *** ilovezfs has quit IRC[08:52:37] *** ilovezfs has joined #smartos[08:55:37] *** jaimef has quit IRC[08:57:31] *** jaimef has joined #smartos[09:00:04] *** ilovezfs has quit IRC[09:01:25] *** yruss972 has quit IRC[09:04:35] *** julianduque has quit IRC[09:06:46] *** ilovezfs has joined #smartos[09:11:27] *** yruss972 has joined #smartos[09:20:02] *** texarcana has quit IRC[09:22:06] *** texarcana has joined #smartos[09:27:50] *** vaskas has quit IRC[09:30:36] *** wolfeida_ has quit IRC[09:40:22] *** ilovezfs has quit IRC[09:41:49] *** KermitTheFragger has joined #smartos[09:42:39] *** pringlescan has joined #smartos[09:47:42] *** pringlescan has quit IRC[09:47:55] *** ilovezfs has joined #smartos[10:01:21] *** vizant0 has joined #smartos[10:10:29] *** bens1 has joined #smartos[10:22:56] *** andy_js has joined #smartos[10:35:14] *** wolfeidau has joined #smartos[10:42:52] *** pringlescan has joined #smartos[10:44:46] *** system5 has quit IRC[10:46:39] *** system5 has joined #smartos[10:48:09] *** pringlescan has quit IRC[11:05:14] *** yruss972 has quit IRC[11:13:24] *** W0rmDr1nk has joined #smartos[11:25:12] <LnxTx> how can i update manifest of installed image? (imgadm install -P tank -f img.zvol.bz2 -m manifest.json)[11:25:21] <LnxTx> or add property?[11:27:07] <jperkin> just change the manifest.json, or am I misunderstanding the question?[11:28:45] <LnxTx> ok, image was installed and few kvms was created. where is images manifest after installation?[11:28:52] *** yruss972 has joined #smartos[11:29:12] <jperkin> you can't modify it once it's installed, but you could just change the manifest and update the uuid and re-import and use the new one[11:36:31] <LnxTx> hmm, what if i edit file /var/imgadm/images/POOL-UUID.json?[11:38:10] <LnxTx> works :)[11:41:58] *** khushildep has joined #smartos[11:43:18] *** pringlescan has joined #smartos[11:47:57] *** pringlescan has quit IRC[12:04:43] *** wolfeidau has quit IRC[12:14:19] *** trentster has quit IRC[12:18:35] *** trentster has joined #smartos[12:41:42] *** backjlack has quit IRC[12:43:35] *** pringlescan has joined #smartos[12:47:27] *** neophenix has joined #smartos[12:48:24] *** pringlescan has quit IRC[12:58:28] *** oddsignals has joined #smartos[12:59:11] *** TMinus36 has quit IRC[13:14:57] *** hadret has joined #smartos[13:18:48] *** sebasp has quit IRC[13:19:06] *** sebasp has joined #smartos[13:28:09] <nikolam> what is usually used to access KVM machines?, to run their GUI remotely (except native solutions like rdp for windows, vnc, ssh -X etc)[13:29:12] <jperkin> vnc to the console port[13:36:39] *** papasmurf has quit IRC[13:36:48] *** papasmurf has joined #smartos[13:37:29] *** szaydel has joined #smartos[13:43:55] *** pringlescan has joined #smartos[13:48:51] *** pringlescan has quit IRC[13:52:58] <jesse_> something doesn't add up[13:53:10] <jesse_> zpool status says 322M/s for scrub[13:53:46] <jesse_> zpool iostat has not seen more than 280M (zpool iostat pool 30)[13:54:11] <jesse_> and it's a 5-disk raidz, so iostat should be higher?[13:55:10] <jesse_> (that 322M/s is average? total? of the past 10 hours the scrub has been running)[13:55:18] <jesse_> s/total/max/[13:55:30] *** chakatz has quit IRC[14:00:27] *** ilovezfs has quit IRC[14:01:02] *** ilovezfs has joined #smartos[14:04:12] *** yruss972 has quit IRC[14:08:50] *** ilovezfs has quit IRC[14:10:33] *** ilovezfs has joined #smartos[14:20:13] *** ira has joined #smartos[14:23:16] *** nikolam has quit IRC[14:25:16] *** yruss972 has joined #smartos[14:25:44] *** chakatz has joined #smartos[14:28:08] <jesse_> and another thing that has been bugging me for a long time[14:28:20] <jesse_> why does otherwise idle pool write while doing scrub?[14:28:31] <jesse_> I thought scrub was supposed to only read[14:43:32] *** oddsignals has quit IRC[14:44:17] *** pringlescan has joined #smartos[14:48:34] *** jim80net has quit IRC[14:49:18] *** pringlescan has quit IRC[14:54:03] *** ktk is now known as ktkNA[14:54:15] *** chakatz has quit IRC[14:56:10] *** TMinus36 has joined #smartos[14:59:18] *** ahaydock has quit IRC[15:03:47] *** chakatz has joined #smartos[15:06:42] *** W0rmDr1nk has quit IRC[15:08:25] *** ahaydock has joined #smartos[15:09:33] *** nikolam has joined #smartos[15:12:04] <nahamu> jesse_: perhaps making note of its progress so it knows where to pick up on a reboot... check the source if you really want to know. ;)[15:12:24] *** W0rmDr1nk has joined #smartos[15:12:47] <nahamu> (recall that if your scrub is half done and you reboot, the scrub continues rather than starting from scratch)[15:13:00] <jesse_> nahamu, that would make kind of sense. But why do that 41 iops/250KB/s?=)[15:13:15] <nahamu> dtrace and find out![15:13:41] *** ira has quit IRC[15:14:22] *** pringlescan has joined #smartos[15:14:57] <jesse_> ah, but that's how it's always with dtrace - takes a day to figure out how to get the right information[15:15:11] <jesse_> I'm not that interested, just a bit curious=)[15:17:55] *** W0rmDr1nk has quit IRC[15:19:12] *** pringlescan has quit IRC[15:24:07] *** W0rmDr1nk has joined #smartos[15:25:06] *** jim80net has joined #smartos[15:38:28] <nahamu> Ask in #illumos perhaps. Someone in there might know offhand.[15:44:56] <jesse_> scrub status update is quite plausible. if it's written with parity, that would increase the amount written[15:45:36] <jesse_> probably not something one would want to do with some configurations, though[15:45:52] <jesse_> like big read-only SSD pool[15:49:14] *** pringlescan has joined #smartos[15:53:29] *** nikolam has quit IRC[15:55:32] *** nikolam has joined #smartos[16:04:23] *** AlainODea has quit IRC[16:10:56] *** chorrell has joined #smartos[16:13:30] *** jsm has joined #smartos[16:15:27] <jsm> where is the appropriate place to file bugs against pkg_info? it is throwing a segfault with one of our packages when doing pkg_info -X <filename> .. package is here: http://s3.amazonaws.com/downloads.basho.com/riak/1.4/1.4.2/smartos/13.1/riak-1.4.2-SmartOS-x86_64.tgz[16:16:17] *** nefilim has joined #smartos[16:17:39] <jperkin> jsm: the best place would be the netbsd pr database, but you can probably log it as a github issue against joyent/pkgsrc and we can look.[16:18:58] <jperkin> it's probably that you are missing a particular field that the pkg tools are expecting[16:19:19] <jsm> jperkin: yeah, i'm sure of it, I just don't think that should cause a segfault :)[16:19:36] <jperkin> sure, it's a legitimate bug[16:19:56] <mamash> ah, there's 1.4.2 out, i'll take that as a notice :)[16:20:30] <jsm> i thought of filing a bug here: http://www.netbsd.org/cgi-bin/sendpr.cgi?gndb=netbsd, but that asks me "Which NetBSD Release You Are Using:"[16:24:00] <jperkin> ignore that, just ensure it's in the 'pkg' category[16:27:58] <jsm> jperkin: cool, i just put SmartOS s the NetBSD release, and used the pkg category. thanks for the tip[16:30:32] *** Trixboxer has joined #smartos[16:30:54] <Trixboxer> Hi, how can I install http://search.cpan.org/~makamaka/JSON-RPC-0.96/lib/JSON/RPC/Server.pm in smartos ?[16:31:28] *** chakatz has quit IRC[16:31:33] *** axonpoet has joined #smartos[16:32:54] *** nikolam has quit IRC[16:36:40] *** bradley0 has joined #smartos[16:37:02] *** chakatz has joined #smartos[16:41:53] *** mary5030 has joined #smartos[16:44:42] <pringlescan> I'm not even sure where to ask, but, I was wondering if anyone has ever set up a temporary mailserver that simply forwards mail based on a set of rules when doing an E-mail migration?[16:45:42] <bradley0> like a postfix smtp relay?[16:47:30] <daleg> sure, and what you can use to do that depends on how complex or simple your rules need to be[16:47:56] <yofuh> oone has ever done a backup mx![16:47:57] <pringlescan> bradley0: exactly[16:48:12] <yofuh> noone[16:48:31] <pringlescan> I have to leave people's google apps account open so they can retrieve content, but I want all outgoing mail to go to their new google apps account (in a different … google apps account)[16:49:02] <pringlescan> I mean, I myself, have two separate google apps accounts, they're not in separate organizations, they're completely separate[16:49:14] <pringlescan> the admin api won't let me mass create forwarding rules or add pop3 accounts[16:49:38] <pringlescan> so other than forwarding it using a relay and doing an exchange migration using their tool I don't see any other good options[16:50:01] *** nikolam has joined #smartos[16:50:42] *** backjlack has joined #smartos[16:53:30] <jperkin> I'm sure that's possible with newfangled MTAs, I used to do similar with sendmail's generics table[16:56:31] *** orangeroo has joined #smartos[16:58:58] <orangeroo> Hi all. Anyone familiar with a referservation error when creating a new virtual machine?[17:00:15] *** tbatchelli_m has joined #smartos[17:02:31] *** nikolam has quit IRC[17:05:30] <orangeroo> I've read the logged chat sessioned, but I've run into something that isn't solved by matching image_size.[17:05:41] <orangeroo> Or sessions[17:06:40] *** gkyildirim has joined #smartos[17:07:00] <orangeroo> Exported a custom dataset from SDC 6 into a somewhat recent release of SmartOS[17:09:33] *** eate has joined #smartos[17:09:57] *** eate has quit IRC[17:13:43] *** trentster has quit IRC[17:15:33] *** tbatchelli_m has quit IRC[17:15:36] <hadret> hi! to have network working in non-global zone, do I need to create for each and one of them vlan bridged to physical NIC?[17:15:37] *** deirdres has joined #smartos[17:16:20] <orangeroo> hadret: depends on what you're trying to do.[17:17:10] <orangeroo> You don't have to have a physical nic for each instance.[17:17:26] *** nikolam has joined #smartos[17:17:40] <hadret> I get that, but what about vlan?[17:18:04] <hadret> I want to have access to outside world from within the non-global zone[17:18:10] *** leecallen35 has quit IRC[17:18:47] *** leecallen has joined #smartos[17:18:55] <orangeroo> Your nic is already doing 802.1q?[17:20:03] <hadret> no[17:20:16] <hadret> I haven't touch it yet[17:22:38] <hadret> one physical NIC, how can I configure non-global zone to have access to outside world?[17:23:10] *** nikolam has quit IRC[17:23:36] <orangeroo> Your zone can only access the network that your one nic is exposed to if your switch is configured as an access port.[17:24:37] <orangeroo> You have to enable 802.1q on your switch and nic to make all the vlans you need available to a zone or zones.[17:24:54] <orangeroo> I may not be understanding what you are trying to do though.[17:25:08] <hadret> cool, that's helpful enough[17:25:14] <hadret> thank you[17:25:23] <orangeroo> Take a look at this: http://wiki.smartos.org/display/DOC/vmadm+JSON+Quick+Reference[17:25:44] <orangeroo> when you configure your nic, you may need to specify the vlan_id.[17:26:37] <hadret> it will create vlan then automatically?[17:27:17] <orangeroo> No vlans get created. It just allows the vnic to recieve packets that are tagged for a particular vlan.[17:29:06] *** daleg has quit IRC[17:29:22] *** jsm has quit IRC[17:29:53] *** pringlescan has quit IRC[17:30:13] <hadret> so by default, the one physical NIC should be exposed and used by non-global zone? or do you need to somehow define it?[17:30:27] *** swaagie has joined #smartos[17:30:59] <orangeroo> I don't think there's a default. I believe you have to define a nic in the json.[17:31:19] <orangeroo> If your host only has a single nic, it would probably be the admin nic.[17:31:51] <orangeroo> I like using etherstubs and a NAT myself.[17:31:59] *** jsm has joined #smartos[17:32:26] <hadret> etherstubs, will have a look[17:32:49] <orangeroo> You do have to do some work to get the NAT part working though...[17:32:54] *** nikolam has joined #smartos[17:33:10] <hadret> ok, so last thing -- when I'm defining the admin NIC card, the only one in the system for the non-global zone, IP, gateway etc. needs to be the same as for the global zone, correct?[17:33:58] <orangeroo> gateway, yet, but you have to use a unique IP.[17:34:12] <orangeroo> or yes instead of yet[17:35:48] *** daleg has joined #smartos[17:35:56] <hadret> seems alrigh, though I have a dedicated IP on this machine[17:36:02] <hadret> so one NIC with IP[17:36:19] <hadret> guess I will need to create vlan and bridge it to the NIC for earch zone[17:41:54] *** chakatz has quit IRC[17:48:13] *** tylerflint has joined #smartos[17:49:54] <tylerflint> we have a smartos machine, that keeps dropping it's default route, about 1 per day, at various times. Has anybody seen that?[17:50:10] <tylerflint> s/1/once/[17:51:09] <orangeroo> Haven't seen that myself.[17:52:35] <orangeroo> tylerflint: How do you get your route?[17:54:57] <tylerflint> it's all set by the /etc/usbkey/config[17:55:12] <tylerflint> so the smf logs for network-routing-route say the service is restarting, and the reason is "no refresh method defined"[17:55:25] *** nefilim has quit IRC[17:55:27] <tylerflint> which would make me think something is trying to reload the service[17:56:16] <orangeroo> Dynamic routing?[17:57:01] <orangeroo> I mean, a static default route shouldn't go anywhere.[17:58:09] *** mamash has left #smartos[17:58:15] *** bradley0 has quit IRC[18:01:09] *** nikolam has quit IRC[18:02:18] *** orangeroo has quit IRC[18:02:25] <tylerflint> it's just the default route that's generated on boot[18:07:51] *** FatDarrel has joined #smartos[18:10:30] *** nikolam has joined #smartos[18:18:16] *** bens1 has quit IRC[18:22:54] *** cjones_ has joined #smartos[18:32:16] *** ryancnelson has joined #smartos[18:40:18] <jsm> jperkin: a patch has already been delivered, that was quick: http://gnats.netbsd.org/cgi-bin/query-pr-single.pl?number=48207[18:41:41] *** Alek has quit IRC[18:43:22] *** Kristoomus has joined #smartos[18:43:25] <Kristoomus> New website up and running! Please write into this website contact information->comments what do you think about it. http://webchef.netau.net[18:45:38] *** tbatchelli_m has joined #smartos[18:46:05] *** KermitTheFragger has quit IRC[18:52:12] *** jedahan has joined #smartos[18:54:55] *** wolstena has joined #smartos[19:00:06] *** yruss972 has quit IRC[19:15:23] *** jsm_ has joined #smartos[19:16:18] *** jsm has quit IRC[19:16:18] *** jsm_ is now known as jsm[19:17:19] *** khushildep has quit IRC[19:22:57] *** Kristoomus has quit IRC[19:27:41] <pettson> kwork: Where is the new website?[19:28:01] <pettson> kwork: Sorry, I meant Kristoomus.[19:30:48] <kwork> pettson: i have mad frontpage skillz, i can'z help youz outz[19:31:33] *** ira has joined #smartos[19:36:53] *** W0rmDr1nk has quit IRC[19:38:37] *** jedahan has quit IRC[19:39:15] *** potatosalad has joined #smartos[19:42:53] *** deirdres has quit IRC[19:52:24] *** trentster has joined #smartos[19:55:18] *** leecallen has quit IRC[19:57:09] *** trentster has quit IRC[20:01:55] *** trentster has joined #smartos[20:05:51] *** ryancnelson has quit IRC[20:07:10] *** yruss972 has joined #smartos[20:07:23] *** pringlescan has joined #smartos[20:09:38] *** nikolam has quit IRC[20:12:00] *** yruss972 has quit IRC[20:12:13] *** Trixboxer has quit IRC[20:18:21] *** neophenix has quit IRC[20:23:35] *** vizant0 has quit IRC[20:25:04] *** Alek has joined #smartos[20:34:39] *** deirdres has joined #smartos[20:37:23] *** hadret has quit IRC[20:41:16] *** jsm_ has joined #smartos[20:41:41] *** jsm has quit IRC[20:41:41] *** jsm_ is now known as jsm[20:48:11] *** tbatchelli_m has quit IRC[20:51:16] *** ilovezfs has quit IRC[20:52:53] *** ilovezfs has joined #smartos[21:06:01] *** Alek2 has joined #smartos[21:07:30] *** Alek has quit IRC[21:08:11] *** cjones_ has quit IRC[21:11:02] <ismell> thats a typical HDD setup look like on a smartos host?[21:11:42] <wesolows> https://github.com/joyent/manufacturing[21:28:44] *** Triskelios has quit IRC[21:35:00] *** knotty66 has joined #smartos[21:38:35] <ismell> i guess what i'm asking is how are the drives configured in the pools?[21:39:14] <ismell> some of that hardware looks fun :)[21:40:35] <wesolows> for SmartOS (the non-product), see https://github.com/joyent/smartos-overlay/blob/master/smartdc/bin/smartos_prompt_config.sh#L397[21:41:25] <wesolows> for SDC, we use https://github.com/joyent/smartos-live/blob/master/src/disklayout.js and in particular https://github.com/joyent/smartos-live/blob/master/src/node_modules/disklayout.js[21:41:37] <wesolows> We're working to unify them (by using the SDC approach)[21:43:30] <ismell> oh fancy[21:44:17] <ismell> are there any plans to release an SDC interface for the home lab ?;)[21:44:57] <ismell> thinking about building me a new box with the Ivy Bridge-EP cpus that just came out[21:45:10] * jim80net likes to just pick two drives out of an array, setup the zones pool, then zpool add mirror the rest,[21:45:57] <mnaser> cyrus-sasl-2.1.26nb2 is failing to install because it can't create a "cyrus" user (ro-mounted /etc/passwd)[21:45:59] <mnaser> any suggestions?[21:46:44] <mnaser> "UX: /usr/sbin/useradd: ERROR: Cannot update system files - login cannot be created."[21:47:02] <mnaser> i dont really need it but need the libraries it pulls in[21:47:18] <jperkin> there are some workarounds to create users in the GZ, but yeh that's just going to be one of those things you need to deal with if you install stuff in the GZ[21:47:52] <mnaser> just saw this http://wiki.smartos.org/display/DOC/Allowing+user+CRUD+in+the+global+zone[21:48:33] <mnaser> i just need the libraries.. so i just need to simulate it[21:49:36] <jperkin> in case it's not obvious, we don't recommend doing most of the stuff on the wiki, unsupported etc etc.[21:51:25] *** Fuzai has quit IRC[21:51:42] *** Fuzai has joined #smartos[21:52:13] *** tylerflint has left #smartos[21:53:49] *** elijah-mbp has quit IRC[21:54:38] *** elijah-mbp has joined #smartos[21:58:50] *** wolstena has quit IRC[22:02:59] *** julianduque has joined #smartos[22:08:10] <backjlack> Has anyone built some kind of "ESXi" like system for SmartOS to run KVM based VMs?[22:08:37] <nathan7> backjlack: You mean like.. vmadm?[22:08:39] *** Alek2 has quit IRC[22:09:00] <nathan7> backjlack: Stock SmartOS is pretty much an ESXi replacement, minus the fancy UI[22:09:00] <wesolows> ismell: not sure what you mean by "an SDC interface". We will sell you SDC licenses for a total of 1 TB of DRAM or greater; smaller configurations aren't supported and you're stuck with SmartOS. That said, if you just mean the storage pool stuff, yes we do plan to replace the existing SmartOS code with the SDC stuff, which is already open source.[22:09:11] *** Alek has joined #smartos[22:09:35] <nathan7> wesolows: 1TB of DRAM for the entire set of servers, or per server?[22:09:42] <backjlack> nathan7: Got it, that seems to lack only a GUI. I'm not that familiar with SmartOS / Illumos.[22:09:45] <wesolows> total[22:09:49] <nathan7> coo'[22:10:01] <ismell> i'll be lucky to push 32gig on my home server ;)[22:10:03] <wesolows> I'm not aware of a single server running SDC that has 1 TB of DRAM :)[22:10:09] <wesolows> though it's supported.[22:11:32] <Alasdairrr> That's because you guys charge by the GB :-p[22:11:38] <nathan7> wasn't sure anymore what kind of setups Joyent runs[22:12:11] <wesolows> it's actually because 32 GB DIMMs are expensive and slow.[22:12:24] <rmustacc> nathan7: https://github.com/joyent/manufacturing[22:12:49] *** trentster has quit IRC[22:13:17] <nathan7> rmustacc: mhm[22:13:20] <backjlack> Does SmartOS support nested KVM on Intel hardware?[22:13:24] <nathan7> Never really did look at that properly[22:13:40] <nathan7> backjlack: SmartOS-in-SmartOS is unsupported, unsure about the rest[22:14:12] <rmustacc> We don't support nested KVM.[22:14:24] <backjlack> rmustacc: Got it, thanks.[22:15:54] *** wolstena has joined #smartos[22:17:19] *** trentster has joined #smartos[22:23:06] *** estibi has quit IRC[22:27:26] *** knotty66 has quit IRC[22:34:34] *** knotty66 has joined #smartos[22:34:47] *** bens1 has joined #smartos[22:37:39] *** andy_js has quit IRC[22:40:35] *** chorrell has quit IRC[22:47:57] *** tbatchelli_m has joined #smartos[22:50:05] *** nikolam has joined #smartos[22:56:43] *** andoriyu has joined #smartos[22:58:02] *** andoriyu has quit IRC[22:58:11] *** bens1 has quit IRC[22:58:58] *** andoriyu has joined #smartos[22:59:21] *** vizant0 has joined #smartos[23:01:13] *** cjones_ has joined #smartos[23:08:46] *** oddsignals has joined #smartos[23:10:02] *** mgls has quit IRC[23:12:33] *** tbatchelli_m has quit IRC[23:13:48] *** knotty66 has quit IRC[23:17:05] *** tbatchelli_m has joined #smartos[23:20:25] *** tbatchelli_m has quit IRC[23:20:31] *** tbatchelli_ has joined #smartos[23:23:45] *** chorrell has joined #smartos[23:26:10] *** tbatchelli_m has joined #smartos[23:28:54] *** tbatchelli_ has quit IRC[23:29:12] *** Alek has quit IRC[23:36:51] *** axonpoet has quit IRC[23:37:58] *** papasmurf is now known as volci[23:49:00] *** Alek has joined #smartos