September 9, 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 9, 2013 [00:02:29] *** prasm_ has joined #smartos[00:02:44] *** roast__ has joined #smartos[00:06:15] *** copec_ has joined #smartos[00:07:06] *** prasm has quit IRC[00:07:10] *** richlowe has quit IRC[00:07:11] *** _dumfries has quit IRC[00:07:12] *** djun has quit IRC[00:07:12] *** roast_ has quit IRC[00:07:15] *** copec has quit IRC[00:07:18] *** tbatchelli_m has quit IRC[00:07:21] *** jellydonut has quit IRC[00:07:34] *** link0 has quit IRC[00:07:40] *** nathan7 has quit IRC[00:07:41] *** scarcry has quit IRC[00:07:46] *** Ymarg has quit IRC[00:07:47] *** nshalman has quit IRC[00:07:51] *** backjlack has quit IRC[00:07:53] *** badboy_ has quit IRC[00:07:54] *** scanf has quit IRC[00:07:55] *** Licenser has quit IRC[00:07:55] *** ssm has quit IRC[00:07:57] *** rmustacc has quit IRC[00:07:58] *** TMinus36 has quit IRC[00:08:00] *** defel_ has quit IRC[00:08:01] *** jgrafton has quit IRC[00:08:14] *** e^ipi has quit IRC[00:08:16] *** david-ec_ has quit IRC[00:08:17] *** wolstena1 has quit IRC[00:08:19] *** texarcana has quit IRC[00:08:20] *** Teknix has quit IRC[00:08:21] *** marsell has quit IRC[00:08:24] *** trentster has quit IRC[00:08:24] *** ivan\ has quit IRC[00:08:29] *** LeftWing has quit IRC[00:08:32] *** bahamas10 has quit IRC[00:08:37] *** mnaser has quit IRC[00:08:39] *** |woody| has quit IRC[00:08:39] *** blu has quit IRC[00:08:39] *** mmalecki has quit IRC[00:08:40] *** njm has quit IRC[00:08:44] *** rodgort has quit IRC[00:08:46] *** estibi has quit IRC[00:08:49] *** nhubbard has quit IRC[00:08:49] *** pettson has quit IRC[00:08:51] *** copec_ is now known as copec[00:08:51] *** mikl has quit IRC[00:08:51] *** ivan\ has joined #smartos[00:09:54] *** andy_js has quit IRC[00:11:55] *** djun has joined #smartos[00:18:22] *** njm has joined #smartos[00:18:22] *** rodgort has joined #smartos[00:18:23] *** estibi has joined #smartos[00:18:23] *** pettson has joined #smartos[00:18:23] *** nhubbard has joined #smartos[00:18:42] *** tbatchelli_m has joined #smartos[00:18:42] *** jellydonut has joined #smartos[00:18:42] *** Ymarg has joined #smartos[00:18:42] *** link0 has joined #smartos[00:18:42] *** nathan7 has joined #smartos[00:18:43] *** scarcry has joined #smartos[00:18:43] *** nshalman has joined #smartos[00:19:42] *** badboy_ has joined #smartos[00:19:42] *** scanf has joined #smartos[00:19:42] *** Licenser has joined #smartos[00:19:42] *** ssm has joined #smartos[00:19:42] *** rmustacc has joined #smartos[00:19:43] *** TMinus36 has joined #smartos[00:19:43] *** defel_ has joined #smartos[00:19:43] *** jgrafton has joined #smartos[00:19:43] *** e^ipi has joined #smartos[00:20:01] *** richlowe` has joined #smartos[00:20:01] *** david-ec_ has joined #smartos[00:20:01] *** wolstena1 has joined #smartos[00:20:01] *** texarcana has joined #smartos[00:20:01] *** Teknix has joined #smartos[00:20:01] *** marsell has joined #smartos[00:20:01] *** trentster has joined #smartos[00:20:02] *** LeftWing has joined #smartos[00:20:02] *** bahamas10 has joined #smartos[00:20:02] *** mnaser has joined #smartos[00:20:02] *** |woody| has joined #smartos[00:20:02] *** blu has joined #smartos[00:20:02] *** mmalecki has joined #smartos[00:20:21] *** richlowe` is now known as richlowe[00:35:19] *** potatosalad has joined #smartos[00:49:28] *** Niamkik_ has quit IRC[00:49:54] *** Niamkik has joined #smartos[01:00:04] *** nefilim has joined #smartos[01:00:52] *** potatosalad has quit IRC[01:07:11] *** cjones_ has joined #smartos[01:14:39] *** cjones_ has quit IRC[01:21:57] *** cjones_ has joined #smartos[01:25:56] *** tbatchelli_m has quit IRC[01:33:43] *** roast__ is now known as roast_[01:33:59] *** gkyildirim has left #smartos[01:34:10] *** roast_ has quit IRC[01:34:10] *** roast_ has joined #smartos[01:36:20] *** cjones_ has quit IRC[01:45:21] *** mary5030 has joined #smartos[01:47:33] *** tnn_ has quit IRC[01:47:37] *** szaydel has joined #smartos[01:49:19] *** szaydel has quit IRC[01:50:29] *** vaskas has joined #smartos[02:16:04] *** szaydel has joined #smartos[02:22:11] *** vaskas has quit IRC[02:36:11] *** cjones_ has joined #smartos[02:37:06] *** vaskas has joined #smartos[02:59:10] *** swaagie has quit IRC[03:02:30] *** cjones_ has quit IRC[03:04:15] *** cjones_ has joined #smartos[03:06:11] *** cjones_ has quit IRC[03:12:05] *** cjones_ has joined #smartos[03:13:10] *** cjones_ has quit IRC[03:16:20] *** vaskas has quit IRC[03:25:41] *** vaskas has joined #smartos[03:26:35] *** szaydel has quit IRC[03:39:10] *** TMinus36 has quit IRC[03:43:11] *** dysinger has joined #smartos[03:54:14] *** potatosalad has joined #smartos[04:18:08] *** trentster has quit IRC[04:22:39] *** trentster has joined #smartos[04:31:42] *** TMinus36 has joined #smartos[04:38:19] *** masquerade has joined #smartos[04:38:43] *** masquerade is now known as Guest49064[04:48:01] *** jellydonut has quit IRC[05:15:43] *** leecallen35 has joined #smartos[05:18:31] *** leecallen has quit IRC[05:18:52] *** cjones_ has joined #smartos[05:42:05] *** tbatchelli_m has joined #smartos[05:45:01] *** potatosalad has quit IRC[05:49:28] *** AlainODea has joined #smartos[05:52:22] *** yruss972 has joined #smartos[05:52:27] *** dysinger has left #smartos[05:56:23] *** trentster has quit IRC[06:00:59] *** tnn_ has joined #smartos[06:01:57] *** trentster has joined #smartos[06:03:21] *** yruss972 has quit IRC[06:11:05] *** yruss972 has joined #smartos[06:12:27] *** tnn_ has quit IRC[06:14:54] *** MerlinDMC has quit IRC[06:17:53] *** potatosalad has joined #smartos[06:21:53] *** MerlinDMC has joined #smartos[06:31:06] *** tnn_ has joined #smartos[06:32:46] *** tnn_ has quit IRC[06:42:22] *** noahmehl has joined #smartos[06:43:17] *** cjones_ has quit IRC[06:56:35] *** mary5030 has quit IRC[06:57:26] *** nefilim has quit IRC[07:05:27] *** potatosalad has quit IRC[07:06:07] *** Guest49064 has quit IRC[07:07:00] *** tnn_ has joined #smartos[07:18:46] *** Hisa_X has quit IRC[07:19:01] *** tnn_ has quit IRC[07:22:04] *** Hisa_X has joined #smartos[07:22:41] *** yruss972 has quit IRC[07:24:13] *** yruss972 has joined #smartos[07:28:49] *** tbatchelli_m has quit IRC[07:33:49] *** masquerade has joined #smartos[07:34:13] *** masquerade is now known as Guest42829[07:40:11] *** Guest42829 has quit IRC[08:10:12] *** mamash has joined #smartos[08:15:41] *** potatosalad has joined #smartos[08:20:36] *** chakatz has joined #smartos[08:31:32] *** masquerade has joined #smartos[08:31:57] *** masquerade is now known as Guest23409[08:40:36] *** mamash has left #smartos[09:06:13] *** vaskas has quit IRC[09:11:08] *** deirdres has quit IRC[09:14:55] *** bixu has quit IRC[09:15:51] *** bixu has joined #smartos[09:18:41] *** mgls has joined #smartos[09:20:02] *** texarcana has quit IRC[09:22:02] *** texarcana has joined #smartos[09:24:18] *** trentster has quit IRC[09:24:52] *** mamash has joined #smartos[09:25:13] *** alcir has joined #smartos[09:46:36] *** trentster has joined #smartos[09:50:58] *** trentster has quit IRC[09:56:07] *** trentster has joined #smartos[10:02:05] *** whyz has quit IRC[10:06:06] *** whyz has joined #smartos[10:17:36] *** backjlack has joined #smartos[10:24:09] *** bens1 has joined #smartos[10:39:29] *** andy_js has joined #smartos[11:42:57] <yo61> Morning[11:43:35] <yo61> I'm giving fifo another go so am running through the install described here: http://project-fifo.net/display/PF/The+Installation+Guide[11:43:52] <yo61> I ran the step to provision the fifo zone: vmadm create -f setupfifo.json[11:44:16] <yo61> And it seemed to take a long time before exiting with: timed out waiting for /var/svc/provisioning to move for 897187a2-9ef5-4650-ab59-3fc210679172[11:45:11] <yo61> I could zlogin to the zone while the vmadm create was running, but now it's showing as "failed"[11:46:16] <yo61> Any idea what might be wrong?[11:47:04] <yofuh> not sure about fifo, but it usually happens if you run an older image on an newer smartos[11:47:07] <yo61> This is the json I used to create the fifo zone: https://gist.github.com/robinbowes/6493592[11:47:16] <yo61> Ah, OK[11:48:03] <yo61> I just updated and am using joyent_20130905T204057Z[11:48:04] <yofuh> in the mist cases you can just remove the var/avc/provisioning file manually during provisioning but that doesnt give you a guaranty that it really has finished[11:48:58] <yofuh> once its done as "failed" you can just remove the failed attribute line from the zones xml in /etc/zones but again, no guaratnie that the zone has completely configured[11:49:08] <yo61> OK[11:51:27] <yo61> Mmm, it's not. No network[11:55:29] <yofuh> could be one reason why it failed in the first place[11:55:35] <yo61> Could be[11:56:01] <yo61> Does the json I used to create the zone look OK?[11:57:39] <yofuh> does so, but of course i do not know anything about your network. "interface": "net0" doesn't hava any semantic here, it might broke something[11:57:57] <yo61> I tried again without that - same problem[11:58:22] <yofuh> well, as you can boot the zone now, you can debug[11:59:59] <LnxTx> yofuh: try "autoboot": false in json then start manually after create[12:00:29] <yofuh> LnxTx: no[12:10:02] <yo61> I'm struggling to know how to debug this[12:11:39] <yofuh> you could start by checking if the network devices did appear, network ifconfig output seems correct, default route exist... the ususal crap[12:11:55] <yo61> The NICs didn't appear[12:12:25] <yofuh> dladm shows you nithing?[12:13:00] <yo61> Nothing in the zone[12:13:19] <yo61> I tried creating a zone with no NICs - I get the same hang[12:14:12] <yofuh> that "hang" is not really a hang, it is just that the setup script fails and nobody removed var/svc/provisioning[12:14:28] <yo61> Yes, I understand that[12:14:41] <yo61> Am searching for setup script output/error msgs[12:15:05] <yofuh> the mechanism is anything but robust, so it really can be "anything", however, you problem is obviously that vmadm failed to create a vnic[12:15:55] <yofuh> i would check if a vnic for that zone exist in the gz (assume not) and check if you did somewhing strange to tha admin interface[12:16:35] <MerlinDMC> yo61, check if the selected IP is really unused in that network[12:17:35] <yofuh> sysinfo gives you some idea of you have a network interface named "admin" or not[12:17:58] <yo61> OK[12:18:01] <MerlinDMC> yo61, and you'll want to check the zoneinit logs after the zone got to failed state[12:19:22] <yo61> https://gist.github.com/robinbowes/6493842[12:21:26] <yo61> Hmm, I appear to also be hitting the zones/dump error that has been widely reported on the mailing list[12:29:30] <yo61> Fixed that[12:30:18] <yo61> The IP I'm using is definitely in used on the network[12:30:38] <yo61> Where do I find the zoneinit logs?[12:31:17] <yo61> nm - goti t[12:37:29] <yo61> I re-created the zone, the creation failed with the timeout error, I removed that attribute from /etc/zones/uuid.yml and restarted the zone[12:37:45] <yo61> I've logged in with zlogin and am looking in /var/log[12:38:46] <yo61> all logs are 0 bytes, except auth.log which has a couple of old msgs in it, ie. not from my most recent boot[12:47:00] <MerlinDMC> yo61, in /zones/<uuid>/root/var/svc/log/ ... named like *zoneinit*[12:47:15] *** Guest23409 has quit IRC[12:48:42] <yo61> No file named like that in there, just system-early-manifest-import:default.log[12:49:05] <MerlinDMC> then some service prior to zoneinit did fail[12:49:51] <MerlinDMC> I would (stupid but works) open a second ssh console and zlogin to the zone while it provisions to see what's failing ... you'll have about 5 minutes till the timeout event kills the zone[12:50:00] <yo61> https://gist.github.com/robinbowes/6494065[12:51:05] <MerlinDMC> and take a look at your GZ ... see if all services are running and don't sit in maintenance state[12:51:21] <yo61> Just ntp has failed[13:02:24] <yo61> fixed ntp[13:07:29] <yo61> Looks like I may be (still) hitting this: https://github.com/joyent/smartos-live/issues/158[13:16:13] *** nikolam has joined #smartos[13:31:24] *** yruss972 has quit IRC[13:37:34] *** chakatz has quit IRC[13:48:21] *** Hisa_X has quit IRC[13:51:04] <MerlinDMC> yo61, issue #158 is fixed for me I think ... I didn't hit that one anymore[13:51:22] <MerlinDMC> but i'll recheck that when i'm back home[13:51:53] *** Hisa_X has joined #smartos[13:57:50] <yo61> thanks[14:02:28] *** yruss972 has joined #smartos[14:06:49] *** nikolam has quit IRC[14:11:02] *** yruss972 has quit IRC[14:11:40] *** ira has joined #smartos[14:15:42] *** szaydel has joined #smartos[14:18:13] *** Hisa_X has quit IRC[14:31:22] *** jellydonut has joined #smartos[14:36:36] *** hadret has joined #smartos[14:38:07] <yo61> Well, I tried adding a tag to the 2nd nic in the machine (external) and using that for the new VM instead. Same issue[14:38:13] <yo61> Wonder if my image is screwed?[14:38:21] *** szaydel has quit IRC[14:40:33] <MerlinDMC> did you create your own image?[14:40:39] <yo61> No, I imported it[14:41:04] <MerlinDMC> a base image? ... that one for the fifo zone?[14:41:06] <yo61> Am going to re-import[14:41:16] *** Hisa_X has joined #smartos[14:41:32] <yo61> Yes[14:41:41] <yo61> base64 1.8.4[14:41:47] <yo61> fdea06b0-3f24-11e2-ac50-0b645575ce9d[14:42:03] <MerlinDMC> wait ... i'll check my vmware fifo install[14:42:25] <yo61> I already deleted it and am downloading[14:42:33] <yo61> Well, I say I'm downloading...[14:42:38] <yo61> Not a lot of progress so far...[14:43:26] <MerlinDMC> I did use a 13.2.0 base64 image ^^[14:43:37] <yo61> Oh, will that work with fifo?[14:43:56] <MerlinDMC> the UI at least works ... I just needed to check some changes there[14:44:11] <MerlinDMC> don't know if there will be issues[14:44:33] <MerlinDMC> but i'll try with a 1.9.x image .. I think the 1.8.x ones did bark while creating[14:44:45] <yo61> The wiki says 1.9.x doesn't wokr[14:44:49] <Licenser> yo61 you can try out the 1.13.2 image[14:44:57] <Licenser> 1.9 will not work[14:45:16] <yo61> 1.13.2?[14:45:31] <Licenser> erm 13.2.0 not 1.13.2[14:45:43] <yo61> :) I assumed that's what you meant[14:45:47] <Licenser> the newest image :) 13.2.0[14:45:49] <Licenser> sorry[14:45:56] <yo61> I'm getting imgadm import 0084dad6-05c1-11e3-9476-8f8320925eea[14:46:24] *** szaydel has joined #smartos[14:46:35] <Licenser> it's not well tested but it should work[14:46:48] <Licenser> the problem with 1.9 that there was no working nginx on it[14:47:38] <yo61> few mins to get... (damned slow broadband...)[14:47:55] <Licenser> but the nginx in 13.2.0 is new enough[14:48:02] <Licenser> but it's an adventure :)[14:48:11] * Licenser 'll be back in ~20[14:48:21] <yo61> Woo, just what I need - adventures[14:52:08] <yo61> <sigh> same issue[14:52:57] <MerlinDMC> if 13.2.0 does not provision then you have some major problem there[14:53:07] <yo61> If only I could work out what...[14:53:49] <MerlinDMC> yo61, try and see if that provisions: https://gist.github.com/MerlinDMC/e0b464c3f66e2e4df21f[14:54:33] <yo61> How long should it take?[14:55:57] <yo61> It appears to be doing the same as the previous attempts[14:56:45] <MerlinDMC> some seconds ... depending on your disks[14:57:10] <yo61> Still "hung" - I expect it to time out as did the previous zones[14:57:42] <MerlinDMC> but could also be that the newer image will test network connectivity and fail after a timeout ... here it took about 40 seconds to provision[14:58:03] <yo61> OK[14:58:16] <yo61> It's still "provisioning"[14:58:39] <MerlinDMC> then check the GZ logs for vmadmd and vmadm[14:59:25] <MerlinDMC> i need to get some customers happy will be back later[14:59:29] <yo61> kk[15:03:05] *** yruss972 has joined #smartos[15:04:31] *** chakatz has joined #smartos[15:11:19] *** yruss972 has quit IRC[15:12:50] *** yruss972 has joined #smartos[15:18:40] *** W0rmDr1nk has joined #smartos[15:19:38] <yo61> Am giving up for now - will wait until you guys get back. I have no idea what the issue is[15:19:41] *** xinkeT has joined #smartos[15:22:05] *** Tekni has quit IRC[15:32:02] <Licenser> yo61 back now[15:32:07] <Licenser> gut feeling says it's a networking problem[15:33:45] *** jim80net has joined #smartos[15:34:12] <yo61> Any idea how to diagnose?[15:34:20] <Licenser> zlogin and ping something :)[15:34:47] <yo61> I'm a bit of a smartos noob[15:35:02] <yo61> There are no network interfaces created in the zone[15:35:32] <yo61> eg. ping 8.8.8.8 returns unknown host 8.8.8.8[15:44:46] <MerlinDMC> Licenser, provisioning a zone with no networking also doesn't succeed[15:45:18] *** nikolam has joined #smartos[15:48:17] <Licenser> oh[15:50:27] <Licenser> hmm hmm[15:51:10] <yo61> I'm open to all and any ideas...[15:51:31] <Licenser> have you tried the mailing list?[15:51:36] <yo61> No, not yet[15:52:26] <Licenser> I've no ideas, vmadm is not very nice to debug if it does not work sadly :([15:52:56] <yo61> OK, will try the list[15:58:46] <yo61> One other thing - I have pkgin installed as per here: http://wiki.smartos.org/display/DOC/Installing+pkgin[15:59:33] <yo61> If I then run "pkgin ug" it uninstalls pkgin[15:59:47] <yo61> It says:[15:59:49] <yo61> 1 packages to be upgraded: pkgin-0.6.4[15:59:57] <yo61> 1 packages to be installed: pkgin-0.6.4nb1 (0B to download, -32B to install[16:00:11] <jperkin> pkgin doesn't handle upgrading itself yet..[16:00:16] <yo61> Then removes pkgin and doesn't install the new version[16:00:24] <jperkin> 'pkg_add -u pkgin' first[16:01:04] <yo61> Ah, OK, thx[16:01:42] <yo61> Hmm, that installs pkgin-0.6.0nb1[16:02:20] <jperkin> sounds like you installed the old bootstrap on that page then tried to manually update to a newer repo[16:02:26] <jperkin> in which case you will have an obsolete pkg_install.conf[16:02:48] <yo61> Ah yes[16:02:54] <yo61> PKG_PATH=http://pkgsrc.joyent.com/sdc6/2012Q2/x86_64/All[16:02:56] <jperkin> easier to just use a new bootstrap from the start: http://pkgsrc.joyent.com/packages/SmartOS/bootstrap/[16:03:02] *** masquerade has joined #smartos[16:03:25] *** masquerade is now known as Guest49298[16:03:55] <yo61> I used http://pkgsrc.joyent.com/packages/SmartOS/bootstrap/bootstrap-2013Q2-x86_64.tar.gz[16:04:46] <yo61> ie. I followed http://wiki.smartos.org/display/DOC/Installing+pkgin but used the later bootstrap package[16:05:00] <jperkin> ah, I didn't update the i386/x86_64 bootstraps, I'll fix that..[16:05:25] <yo61> OK, great[16:05:36] <yo61> Anything I can change manually? Or wait for you?[16:06:17] <jperkin> just unpack the bootstrap then do the pkg_add -u thing, that'll do the same thing[16:06:40] <jperkin> oh right, pkg_install.conf is done at base-install time, ignore..[16:06:54] <jperkin> we should move that[16:07:29] <yo61> Yes, pkg_add -u pkgin still uses pkgin-0.6.0nb1[16:09:49] <yo61> shall I remove pkg_install.conf ?[16:13:57] <yo61> Ok, got to go do the school run. Back later.[16:14:20] *** nefilim has joined #smartos[16:16:35] *** neophenix has joined #smartos[16:18:38] *** daleg has joined #smartos[16:28:46] *** nikolam has quit IRC[16:31:33] <alcir> how to purge pkgin? rm -rf /opt/local?[16:31:45] <jperkin> and /var/db/pkgin[16:32:17] <alcir> ok[16:34:03] *** axonpoet has joined #smartos[16:37:00] *** deirdres has joined #smartos[16:37:58] *** Guest49298 has quit IRC[16:44:25] *** mariusp has joined #smartos[16:44:55] *** mamash has left #smartos[16:51:00] *** mary5030 has joined #smartos[16:51:51] *** elijah-mbp has quit IRC[16:52:12] *** elijah-mbp has joined #smartos[16:54:29] *** masquerade has joined #smartos[16:54:53] *** masquerade is now known as Guest94807[16:59:57] *** tbatchelli_m has joined #smartos[17:14:43] *** cjones_ has joined #smartos[17:16:30] *** leecallen has joined #smartos[17:16:32] *** alcir has quit IRC[17:17:26] *** leecallen35 has quit IRC[17:18:21] *** potatosalad has quit IRC[17:19:26] *** Guest94807 has quit IRC[17:20:55] <jperkin> yo61: I updated the i386 and x86_64 bootstraps if you wanted to just use those now..[17:25:09] *** deirdres has quit IRC[17:30:47] <yo61> OK, pulling now[17:30:52] <yo61> Thanks[17:32:14] *** tbatchelli_m has quit IRC[17:33:02] <yofuh> jperkin: actually that "purge" approach is not enough since you have that dynamic UID crap(sorry, but it really sucks to have half installed cyrus and postfix packages because the user is already on the system before upgrade/install)[17:33:38] <jperkin> that bug is fixed with the latest pkg_install[17:33:41] <yofuh> you really need to remove all users that "could" be used before any postinstall script runs[17:33:58] <jperkin> but yes, it's not a complete purge, same for any databases in /var etc.[17:34:45] <yofuh> when (which pkgsrc release) and how did you fix it?[17:35:39] <jperkin> in the latest 2013Q2 bootstraps, one of the side-effects of some bad pkg_add -f code was that the metadata was left behind on a failed package install[17:36:36] <yofuh> ok did not try Q2 so far[17:37:24] <yofuh> on 2013Q1 the situation is bad as pkgin doesn't even notice the failed install[17:39:12] *** chakatz has quit IRC[17:49:46] *** tbatchelli_m has joined #smartos[17:49:57] *** yruss972 has quit IRC[17:52:35] *** potatosalad has joined #smartos[17:54:59] *** Shadow_aok has joined #smartos[17:55:01] <Shadow_aok> hi guys[17:55:54] <Shadow_aok> I have a question regarding the arch.[17:56:03] <Shadow_aok> My server has 4 Quad-Core AMD Opteron(tm) Processor 8356 U[17:56:24] <Shadow_aok> uname -a give me a i386 arch[17:56:56] <Shadow_aok> isainfo -v says I can run i386 and amd64 applications[17:57:10] <Shadow_aok> so, is smartos running in 32 or 64 bits ?[17:57:20] <wesolows> the kernel is always 64-bit[17:57:20] <jperkin> always 64-bit (isainfo -k)[17:57:36] <wesolows> it provides both the ability to build and run both 32- and 64-bit programs, always[17:58:13] <Shadow_aok> ok, isainfo givs me amd64[17:58:28] <Shadow_aok> it's troubling that uname doesn't display amd64 :)[17:58:30] <Shadow_aok> thanks[17:58:38] <wesolows> it's actually correct, though.[17:59:04] <Shadow_aok> huh ?[17:59:23] <wesolows> The documentation defines the meaning of these fields very clearly.[17:59:33] <wesolows> Your system's architecture (ISA) is in fact i386.[18:00:17] <wesolows> That's what uname -p should show. amd64 is just an extension, which isainfo correctly identifies.[18:00:22] <Shadow_aok> i see[18:01:44] *** szaydel_ has joined #smartos[18:02:25] *** chorrell has joined #smartos[18:02:50] *** szaydel has quit IRC[18:13:38] *** cjones_ has quit IRC[18:20:41] *** Shadow_aok has quit IRC[18:21:26] *** masquerade has joined #smartos[18:21:50] *** masquerade is now known as Guest42028[18:24:50] *** deirdres has joined #smartos[18:28:48] *** dap has joined #smartos[18:30:35] *** cjones_ has joined #smartos[18:31:36] *** ryancnelson has joined #smartos[18:32:31] *** jellydonut has quit IRC[18:33:13] *** jellydonut has joined #smartos[18:33:17] *** jellydonut has quit IRC[18:35:11] *** jellydonut has joined #smartos[18:37:09] <aszeszo> hi all, is it possible to do mount -F lofs /zones/zoneuuid/.zfs/snapshot /zones/zoneuuid/root/.snapshots to allow creating new zfs snapshots by creating new directories in /.snapshots/newsnapshotname inside zone?[18:37:25] <aszeszo> hope it makes sense :)[18:37:42] <rmustacc> You want mkdir to create a snapshot?[18:37:50] <aszeszo> yeah[18:37:53] <szaydel_> aszeszo: Not going to happen.[18:37:54] <aszeszo> works in gz[18:37:56] <rmustacc> Does that currently even work?[18:38:14] <aszeszo> but not via lofs mount to inside of a zone[18:38:33] <rmustacc> Your zone probably doesn't have permissions.[18:38:37] <rmustacc> The same way the zone can't zfs snapshot.[18:43:15] <nahamu> aszeszo: does it work if you do delegate_dataset ?[18:43:38] <nahamu> oh nevermind, you want to snapshot precisely the root filesystem of the zone...[18:44:19] *** nefilim has quit IRC[18:45:55] <nahamu> I guess one question is, is there any reason a similar flag couldn't just delegate the zone's root filesystem down to the zone?[18:46:27] <nahamu> sounds like even if it could, SDC wouldn't use the feature, so it would be up to someone else to implement it carefully.[18:46:51] *** tbatchelli_m has quit IRC[18:47:39] *** Alek has joined #smartos[18:48:03] <aszeszo> this stuff works from gz http://paste.ec/?75aa1e28b5a832a9#D/msUOOfUAjwqh4pNaXcn/2JXzwD3fCJiqGttTQ+jXQ=[18:49:14] <nahamu> you could probably do it on the delegated "data" filesystem if you used the "delegate_dataset" feature.[18:50:39] * nahamu .oO( What would happen if you used hyperlofs instead? )[18:51:16] <nahamu> it would probably still fail...[18:52:04] <nahamu> aszeszo: am I correct that your goal really is tied to the root filesystem so delegate_dataset is insufficient?[18:52:44] *** dap has quit IRC[18:53:02] *** dap has joined #smartos[18:53:26] *** ryancnelson has left #smartos[18:55:51] *** tbatchelli_m has joined #smartos[18:55:57] *** deirdres_ has joined #smartos[18:57:21] <aszeszo> nahamu: yes, want to use this hack inside existing zone without moving data around[18:57:30] <aszeszo> withouth any delegated datasets[18:59:24] *** deirdres has quit IRC[18:59:24] *** deirdres_ is now known as deirdres[19:07:07] *** gkyildirim has joined #smartos[19:07:37] <aszeszo> including sys_mount in the zonecfg limitpriv makes the hack work from inside zone[19:08:11] <aszeszo> ignore me[19:08:49] <aszeszo> (lofs mount disappeared after zone reboot)[19:18:01] *** deirdres has quit IRC[19:18:20] *** deirdres has joined #smartos[19:18:24] *** wramthun has joined #smartos[19:26:29] *** W0rmDr1nk has quit IRC[19:47:28] *** hadret has quit IRC[19:49:56] *** jaimef has quit IRC[19:52:51] *** Alek has quit IRC[19:56:14] *** killfill has quit IRC[19:57:07] *** jaimef has joined #smartos[19:59:18] *** mamash has joined #smartos[20:17:26] *** bens1 has quit IRC[20:26:38] *** Alek has joined #smartos[20:28:51] *** mamash has left #smartos[20:34:59] *** vertigo has joined #smartos[20:42:14] *** ryancnelson has joined #smartos[20:54:29] *** killfill has joined #smartos[21:08:52] <leecallen> I find that my zone has access to the snapshot files. It can't TAKE the snapshots, or roll them back - those must be done by the GZ - but the zone has access to the .zfs directory of the 'lofs' mounted filesystem.[21:10:08] <richlowe> not keen on that at all.[21:10:45] <ira> If you can see the .zfs directory, you should be able to see the snapshots.[21:10:56] <ira> I actually wrote a patch to fix that hole that does it cleanly.[21:11:31] *** julianduque has joined #smartos[21:11:52] <ira> but you must have access to the trigger directory… no trigger directory access, no snapshots.[21:13:03] <ira> taking and removing, is another story… and another problem.[21:18:34] <richlowe> and likely a bad idea.[21:18:51] <ira> That I'd agree with.[21:19:06] <ira> But saying "You can see the snapshot, but not look at it." just feels absurd.[21:20:09] <ira> And that issue replicates.[21:20:19] <yo61> I was talking earlier about a problem I have creating local zones[21:20:29] <yo61> I sent a mail to the list about it[21:20:57] <yo61> Anyone able to help me dig into it a bit further?[21:21:24] *** mariusp_ has joined #smartos[21:22:50] <ira> Also note: I specifically only mentioned the trigger directory, as the only way a zone could do the mount of a snapshot. ;)[21:23:14] <ira> (Via traversal only I believe the way I wrote it.)[21:24:37] *** mariusp has quit IRC[21:25:39] *** dap has quit IRC[21:26:23] <leecallen> ira: I don't understand the 'trigger directory'. Is that .zfs ?[21:26:41] <ira> .zfs/snapshot[21:26:58] <ira> When you cd from there, it actually "triggers" a mount.[21:27:21] <ira> Zones by default do not have permission to mount snapshots, so this fails.[21:28:13] <leecallen> ah. I had no idea that occurred (cd --> mount). I swear I have cd'd into snapshot directories in the zone. I will test.[21:28:56] *** mariusp_ has quit IRC[21:29:09] <ira> If the GZ has mounted them ahead of time.. it works fine.[21:29:19] <ira> This is part of why I think it is totally broken the way it is.[21:29:24] <leecallen> no, you are right, the cd fails[21:29:34] <ira> Now, goto the GZ, cd in.[21:29:39] <ira> And it'll work from the zone.[21:30:18] *** W0rmDr1nk has joined #smartos[21:31:07] <leecallen> correct again. that explains some behaviour I saw before that I thought was flaky.[21:31:13] <ira> Yeop.[21:31:30] <ira> 100% broken semantics IMHO. I have a like 5 line patch that fixes this.[21:32:08] <leecallen> fixes it how? does it allow the zones access to the snapshots?[21:32:12] <ira> Yep.[21:32:25] <leecallen> what does it patch?[21:32:31] <ira> Only via the trigger directory.[21:32:35] <ira> The platform.[21:32:50] <ira> It's pretty mild for what it does. ;)[21:32:59] <leecallen> will it be included in some future release, or do I need to 'roll my own' to get your patch?[21:33:20] <leecallen> (assuming you will share it with me)[21:33:41] <ira> I havent sent it out yet. I need to find out if upstream (Joyent/illumos) even shares my opinion on this issue.[21:33:53] <ira> I have 100% permission to share it.[21:34:06] <rmustacc> Well, you could ask developer@ or zfs@[21:34:19] <rmustacc> Or just mail a few people the details to get their take who are familiar with it.[21:34:25] <leecallen> ira would you keep me posted? I need this[21:34:39] <ira> lee, I'll gladly send you the patch we've qualified.[21:34:45] <leecallen> thanks[21:35:09] <leecallen> do you (still) have my PM ?[21:35:15] <ira> I think so…[21:35:32] <ira> It is more a matter of my pulling it out of my GIT tree.[21:35:41] <ira> I own rm stuff too.[21:35:46] <ira> owe partdon.[21:36:03] <leecallen> I would really rather you 'update' me on the status of getting it included in a future release...[21:36:21] <leecallen> if that happens I will just grab the release. else I will need to build my own smartos with your patch.[21:36:30] <ira> Oh yeah. But at the least, you should have it heaven forbig the shed stays the same color.[21:36:45] <leecallen> sure[21:37:04] <ira> And I should go through, and bug report all my "backlog".[21:37:20] <ira> It's all small annoying crud.[21:37:46] <ira> But that issue REALLY ticked off some users. ;)[21:38:21] <ira> BTW: If you expose .zfs in manta, it is also impacted by this bug of wonderfulness ;)[21:38:47] <rmustacc> I don't think we have much desire to expose .zfs for the root dataset for manta compute zones.[21:39:02] <ira> For us it is sub volumes.[21:39:22] <ira> But I cna 100% understand that.[21:41:37] *** potatosalad has quit IRC[21:41:47] <ira> You probably want to alter the commit message.[21:42:57] *** potatosalad has joined #smartos[21:43:40] <rmustacc> ira: Well, I'd rather you just got changes that seemed useful into illumos.[21:43:51] <ira> Agreed.[21:43:54] <rmustacc> So then we all benefit from it.[21:44:45] <rmustacc> If you need a hand with that part of it I'm happy to help.[21:46:13] <ira> I need to inventory it all.[21:46:49] *** dap has joined #smartos[21:47:37] <ira> I think there's only 2-3 real ones people might want. The one Lee hit is a big one of them,[21:48:06] <ira> One of them I believe is 100% unreviewable, and w/o the application using it, I can't really push it.[21:48:16] <ira> (or damn close)[21:48:43] <ira> Yeah, the rest is minor stuff, where I disagree with Joyent ;)[21:49:13] <ira> Well, there is a 3rd, but I can't prove it fixes the bug we are hitting… merely that it fixes some crappy code.[21:49:50] <rmustacc> Well, I wouldn't hold yourself back.[21:50:12] <rmustacc> Is the stuff we disagree with though in illumos?[21:50:16] <rmustacc> Or the rest of the system?[21:50:32] <ira> I ripped out drv/ipnet is the big one.[21:51:13] <ira> The rest is all minor, and really specific to what we do.[21:51:44] <ira> (And yes, I know many people can't do that. That's why I never pushed on doing it upstream.)[21:52:14] <ira> But drv/ipnet can consume your entrie machine w/o you thinking twice. :/[21:52:46] <rmustacc> Right, that needs to be fixed.[21:53:05] <ira> I should file a bug listing removing it as a work around.[21:53:38] <ira> Because that one sucked to find. Thank you richlowe for the help on that one ;)[21:54:16] *** mamash has joined #smartos[22:02:42] *** mary5030 has quit IRC[22:06:00] *** volci has quit IRC[22:14:07] *** wolfeidau has quit IRC[22:19:26] *** Alek has quit IRC[22:37:19] *** dap has quit IRC[22:47:51] *** chorrell has quit IRC[22:50:31] *** nikolam has joined #smartos[22:51:41] *** mary5030 has joined #smartos[22:54:21] <MerlinDMC> rmustacc, any hint what would make a zone fail provisioning like this: https://gist.github.com/MerlinDMC/10215843ef85861c444b ... it gets a timeout after 5 minutes waiting[22:55:47] *** volci has joined #smartos[22:58:19] <rmustacc> MerlinDMC: Why is EMI disabled?[22:58:46] <rmustacc> Oh looks like it's offline as opposed to disabled.[22:59:40] *** bens1 has joined #smartos[23:03:43] *** julian_duque has joined #smartos[23:04:00] *** julianduque has quit IRC[23:08:08] *** julian_duque is now known as julianduque[23:10:00] *** mgls has quit IRC[23:14:41] *** dap has joined #smartos[23:14:56] *** Alek has joined #smartos[23:17:42] <MerlinDMC> rmustacc, https://gist.github.com/MerlinDMC/ddc61eaef9bd5797632f ... GZ is fine, other kvm zones are running fine as well just smartos zones get stuck while provisioning[23:18:51] <MerlinDMC> I never got them stuck that early so I'm quiet out of ideas where to dig for the cause[23:23:27] *** mamash has left #smartos[23:23:51] <rmustacc> A useful little sanity test is does it work with the brand is joyent-minimal instead of joyent?[23:25:17] <MerlinDMC> rmustacc, joyent-minimal provisions fine as it seems ...[23:26:40] <rmustacc> Well, that's good at least.[23:27:14] <rmustacc> What does svcs -xv early-manifest-import say in the bad case?[23:28:59] <MerlinDMC> this: https://gist.github.com/MerlinDMC/d8207009758f165bd2de[23:29:41] <rmustacc> What about ptree?[23:30:07] <MerlinDMC> https://gist.github.com/MerlinDMC/e3108893bb4969577b6d[23:30:48] <rmustacc> Weird, seems like the smf database in the image is in bad shape for some reason.[23:30:57] *** wolstena1 has quit IRC[23:32:20] *** wolstena has joined #smartos[23:32:29] <MerlinDMC> could that happen if the GZ was upgraded from a way older version ... or is that database recreated from scratch for each zone?[23:34:28] <rmustacc> That database comes from the image for the joyent brand.[23:34:45] <MerlinDMC> I can see a non merged journal in /etc/svc after the thing goes into failed state[23:41:14] *** ilovezfs has quit IRC[23:42:34] * MerlinDMC have to get some sleep - will look into the repository.db file tomorrow[23:45:55] *** chorrell has joined #smartos[23:49:32] *** ilovezfs has joined #smartos