| vladimir3p | btw, have you tried euca utils with those? | 00:00 |
|---|---|---|
| Kiall | Yea, euca2tools works with them.. | 00:00 |
| vladimir3p | great | 00:01 |
| Kiall | The big issue with installing from source is, you can basically never be sure of a clean uninstall / upgrade / downgrade.. So if you try diablo, then try essex, then want to go back, you're screwed etc etc ;) | 00:02 |
| vladimir3p | yep, exactly. | 00:02 |
| vladimir3p | ok, thanks for all your work. Will try to take advantage of it ;-) | 00:03 |
| Kiall | yea - hopefully it wont be needed anymore soon enough ;) The ubuntu packages are finally getting fixed up :) | 00:03 |
| vladimir3p | you still need to go through the Nova configuration though... | 00:05 |
| Kiall | Yea - but that part is easy.. Once you know what knobs need turning! | 00:05 |
| vladimir3p | ok, another one for you - FIXED_RANGE_NETWROK_COUNT | 00:06 |
| vladimir3p | :-) | 00:06 |
| Kiall | WTH.That explains the error someone was telling me about the other day! | 00:07 |
| Kiall | (I've not used the scripts since I added some extra settings for that kind of stuff! Whoops...) | 00:08 |
| vladimir3p | you may consider using something like {xxx:-yyy} | 00:08 |
| vladimir3p | it will give you default values in scripts | 00:09 |
| Kiall | Yea - I could, but generally there is no need for fallbacks since there shouldnt be typos ;) | 00:10 |
| Kiall | And its a use once then forgot about it script, Only useful to get everything going the first time! | 00:11 |
| vladimir3p | yep, true | 00:12 |
| vladimir3p | ok, will continue digging in... | 00:12 |
| *** Sander^work has quit IRC | 00:18 | |
| *** Sander^work has joined #openstack-dev | 00:18 | |
| *** adjohn has quit IRC | 00:29 | |
| vladimir3p | Kiall: another quick question if you are still there... | 00:35 |
| vladimir3p | on compute node you disable dnsmasq. I'm not really familiar with this part, but is it really required? | 00:36 |
| vladimir3p | moreover, in nova-all you execute both nova-controller & nova-compute stripts (as a result dnsmasq will be disabled) | 00:38 |
| *** pixelbeat has quit IRC | 01:08 | |
| *** mwhooker has quit IRC | 01:19 | |
| *** pixelbeat has joined #openstack-dev | 01:21 | |
| *** crobinso has quit IRC | 01:23 | |
| *** andrewbogott has quit IRC | 01:33 | |
| *** pixelbeat has quit IRC | 01:35 | |
| *** lloydde has quit IRC | 01:38 | |
| *** vladimir3p has quit IRC | 01:42 | |
| *** sleepsonthefloo has quit IRC | 01:45 | |
| *** bengrue has quit IRC | 01:49 | |
| *** hugokuo has joined #openstack-dev | 02:11 | |
| *** elasticdog has quit IRC | 02:18 | |
| openstackgerrit | gholt proposed a change to openstack/swift: Work in progress for TempURL middleware. https://review.openstack.org/2246 | 02:27 |
| *** novas0x2a|laptop has quit IRC | 02:39 | |
| *** maplebed has quit IRC | 02:50 | |
| *** newnode has quit IRC | 02:53 | |
| *** nati2 has joined #openstack-dev | 03:09 | |
| *** adjohn has joined #openstack-dev | 03:14 | |
| *** sleepsonthefloo has joined #openstack-dev | 03:15 | |
| *** jog0 has quit IRC | 03:22 | |
| *** vladimir3p has joined #openstack-dev | 03:27 | |
| *** adjohn has quit IRC | 03:42 | |
| *** jeremy has quit IRC | 03:52 | |
| *** jeremy has joined #openstack-dev | 03:52 | |
| *** jeremy has joined #openstack-dev | 03:52 | |
| *** vladimir3p has quit IRC | 04:09 | |
| *** mjfork has quit IRC | 04:20 | |
| *** jdurgin has quit IRC | 04:27 | |
| *** cdub has quit IRC | 04:32 | |
| *** hugo_kuo has joined #openstack-dev | 04:34 | |
| *** hugokuo has quit IRC | 04:34 | |
| *** scottjg has joined #openstack-dev | 04:35 | |
| *** hugo_kuo has quit IRC | 04:37 | |
| *** andrewbogott has joined #openstack-dev | 04:52 | |
| *** medberry has quit IRC | 04:59 | |
| *** medberry has joined #openstack-dev | 05:00 | |
| *** medberry has quit IRC | 05:00 | |
| *** medberry has joined #openstack-dev | 05:00 | |
| *** errr has quit IRC | 05:09 | |
| *** errr has joined #openstack-dev | 05:09 | |
| *** dwalleck has joined #openstack-dev | 05:32 | |
| *** rsampaio has joined #openstack-dev | 05:33 | |
| *** dwalleck has quit IRC | 05:37 | |
| *** sleepsonthefloo has quit IRC | 05:41 | |
| *** media-786 has joined #openstack-dev | 05:50 | |
| *** hugokuo has joined #openstack-dev | 05:53 | |
| openstackgerrit | gholt proposed a change to openstack/swift: Work in progress for TempURL middleware. https://review.openstack.org/2246 | 06:03 |
| *** rsampaio has quit IRC | 06:05 | |
| *** hugokuo has quit IRC | 06:09 | |
| *** hugokuo has joined #openstack-dev | 06:10 | |
| *** hugokuo has quit IRC | 06:12 | |
| *** hugokuo has joined #openstack-dev | 06:13 | |
| *** rsampaio has joined #openstack-dev | 06:20 | |
| *** media-786 has quit IRC | 06:21 | |
| *** rsampaio has quit IRC | 06:24 | |
| *** rsampaio has joined #openstack-dev | 06:25 | |
| *** rsampaio has quit IRC | 06:28 | |
| *** zaitcev has quit IRC | 06:35 | |
| openstackgerrit | Naveed Massjouni proposed a change to openstack/nova: Renaming instance_actions.instance_id column to instance_uuid. blueprint: internal-uuids https://review.openstack.org/2244 | 07:09 |
| *** hugo_kuo has joined #openstack-dev | 07:22 | |
| *** hugokuo has quit IRC | 07:23 | |
| *** novas0x2a|laptop has joined #openstack-dev | 07:30 | |
| *** novas0x2a|laptop has quit IRC | 08:08 | |
| *** novas0x2a|laptop has joined #openstack-dev | 08:12 | |
| *** hugo_kuo has quit IRC | 08:12 | |
| *** nati2 has quit IRC | 08:17 | |
| *** scottjg has quit IRC | 08:38 | |
| *** novas0x2a|laptop has quit IRC | 09:30 | |
| *** novas0x2a|laptop has joined #openstack-dev | 09:31 | |
| *** novas0x2a|laptop has quit IRC | 11:00 | |
| *** novas0x2a|laptop has joined #openstack-dev | 11:01 | |
| *** pixelbeat has joined #openstack-dev | 11:34 | |
| *** pixelbeat has quit IRC | 11:55 | |
| *** rkukura has joined #openstack-dev | 14:02 | |
| *** pixelbeat has joined #openstack-dev | 14:20 | |
| *** mdomsch has joined #openstack-dev | 14:39 | |
| *** mdomsch has quit IRC | 15:04 | |
| *** pixelbeat has quit IRC | 15:12 | |
| Kiall | vladimir3p, nova-network manage dnsmasq instances itself, I disable the init script managed one to prevent issues.. | 15:49 |
| *** nati2 has joined #openstack-dev | 16:02 | |
| *** nati2_ has joined #openstack-dev | 16:28 | |
| *** nati2 has quit IRC | 16:30 | |
| *** nati2 has joined #openstack-dev | 16:35 | |
| *** nati2_ has quit IRC | 16:36 | |
| *** mjfork has joined #openstack-dev | 16:40 | |
| *** pixelbeat has joined #openstack-dev | 16:45 | |
| *** zns has joined #openstack-dev | 16:56 | |
| *** vladimir3p has joined #openstack-dev | 17:10 | |
| *** vladimir3p has quit IRC | 17:10 | |
| *** darthflatus has joined #openstack-dev | 17:32 | |
| *** darthflatus1 has joined #openstack-dev | 17:38 | |
| *** darthflatus has quit IRC | 17:41 | |
| *** zaitcev has joined #openstack-dev | 17:50 | |
| *** Mandell has quit IRC | 17:56 | |
| openstackgerrit | p-draigbrady proposed a change to openstack/nova: Bug#898257 support handling images with libguestfs https://review.openstack.org/1994 | 17:59 |
| openstackgerrit | p-draigbrady proposed a change to openstack/nova: Bug#898257 abstract out disk image access methods https://review.openstack.org/1993 | 17:59 |
| *** Mandell has joined #openstack-dev | 18:07 | |
| *** pixelbeat has quit IRC | 18:07 | |
| zaitcev | I learned two lessons. | 18:13 |
| zaitcev | #1. Do not run Swift proxy server on a VM with less than 1GB RAM. It just won't work - even on a testbed with no load! | 18:14 |
| zaitcev | #2. Swift on ext4 does not work. The objects are stored fine, but updaters run rsync for replication and it fails to use EA in ext4 and goes kaputt, so replication does not work. | 18:15 |
| notmyname | zaitcev: rsync wasn't copying the xattrs on ext4? | 18:20 |
| zaitcev | notmyname: That's correct. | 18:21 |
| notmyname | zaitcev: interesting. I've not tested swift with ext4 | 18:21 |
| notmyname | and I've got my swift dev env on a 1GB VM. proxy, storage, and background services. it works fine, so I'm not sure why you weren't getting the proxy to work with 1GB of memory just for it | 18:22 |
| zaitcev | notmyname: it is not blind, so not difficult to discover. there's a very obvious message in logs | 18:22 |
| zaitcev | 2011/12/09 19:32:54 [16105] rsync: rsync_xal_set: lsetxattr("a41/255e1882eaced58a011ef64378bb5a41/.1322692416.79853.data.hmjuc9","user.swift.metadata") failed: Operation not supported (95) | 18:22 |
| notmyname | interesting | 18:23 |
| zaitcev | It's RHEL 6 (fork off Fedora 13) | 18:23 |
| notmyname | zaitcev: can you add a bug in launchpad with that message (and that it's on ext4)? it will be pretty low priority, but it will be good to document | 18:23 |
| zaitcev | notmyname: I can, but wouldn't it be better to file a bug against rsync at Samba.org somewhere? | 18:24 |
| notmyname | zaitcev: ah. probably so :-) | 18:24 |
| zaitcev | ok | 18:24 |
| notmyname | zaitcev: if it's something that can be fixed with rsync options (ie inside swift), then we should have it documented. if it's entirely with rsync, then obviously it needs to go there | 18:25 |
| notmyname | zaitcev: thanks | 18:25 |
| zaitcev | notmyname: About the proxy, 1GB works, but on 512 MB it OOMs | 18:26 |
| zaitcev | I am quite shocked that essentially idle node can do that, but whatever. Modern computing in action. | 18:26 |
| zaitcev | Truth to be told, I had it co-located with auth node, so maybe it's Keystone actually | 18:27 |
| *** novas0x2a|laptop has quit IRC | 18:29 | |
| *** adjohn has joined #openstack-dev | 18:31 | |
| zns | zaitcev: Unless I don't understand the whole issue, I don't see how Keystone could impact this (it's pretty self-contained). But if I can help let me know... | 18:32 |
| zaitcev | zns: It looks like every Python process settles at about the same amount of RSS, which is quite high in my case - they all have about 16 MB. Piling them up causes the box to OOM due to number of them. | 18:33 |
| zns | zaitcev: got it. I also know that using keystone-control (copied from glance-control I think) tries to allocate extra memory. So if you're starting Keystone with that... | 18:35 |
| zaitcev | yeah, some straw broke camel's back | 18:35 |
| notmyname | zaitcev: FWIW, on my 1GB dev instance: http://paste.openstack.org/show/3747/ | 18:36 |
| *** GheRivero_ has joined #openstack-dev | 18:39 | |
| *** jakedahn has joined #openstack-dev | 18:54 | |
| *** jmckenty has joined #openstack-dev | 18:57 | |
| *** adjohn has quit IRC | 19:07 | |
| *** adjohn has joined #openstack-dev | 19:31 | |
| *** adjohn has quit IRC | 19:36 | |
| openstackgerrit | Jake Dahn proposed a change to openstack/horizon: Merging Instances and Volumes into a single panel. https://review.openstack.org/2152 | 19:41 |
| *** jmckenty has quit IRC | 19:45 | |
| *** GheRivero_ has quit IRC | 19:46 | |
| *** jmckenty has joined #openstack-dev | 19:49 | |
| *** Mandell has quit IRC | 20:01 | |
| *** rsampaio has joined #openstack-dev | 20:12 | |
| *** maplebed has joined #openstack-dev | 21:01 | |
| *** Mandell has joined #openstack-dev | 21:02 | |
| *** scottjg has joined #openstack-dev | 21:24 | |
| *** rsampaio has quit IRC | 21:27 | |
| *** maplebed has quit IRC | 21:28 | |
| *** jakedahn has quit IRC | 21:28 | |
| *** jakedahn has joined #openstack-dev | 21:41 | |
| *** nati2 has quit IRC | 21:48 | |
| openstackgerrit | anotherjesse proposed a change to openstack-dev/devstack: move back to nova.sh method of configuring screen https://review.openstack.org/2247 | 21:55 |
| openstackgerrit | anotherjesse proposed a change to openstack-dev/devstack: move back to nova.sh method of configuring screen https://review.openstack.org/2247 | 21:57 |
| openstackgerrit | Vish Ishaya proposed a change to openstack/nova: Moves find config to utils because it is useful https://review.openstack.org/2248 | 22:03 |
| *** rkukura has quit IRC | 22:05 | |
| openstackgerrit | Edgar Magana Perdomo proposed a change to openstack/quantum: blueprint services-insertion-wrapper https://review.openstack.org/2249 | 22:09 |
| openstackgerrit | Edgar Magana Perdomo proposed a change to openstack/quantum: blueprint services-insertion-wrapper https://review.openstack.org/2249 | 22:16 |
| *** jmckenty has quit IRC | 22:25 | |
| *** adjohn has joined #openstack-dev | 22:35 | |
| *** jmckenty has joined #openstack-dev | 22:38 | |
| *** adjohn has quit IRC | 22:54 | |
| openstackgerrit | Edgar Magana Perdomo proposed a change to openstack/quantum: blueprint services-insertion-wrapper https://review.openstack.org/2249 | 23:16 |
| openstackgerrit | Edgar Magana Perdomo proposed a change to openstack/quantum: blueprint services-insertion-wrapper https://review.openstack.org/2249 | 23:18 |
| *** jakedahn has quit IRC | 23:37 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!