| *** adiantum has quit IRC | 00:01 | |
| *** adiantum has joined #openstack | 00:06 | |
| *** Edwin__ has quit IRC | 00:29 | |
| *** adiantum has quit IRC | 00:36 | |
| *** Ryan_Lane|away is now known as Ryan_Lane | 00:37 | |
| *** thatsdone has joined #openstack | 00:46 | |
| *** adiantum has joined #openstack | 00:48 | |
| *** zul has joined #openstack | 01:08 | |
| *** miclorb_ has quit IRC | 01:09 | |
| *** miclorb_ has joined #openstack | 01:09 | |
| *** trin_cz has quit IRC | 01:34 | |
| *** adiantum has quit IRC | 01:38 | |
| *** adiantum has joined #openstack | 01:40 | |
| *** dirakx has joined #openstack | 01:44 | |
| *** westmaas1 has joined #openstack | 01:59 | |
| *** adiantum has quit IRC | 02:08 | |
| *** drico has quit IRC | 02:14 | |
| *** adiantum has joined #openstack | 02:20 | |
| *** drico has joined #openstack | 02:21 | |
| *** chilts has quit IRC | 02:22 | |
| *** jfluhmann_ has quit IRC | 02:28 | |
| *** kashyapc has joined #openstack | 02:40 | |
| *** jfluhmann_ has joined #openstack | 02:40 | |
| *** DubLo7 has left #openstack | 03:03 | |
| *** HugoKuo has joined #openstack | 03:04 | |
| *** masumotok has quit IRC | 03:23 | |
| *** adiantum has quit IRC | 03:38 | |
| *** adiantum has joined #openstack | 03:50 | |
| *** drico has quit IRC | 04:17 | |
| *** drico has joined #openstack | 04:21 | |
| *** Cybodog has quit IRC | 04:25 | |
| *** chilts has joined #openstack | 04:39 | |
| *** omidhdl has joined #openstack | 05:18 | |
| *** adiantum has quit IRC | 05:38 | |
| *** kashyapc has quit IRC | 05:40 | |
| *** adiantum has joined #openstack | 05:44 | |
| *** f4m8_ is now known as f4m8 | 05:45 | |
| *** f4m8 is now known as f4m8_ | 05:46 | |
| *** mdomsch has joined #openstack | 05:48 | |
| *** adiantum has quit IRC | 05:51 | |
| *** adiantum has joined #openstack | 05:52 | |
| *** drico has quit IRC | 06:19 | |
| *** drico has joined #openstack | 06:21 | |
| *** ramkrsna has joined #openstack | 06:21 | |
| *** mdomsch has quit IRC | 06:24 | |
| *** miclorb_ has quit IRC | 06:41 | |
| *** thatsdon_ has joined #openstack | 06:47 | |
| *** thatsdone has quit IRC | 06:48 | |
| *** perestre1ka has quit IRC | 07:01 | |
| *** adiantum has quit IRC | 07:03 | |
| *** adiantum has joined #openstack | 07:04 | |
| *** drico has quit IRC | 07:19 | |
| *** drico has joined #openstack | 07:21 | |
| *** CloudChris has joined #openstack | 07:26 | |
| *** CloudChris has quit IRC | 07:35 | |
| *** adiantum has quit IRC | 07:35 | |
| *** adiantum has joined #openstack | 07:40 | |
| *** rcc has joined #openstack | 07:40 | |
| openstackhudson | Project nova-tarmac build #53,149: FAILURE in 1.3 sec: http://hudson.openstack.org/job/nova-tarmac/53149/ | 07:40 |
|---|---|---|
| *** guigui has joined #openstack | 07:41 | |
| openstackhudson | Project nova-tarmac build #53,150: STILL FAILING in 1.4 sec: http://hudson.openstack.org/job/nova-tarmac/53150/ | 07:45 |
| *** miclorb_ has joined #openstack | 07:49 | |
| openstackhudson | Project nova-tarmac build #53,151: STILL FAILING in 1.4 sec: http://hudson.openstack.org/job/nova-tarmac/53151/ | 07:50 |
| openstackhudson | Yippie, build fixed! | 07:55 |
| openstackhudson | Project nova-tarmac build #53,152: FIXED in 2.3 sec: http://hudson.openstack.org/job/nova-tarmac/53152/ | 07:55 |
| *** brd_from_italy has joined #openstack | 08:01 | |
| *** guynaor has joined #openstack | 08:03 | |
| *** aimon has quit IRC | 08:04 | |
| *** aimon has joined #openstack | 08:05 | |
| *** adiantum has quit IRC | 08:05 | |
| *** aimon has quit IRC | 08:19 | |
| *** drico has quit IRC | 08:19 | |
| *** aimon has joined #openstack | 08:20 | |
| *** drico has joined #openstack | 08:21 | |
| *** fabiand_ has joined #openstack | 08:27 | |
| *** allsystemsarego has joined #openstack | 08:30 | |
| *** allsystemsarego has joined #openstack | 08:30 | |
| *** DigitalFlux has joined #openstack | 08:36 | |
| ttx | DigitalFlux: about package install failing... what packages are you installing ? | 08:36 |
| DigitalFlux | Hi | 08:37 |
| DigitalFlux | ttx: you are monitoring the channel logs :) | 08:37 |
| DigitalFlux | Cool | 08:37 |
| DigitalFlux | nova-compute is failing .. | 08:37 |
| DigitalFlux | ttx: Here is a full log | 08:38 |
| DigitalFlux | http://pastebin.ubuntu.com/555006/ | 08:38 |
| ttx | ok, so maverick / 2011.1~bzr569 | 08:38 |
| DigitalFlux | Yup | 08:39 |
| DigitalFlux | annegentle: there ? | 08:40 |
| *** calavera has joined #openstack | 08:40 | |
| ttx | looks like a libvirt0 install issue, not a nova one | 08:40 |
| ttx | but the end of your logs seem to say it's ok now ? | 08:41 |
| DigitalFlux | Yeah, but this is after resolving those issues | 08:46 |
| DigitalFlux | by answering those bunch of aptitude questions | 08:46 |
| DigitalFlux | I need to automate the installation non-interactively .. | 08:46 |
| ttx | it was a libvirt0 install error (bad deps), can you reproduce it on a new system ? | 08:47 |
| DigitalFlux | I actually had this on two different systems | 08:48 |
| DigitalFlux | But OK, trying with another new one .. | 08:48 |
| ttx | soren: ping me when around, I need your help for the final branch push | 08:48 |
| ttx | DigitalFlux: try installing libvirt0 first to see if you reproduce the issue outside nova | 08:49 |
| DigitalFlux | ttx: I see, OK will do | 08:49 |
| *** jfluhmann__ has joined #openstack | 08:56 | |
| *** kashyapc has joined #openstack | 08:58 | |
| *** jfluhmann_ has quit IRC | 09:00 | |
| *** pcoca has joined #openstack | 09:13 | |
| *** MarkAtwood has joined #openstack | 09:13 | |
| *** drico has quit IRC | 09:18 | |
| *** drico has joined #openstack | 09:21 | |
| *** masumotok has joined #openstack | 09:22 | |
| *** masumotok has quit IRC | 09:24 | |
| *** MarkAtwood has quit IRC | 09:40 | |
| *** MarkAtwood has joined #openstack | 09:45 | |
| *** thatsdone has joined #openstack | 09:47 | |
| *** thatsdon_ has quit IRC | 09:49 | |
| DigitalFlux | ttx: seemed like a problem with cached packages .. | 10:01 |
| DigitalFlux | ttx: installing fine on a different node .. | 10:01 |
| ttx | DigitalFlux: ok | 10:02 |
| *** ramkrsna has quit IRC | 10:11 | |
| soren | ttx: what's up? | 10:12 |
| *** drico_ has joined #openstack | 10:21 | |
| *** drico has quit IRC | 10:23 | |
| *** ramkrsna has joined #openstack | 10:24 | |
| *** perestrelka has joined #openstack | 10:34 | |
| *** miclorb_ has quit IRC | 10:39 | |
| *** openstackhudson has quit IRC | 10:46 | |
| *** openstackhudson has joined #openstack | 10:47 | |
| *** pcoca has quit IRC | 11:15 | |
| *** adiantum has joined #openstack | 11:20 | |
| *** kashyapc has quit IRC | 11:21 | |
| *** kashyapc has joined #openstack | 11:22 | |
| *** MarkAtwood has quit IRC | 11:26 | |
| *** adiantum has quit IRC | 11:31 | |
| *** adiantum has joined #openstack | 11:37 | |
| *** adiantum has quit IRC | 11:42 | |
| *** f4m8_ is now known as f4m8 | 11:45 | |
| *** thatsdone has quit IRC | 11:46 | |
| sandywalsh | o/ | 11:48 |
| *** arthurc has joined #openstack | 11:58 | |
| rcc | hi guys, just started to fiddle with openstack wiki and source code. I'd like to contribute, any easy bug from nova you might suggest to start with? | 12:06 |
| *** DigitalFlux has quit IRC | 12:12 | |
| *** DigitalFlux has joined #openstack | 12:12 | |
| *** omidhdl has left #openstack | 12:24 | |
| * ttx looks | 12:33 | |
| *** ctennis has quit IRC | 12:33 | |
| ttx | rcc: maybe bug 681417 | 12:34 |
| uvirtbot | Launchpad bug 681417 in nova "All errors about quotas should include a unit" [Low,Triaged] https://launchpad.net/bugs/681417 | 12:34 |
| ttx | soren: we need another review pass on https://code.launchpad.net/~nttdata/nova/live-migration/+merge/44940 | 12:35 |
| soren | ttx: I know. I'm procrastinating. | 12:36 |
| ttx | with an assessment of how far it seems to be from being mergeable, so that we can give masumotok some FFe answer | 12:36 |
| ttx | if there is no chance that can get merged in time, no need to spend too much review effort on it, better defer | 12:37 |
| ttx | soren: ^ | 12:37 |
| soren | ttx: python-glance installed on Hudson. | 12:38 |
| * soren retries https://code.launchpad.net/~rconradharris/nova/xs-snap-return-image-id-before-snapshot/+merge/45494 | 12:38 | |
| *** openstackhudson has quit IRC | 12:44 | |
| *** openstackhudson has joined #openstack | 12:45 | |
| *** jfluhmann__ has quit IRC | 12:46 | |
| *** ctennis has joined #openstack | 12:48 | |
| rcc | thx ttx, i'll have a look | 12:56 |
| *** dendrobates is now known as dendro-afk | 13:02 | |
| *** adiantum has joined #openstack | 13:11 | |
| * annegentle is up and at 'em. Whoever 'em are. Bring 'em. :) | 13:22 | |
| soren | annegentle: That's the spirit! :) | 13:25 |
| annegentle | I've worked through a proof of concept process for some documents to populate docs.openstack.org with docbook as source, outputting their "webhelp" format - woohoo. And whew. I'm gonna make the Bexar deadline. | 13:25 |
| annegentle | I'll have a site up for people to comment on, critique, test, whatnot, in the next week or so. | 13:26 |
| *** DubLo7 has joined #openstack | 13:32 | |
| *** lvaughn has quit IRC | 13:32 | |
| *** EdwinGrubbs has joined #openstack | 13:37 | |
| *** EdwinGrubbs has joined #openstack | 13:37 | |
| *** thatsdone has joined #openstack | 13:38 | |
| *** hadrian has joined #openstack | 13:38 | |
| *** pvo_away is now known as pvo | 13:41 | |
| sandywalsh | ttx, are the last of the merges done now? | 13:52 |
| sandywalsh | (feature merges) | 13:52 |
| *** ramkrsna_ has joined #openstack | 13:54 | |
| ttx | sandywalsh: no | 13:54 |
| ttx | sandywalsh: one more rinse necessary for xs-snapshots | 13:55 |
| ttx | same for pastedeploy | 13:55 |
| sandywalsh | ttx is xenapi-glance-2 still on the table? | 13:55 |
| ttx | db-migration is pending on live-migration | 13:55 |
| ttx | xenapi-glance and live-migration are still on the hook | 13:56 |
| *** masumotok has joined #openstack | 13:56 | |
| ttx | if those two aren't closer to getting merged today, I'll probably propose to defer them. | 13:56 |
| thatsdone | masumotok posted merge request again today (in Japan) | 13:56 |
| ttx | so those last two need all the reviews they can get. | 13:56 |
| *** ramkrsna has quit IRC | 13:57 | |
| sandywalsh | ok, I'll continue on glance, thx | 13:57 |
| ttx | thatsdone: yes, I asked nova-core people to rereview them to assess how far it is from being ready | 13:57 |
| thatsdone | thanks! ttx | 13:57 |
| masumotok | ttx: anyway, please let me know if I can do something | 13:59 |
| ttx | masumotok: soren told me he would rereview soon, I'll also ask vish when he arrives. | 13:59 |
| ttx | any other nova-core person is welcome to review as well. | 14:00 |
| *** ramkrsna has joined #openstack | 14:01 | |
| *** fabiand_ has quit IRC | 14:03 | |
| *** ramkrsna_ has quit IRC | 14:04 | |
| *** spectorclan has joined #openstack | 14:31 | |
| uvirtbot | New bug: #703972 in nova "Nova instance names don't carry to XenAPI instance name" [Undecided,New] https://launchpad.net/bugs/703972 | 14:31 |
| *** jdarcy has joined #openstack | 14:32 | |
| *** pvo is now known as pvo_away | 14:36 | |
| soren | I'm using st to upload a bunch of files (There's 2662 files on the command line, and I'm specifying -c... There are about 800 left to go). It starts out decently (CPU usage of ~7%), but after a while it shoots up to 100% CPU usage, and if I wait even longer, it ends up eating 150% (or however much it can get its hands on). Does this sound familiar at all? | 14:41 |
| *** gondoi has joined #openstack | 14:45 | |
| soren | Hm... One of the threads seems stuck in a tight loop calling read(2) on one of the files, but is getting an empty string back. | 14:45 |
| soren | There seems to be some EOF handling somewhere that needs work. | 14:46 |
| * soren digs | 14:46 | |
| sandywalsh | anyone know how to do the equivalent of the euca-bundle-image/euca-upload-bundle/euca-register dance with glance? | 14:51 |
| *** aliguori has joined #openstack | 14:51 | |
| *** EdwinGrubbs has quit IRC | 14:53 | |
| *** pvo_away is now known as pvo | 14:54 | |
| *** ramkrsna has quit IRC | 14:57 | |
| *** fabiand_ has joined #openstack | 14:57 | |
| *** DigitalFlux has quit IRC | 14:58 | |
| *** guynaor has left #openstack | 14:59 | |
| *** f4m8 is now known as f4m8_ | 15:01 | |
| *** EdwinGrubbs has joined #openstack | 15:02 | |
| *** EdwinGrubbs has joined #openstack | 15:02 | |
| *** lvaughn has joined #openstack | 15:04 | |
| ttx | sirp-: ping | 15:06 |
| *** mdomsch has joined #openstack | 15:06 | |
| *** Ryan_Lane has quit IRC | 15:08 | |
| *** lvaughn has quit IRC | 15:08 | |
| *** lvaughn has joined #openstack | 15:09 | |
| *** jaypipes has joined #openstack | 15:15 | |
| *** j05h has joined #openstack | 15:17 | |
| *** mjmac_ is now known as mjmac | 15:27 | |
| *** johnpur has joined #openstack | 15:32 | |
| *** ChanServ sets mode: +v johnpur | 15:32 | |
| *** DubLo7 has left #openstack | 15:37 | |
| soren | Hm... Ok, instrumenting the put_object function a bit, I can see that it ends up in a state where left=31490737 (the full length of the file), size=65536, it tries to read size bytes, but gets 0. Looking at /proc/<pid>/fdinfo/<fd>, I can see the file pointer is actually pointed at EOF. | 15:38 |
| *** pvo is now known as pvo_away | 15:39 | |
| soren | I'm not sure how this could have happened. | 15:39 |
| soren | Yet, at least. | 15:39 |
| soren | Hmm... Except, of course, if the contents object passed in somehow got fast-forwarded to EOF before it even reached put_object. | 15:41 |
| *** pharkmillups has joined #openstack | 15:47 | |
| *** pvo_away is now known as pvo | 15:47 | |
| *** pvo is now known as pvo_away | 15:48 | |
| *** pvo_away is now known as pvo | 15:48 | |
| *** GaraDE has joined #openstack | 15:54 | |
| *** drico_ has quit IRC | 16:01 | |
| *** guigui has quit IRC | 16:01 | |
| *** clayg_ has joined #openstack | 16:01 | |
| creiht | soren: that is odd | 16:01 |
| creiht | haven't seen that before | 16:01 |
| creiht | though, I admit that I have only used st for small groups of files | 16:02 |
| *** dragondm has joined #openstack | 16:02 | |
| gholt | soren: That is a bit odd. Do you have files that change size? | 16:05 |
| gholt | I've never playing around with st in that sort of scenario. | 16:05 |
| *** GaraDE has quit IRC | 16:05 | |
| *** mdomsch has quit IRC | 16:06 | |
| gholt | But yeah, I could put in a bit better error checking there where if the read gets 0 it throws an exception. | 16:06 |
| *** dabo- is now known as dabo | 16:07 | |
| *** fabiand_ has quit IRC | 16:10 | |
| uvirtbot | New bug: #704015 in swift "st and client.py's reads aren't checking for EOF condition" [Undecided,New] https://launchpad.net/bugs/704015 | 16:16 |
| *** pvo is now known as pvo_away | 16:22 | |
| jaypipes | sirp-: around? | 16:30 |
| sirp- | jaypipes: yep | 16:31 |
| jaypipes | sirp-: heya, ttx and I are trying to get xenapi-glance-2 merged, but we're stuck on https://code.launchpad.net/~rconradharris/nova/xs-snap-return-image-id-before-snapshot/+merge/45494 | 16:31 |
| jaypipes | sirp-: could you fix the merge conflict there and re-push? | 16:31 |
| jaypipes | sirp-: glance is now installed on Hudson and (hopefully!) after resolving that merge conflict in compute/api.py, it should go through... | 16:32 |
| sirp- | jaypipe: yeah ttx pinged me on that just a bit ago, will do | 16:32 |
| *** drico has joined #openstack | 16:34 | |
| jaypipes | sirp-: cheers :) | 16:34 |
| *** kashyapc has quit IRC | 16:34 | |
| *** calavera has quit IRC | 16:35 | |
| *** mdomsch has joined #openstack | 16:35 | |
| *** drico has quit IRC | 16:38 | |
| *** kashyapc has joined #openstack | 16:40 | |
| *** thatsdone has quit IRC | 16:46 | |
| *** brd_from_italy has quit IRC | 16:50 | |
| *** meatz has joined #openstack | 16:56 | |
| *** meatz has quit IRC | 16:59 | |
| uvirtbot | New bug: #704038 in glance "Glance RegistryClient only connects to 0.0.0.0:9292" [High,Confirmed] https://launchpad.net/bugs/704038 | 17:01 |
| *** mdomsch has quit IRC | 17:04 | |
| *** drico has joined #openstack | 17:07 | |
| *** drico has quit IRC | 17:15 | |
| *** ibarrera has quit IRC | 17:16 | |
| sirp- | jaypipes: fyi fixed conflicts and re-pushed xs-snaps | 17:23 |
| jaypipes | sirp-: rock. | 17:24 |
| *** fabiand_ has joined #openstack | 17:25 | |
| *** kashyapc has quit IRC | 17:25 | |
| *** kashyapc has joined #openstack | 17:25 | |
| *** mdomsch has joined #openstack | 17:28 | |
| *** drico has joined #openstack | 17:38 | |
| *** dubsquared has joined #openstack | 17:59 | |
| *** aimon has quit IRC | 18:07 | |
| *** aimon has joined #openstack | 18:08 | |
| sandywalsh | jaypipes, I have a lucid tarball that has the kernel, ramdisk and machine image in it ... is this the same format I should use for glance? | 18:17 |
| openstackhudson | Project nova build #410: SUCCESS in 1 min 30 sec: http://hudson.openstack.org/job/nova/410/ | 18:17 |
| openstackhudson | Tarmac: The Openstack API requires image metadata to be returned immediately after an image-create call. | 18:17 |
| openstackhudson | This is accomplished by having the ImageService create a 'queued' image in Glance. | 18:17 |
| openstackhudson | When the image is subsequently uploaded, the image will go from 'queued' -> 'saving' -> 'queued'. | 18:17 |
| openstackhudson | Related Future Work: | 18:17 |
| openstackhudson | The ImageService needs to be cleaned up so that there is a canonical set of attributes (id, status, etc), and a canonical set of values ('queued', 'saving', etc). Right now, EC2 is fairly coupled to LocalImageService and S3ImageService while OpenStackAPI is coupled to GlanceImageService; ideally, we should be able mix-and-match from any of these. | 18:17 |
| sandywalsh | sirp do I need xs_snapshot to create images first, before I can use xenapi-glance? | 18:17 |
| sandywalsh | sirp- ^ | 18:18 |
| jaypipes | sirp-: hallelujah. xs-snapshots are in trunk. | 18:18 |
| sandywalsh | \o/ | 18:18 |
| sandywalsh | which makes my previous question all the more timely | 18:19 |
| jaypipes | sandywalsh: currently, you can store each of the kernel, ramdisk, and machine image in Glance separately. The discussion on the ML recently hints that we are moving towards getting rid of that and instead having Glance store the disk_format and appliance_format instead of the "type" (which is very AMI specific) | 18:20 |
| sirp- | jaypipes: awesome | 18:20 |
| *** arthurc has quit IRC | 18:21 | |
| sandywalsh | jaypipes so I should probably take my current images, create an instance and use trunk (now) to move the snapshot into glance. | 18:21 |
| jaypipes | sandywalsh: totally depends what you are doing with it... I'm not familiar with the work you are currently doing. Could you elaborate, please? | 18:22 |
| sandywalsh | well, I'm trying to review xenapi-glance, but part of that requires getting images to work with. | 18:22 |
| sandywalsh | and getting them into glance | 18:22 |
| jaypipes | sandywalsh: right, gotcha. Yes, I believe ewan/salvatore need to merge in trunk again and make a couple changes... | 18:25 |
| sandywalsh | jaypipes, but my plan seems reasonable? | 18:25 |
| jaypipes | sandywalsh: so, yes, if you are trying to review/test, you should merge trunk into xenapi-glance-2 and use snapshotting to do images. | 18:25 |
| jaypipes | sandywalsh: yep | 18:25 |
| sandywalsh | jaypipes, groovy ... thanks! | 18:25 |
| jaypipes | mtaylor: heya, is the latest Nova build failure (http://launchpadlibrarian.net/62375654/buildlog_ubuntu-maverick-i386.nova_2011.1~bzr570-0ubuntu0ppa1~maverick1_FAILEDTOBUILD.txt.gz) related to this bug? https://bugs.launchpad.net/glance/+bug/703375 | 18:28 |
| uvirtbot | Launchpad bug 703375 in glance "glance test suite can't run in the context of a debian build" [Undecided,New] | 18:28 |
| *** pharkmillups has quit IRC | 18:28 | |
| *** dfg_ has joined #openstack | 18:31 | |
| *** matclayton has joined #openstack | 18:46 | |
| *** fabiand_ has quit IRC | 18:48 | |
| sandywalsh | so will trunk use glance by default now, or does it need to be enabled? | 18:49 |
| *** fabiand_ has joined #openstack | 18:50 | |
| openstackhudson | Project nova build #411: SUCCESS in 1 min 28 sec: http://hudson.openstack.org/job/nova/411/ | 18:52 |
| openstackhudson | Tarmac: Further decouple api routing decisions and move into paste.deploy configuration. This makes paste back the nova-api binary. | 18:52 |
| fabiand_ | in nova, guest images are retrieved from an object store, right? | 18:52 |
| sirp- | sandywalsh: the default ImageService (as dictated by FLAGS) is till S3ImageService | 18:52 |
| sirp- | fabiand_: right, we have two mechanisms right now, an S3-like mechanism that uses nova-objectstore, and the new way (still under development) that uses an image service (Glance) | 18:53 |
| sandywalsh | sirp, so for me to get a snapshot I'll have to 1. create the instance with objectstore. 2. shutdown 3. switch S3ImageService=Glance (sic?) and 4. do a snapshot, correct? | 18:54 |
| sandywalsh | sirp-, assuming I have glance running | 18:55 |
| sirp- | sandywalsh: exactly, that's what i've been doing (not ideal i know!) | 18:55 |
| sandywalsh | sirp-, that's ok ... so long as I'm on the right path :) | 18:56 |
| *** drico has quit IRC | 18:57 | |
| fabiand_ | sirp-: thanks. so. if i retrieve an image via glance, it get's saved on the local storage and run by nova? | 18:57 |
| *** gondoi has quit IRC | 18:59 | |
| sirp- | fabiand_: right now, you'll want to use objectstore for creating instances (image retreival in glance still needs some work); but that's right; glance fetches the image on to the host machine and then nova then boots the image from there | 18:59 |
| fabiand_ | sirp-: nice :) so what can i do to save an image? | 18:59 |
| fabiand_ | push it back to glance? | 19:00 |
| sirp- | fabiand_: saving an image, at this point, requires you to use XenServer (we haven't built the libvirt version) yet; but assuming you're using XS, you can use the OpenStack API to create an 'instance-snapshot'… this can be done on the CLI using the cloudservers tool: cloudservers image-create <instance_id> "pretty name" | 19:01 |
| fabiand_ | ah okay. great ... so what happens to this snapshot? :) how is it stored? (i am asking ths questions, because i could not find them in the docs) | 19:03 |
| ttx | jaypipes, soren: could you (or another nova-core dude) flip the approved bit on https://code.launchpad.net/~anso/nova/factorycleanup/+merge/45426 ? Now that wsgirouter is merged it should be ok to go | 19:03 |
| *** EdwinGrubbs has quit IRC | 19:04 | |
| sirp- | fabiand_: in the case of XS, the disk-format is VHD, so we bundle up the VHDs (if there is more than one), and stream that into Glance (using a ReST insterface). Glance, in turn, streams that into a backend which could be S3, Swift, or the local filesystem on the Glance machine | 19:05 |
| *** befreax has joined #openstack | 19:06 | |
| *** drico has joined #openstack | 19:07 | |
| ttx | jaypipes, soren; forget about it, xtoddx is on it | 19:08 |
| fabiand_ | sirp-: ah okay ... thanks for that overview :) | 19:08 |
| *** gondoi has joined #openstack | 19:08 | |
| sirp- | fabiand_: np :) | 19:09 |
| *** matclayton has left #openstack | 19:11 | |
| *** drico has quit IRC | 19:16 | |
| *** gondoi has quit IRC | 19:18 | |
| *** gondoi has joined #openstack | 19:18 | |
| *** m_3 has quit IRC | 19:20 | |
| openstackhudson | Project nova build #412: SUCCESS in 1 min 26 sec: http://hudson.openstack.org/job/nova/412/ | 19:22 |
| openstackhudson | Tarmac: Change where paste.deploy factories live and how they are called. They are now in the nova.wsgi.Application/Middleware classes, and call the __init__ method of their class with kwargs of the local configuration of the paste file. | 19:22 |
| ttx | \o/ | 19:23 |
| *** m_3 has joined #openstack | 19:25 | |
| xtoddx | ttx: did that get in? Merge prop still shows "Needs Review" | 19:26 |
| ttx | mtaylor: Hudson/Tarmac got confused about the status of https://code.launchpad.net/~anso/nova/factorycleanup/+merge/45426, it was merged but it set it to Needs Review | 19:26 |
| xtoddx | ttx: should I just set it to "Merged" | 19:26 |
| ttx | xtoddx: I think so, check out http://bazaar.launchpad.net/~hudson-openstack/nova/trunk/revision/572 | 19:27 |
| xtoddx | victory! | 19:27 |
| ttx | xtoddx: thanks ! | 19:27 |
| vishy | ttx: I think we're good on live-migrations. I only see a couple changes needed. I'll let soren chime in | 19:28 |
| ttx | vishy: cool | 19:28 |
| *** adiantum has quit IRC | 19:35 | |
| *** pvo_away is now known as pvo | 19:36 | |
| *** drico has joined #openstack | 19:39 | |
| *** EdwinGrubbs has joined #openstack | 19:45 | |
| *** EdwinGrubbs has joined #openstack | 19:45 | |
| *** MarkAtwood has joined #openstack | 19:46 | |
| mtaylor | ttx: weird | 19:47 |
| soren | vishy: I'm going to look at it once more real soon now. | 19:55 |
| soren | gholt, creiht: No changing of file sizes. However, the fact that the `left` variable in that loop is equal to the size of the file, yet the file pointer is already at the end, suggests something else is afoot. | 19:57 |
| *** Pentheus has joined #openstack | 19:58 | |
| *** drico has quit IRC | 20:03 | |
| Jbain | little offtopic but related: anyone know if there is a /proc/ location where I can find ethernet bridge info? similar to the output of 'brctl show' but hopefully a bit more parseable? | 20:08 |
| *** befreax has quit IRC | 20:09 | |
| dubsquared | not sure, but /proc/net would be my guess? | 20:10 |
| dubsquared | that's where bonding stuff is... | 20:10 |
| dubsquared | shurg | 20:10 |
| *** rcc has quit IRC | 20:11 | |
| *** Ryan_Lane has joined #openstack | 20:13 | |
| fabiand_ | Jbain: /proc/net/bonding/ is bonding.kmod is loaded | 20:14 |
| Jbain | I don't have /proc/net/bonding =\ | 20:15 |
| fabiand_ | sudo modprobe bonding | 20:15 |
| Jbain | well, I'm also not doing any bonding | 20:16 |
| Jbain | I'm really just wanting info on the bridges and vif's that xen is creating | 20:16 |
| fabiand_ | Jbain: argh - i swapped some words :) /sys/class/net/ might help ya | 20:17 |
| *** Ryan_Lane_ has joined #openstack | 20:18 | |
| Jbain | oooh, there's some stuff | 20:18 |
| Jbain | thanks, i'll poke around there a bit :) | 20:19 |
| *** Ryan_Lane has quit IRC | 20:19 | |
| *** dirakx has quit IRC | 20:19 | |
| tr3buchet | vishy: what's the issue with the race condition when associating multiple networks with one project? | 20:20 |
| *** ctennis has quit IRC | 20:21 | |
| *** Ryan_Lane_ is now known as Ryan_Lane | 20:21 | |
| Jbain | fabiand_: this is quite helpful, thanks! | 20:21 |
| fabiand_ | Jbain: np :) | 20:21 |
| soren | Jbain: What exactly are you trying to do? | 20:24 |
| fraggeln | soren: bonding? :) | 20:25 |
| Jbain | well, it started out as a quick script to help me make VM configs for Xen more quickly | 20:25 |
| Jbain | it has evolved into writing a bunch of perl modules for managing all this stuff | 20:26 |
| vishy | tr3buchet: are you referring to one of my code comments? | 20:26 |
| Jbain | right now I'm looking into verifying bridge/vif associations and creating private bridges for VM teams | 20:27 |
| tr3buchet | vishy: yep nova.db.sqlalchemy.models line ~429 | 20:29 |
| *** Ryan_Lane has quit IRC | 20:29 | |
| *** drico has joined #openstack | 20:37 | |
| xtoddx | run_tests (specifically tests/api/openstack/fakes) can't find glance to import client, but it works fine via bpython. I installed in my code directory and made a .pth file in python path, just like i have for nova and swift. any ideas? | 20:40 |
| xtoddx | nm, i probably need to put it in venv lib dir | 20:41 |
| *** opengeard has joined #openstack | 20:44 | |
| *** pvo is now known as pvo_away | 20:49 | |
| *** opengeard has quit IRC | 20:50 | |
| *** adiantum has joined #openstack | 20:54 | |
| soren | Oh! | 20:54 |
| *** dubsquared has quit IRC | 20:55 | |
| soren | creiht, gholt: My suspicion was correct. I logged the current filepos upon entering put_object. It turns out that one of the files is processed twice. | 20:55 |
| *** exlt has quit IRC | 20:56 | |
| soren | creiht, gholt: ...reusing the file object, that is. | 20:56 |
| *** westmaas1 has quit IRC | 20:59 | |
| *** ctennis has joined #openstack | 21:00 | |
| *** ctennis has joined #openstack | 21:00 | |
| *** pharkmillups has joined #openstack | 21:02 | |
| *** exlt has joined #openstack | 21:03 | |
| *** ChanServ sets mode: +v exlt | 21:03 | |
| *** adiantum has quit IRC | 21:06 | |
| *** allsystemsarego has quit IRC | 21:07 | |
| jk0 | looks like unit tests fail since glance isn't installed | 21:08 |
| jk0 | but I'm sure you guys already knew that :) | 21:08 |
| * jk0 installs | 21:08 | |
| jk0 | wonder if we should add it as a dep? | 21:08 |
| *** adiantum has joined #openstack | 21:10 | |
| *** drico has quit IRC | 21:12 | |
| *** pvo_away is now known as pvo | 21:13 | |
| soren | creiht, gholt: Worked it out, I think. | 21:13 |
| soren | creiht: If the upload fails for some reason, an exception is raised. _retry retries the upload, reusing that arguments (including the open file descriptor). | 21:13 |
| dabo | jk0: it's already there: tools/pip-requires | 21:14 |
| jk0 | ah, darned venv! | 21:14 |
| *** MarkAtwood has quit IRC | 21:15 | |
| *** dirakx has joined #openstack | 21:15 | |
| gholt | soren: Bah! Ok, I think there's a bug filed for that already too, just didn't recognize the issue. :/ | 21:16 |
| *** JuanPerez has joined #openstack | 21:17 | |
| *** adiantum has quit IRC | 21:22 | |
| *** miclorb has joined #openstack | 21:22 | |
| gholt | soren: Yeah, found it finally: https://bugs.launchpad.net/swift/+bug/634979 Thanks for tracking that down. | 21:22 |
| uvirtbot | Launchpad bug 634979 in swift "client.Connection doesn't reset stream in put_object retry" [Low,New] | 21:22 |
| soren | gholt: Cool. | 21:23 |
| ttx | soren: do you think you'll have time to rereview live-migration tonight ? | 21:24 |
| ttx | soren: vishy was optimistic. | 21:25 |
| soren | ttx: I'm doing it right now. | 21:25 |
| ttx | soren: ok, I'll stay up a bit more then | 21:26 |
| *** adiantum has joined #openstack | 21:27 | |
| *** gdusbabek has quit IRC | 21:28 | |
| *** gdusbabek has joined #openstack | 21:29 | |
| vishy | tr3buchet: see project_get_network on line 1134 | 21:33 |
| xtoddx | dabo, jk0: i don't see glance in pip-requires? | 21:33 |
| vishy | tr3buchet: the issue is that two components trying to get a network for a project at the same time | 21:33 |
| vishy | tr3buchet: the unique is how we make sure that this doesn't happen. | 21:34 |
| dabo | xtoddx: you're right - I thought that earlier today someone said that it was added to pip-requires | 21:35 |
| xtoddx | dabo: I just uploaded a branch with building the .pth file like we do for nova | 21:35 |
| xtoddx | dabo: lp:~xtoddx/nova/glance-venv | 21:36 |
| dabo | xtoddx: it's just in the xenapi-glance-2 branch | 21:36 |
| tr3buchet | vishy: but supposing we wanted to have two networks? | 21:36 |
| jk0 | xtoddx: I didn't either, but I don't use venv | 21:36 |
| uvirtbot | New bug: #704161 in swift "client.py doesn't properly attempt to green https connections" [Medium,In progress] https://launchpad.net/bugs/704161 | 21:36 |
| uvirtbot | New bug: #704164 in nova "OpenStack API Create function fails with index out of range" [Undecided,New] https://launchpad.net/bugs/704164 | 21:36 |
| vishy | tr3buchet, well you'd have to remove the constraint and figure out a different way of ensuring that two different components are not associating a network at the same time | 21:37 |
| *** drico has joined #openstack | 21:38 | |
| *** adiantum has quit IRC | 21:41 | |
| dabo | xtoddx: instead of the 3-line construct like: f = open(pthfile, 'w'); f.write("%s\n" % glancedir); f.close() | 21:42 |
| dabo | it's cleaner to use: open(pthfile, 'w').write("%s\n" % glancedir) | 21:42 |
| xtoddx | dabo: no close() call? | 21:43 |
| dabo | not needed, since there is no reference hanging around | 21:43 |
| creiht | or | 21:43 |
| creiht | with open(pthfile, 'w') as f: | 21:43 |
| creiht | do stuff | 21:43 |
| creiht | always close your file descriptors | 21:43 |
| dabo | creiht: yeah, that's better for multi-line stuff | 21:43 |
| creiht | dabo: it is better period :) | 21:43 |
| dabo | creiht: there's no need for close() when you don't assign it to a name | 21:44 |
| *** fabiand_ has quit IRC | 21:45 | |
| creiht | That still feels too magical to me | 21:45 |
| dabo | it's the same as exiting a 'with open' construct | 21:45 |
| dabo | when the file-like object is destroyed, close() is always called. | 21:45 |
| *** adiantum has joined #openstack | 21:48 | |
| *** zul has quit IRC | 21:52 | |
| redbo | I don't see how that's magical, it works for very good reasons. But it will suck if we ever get a python that doesn't use refcounting for garbage collection. | 21:54 |
| creiht | Well yes, it would get closed when garbage collection runs, but that is non-determinate right? | 21:55 |
| *** jaypipes has quit IRC | 21:55 | |
| dabo | creiht: Only if you do assign it; if so, you must call close(). If you don't assign it, it is immediately closed. | 21:56 |
| creiht | ahh | 21:56 |
| redbo | With cpython it's going to close as soon as there aren't any references to the object (immediately). But as far as the python language spec, it's not deterministic, yeah. | 21:56 |
| * creiht still likes the with better :) | 21:57 | |
| redbo | I like the with statement. | 21:58 |
| dabo | nothing wrong with the with statement | 21:58 |
| dabo | the problem was that xtoddx was opening, assigning but not closing | 21:58 |
| *** gondoi_ has joined #openstack | 21:58 | |
| xtoddx | dabo: i fixed with block. i like code doing it, and not the garbage collector. | 21:59 |
| *** pharkmillups has quit IRC | 21:59 | |
| creiht | dabo: ahh, that wasn't evident as the example you posted from todd had a close | 21:59 |
| dabo | creiht: yeah, I added that to show that two were equivalent. | 22:00 |
| *** adiantum has quit IRC | 22:00 | |
| *** drico has quit IRC | 22:02 | |
| *** gondoi_ has quit IRC | 22:02 | |
| *** jaypipes has joined #openstack | 22:02 | |
| creiht | more than anything, using the with statement with opening files is just a good habit to get into | 22:02 |
| *** gondoi has quit IRC | 22:03 | |
| *** zul has joined #openstack | 22:04 | |
| *** adiantum has joined #openstack | 22:07 | |
| *** dendro-afk is now known as dendrobates | 22:09 | |
| * creiht wonders if the "waiting for launchpad to update" is the new "waiting for code to compile" | 22:10 | |
| uvirtbot | New bug: #704176 in nova "run_tests.sh -V fails -- Need to add glance to tools/pip-requires" [Low,In progress] https://launchpad.net/bugs/704176 | 22:11 |
| creiht | in reference to: http://xkcd.com/303/ | 22:11 |
| *** adiantum has quit IRC | 22:15 | |
| *** DubLo7 has joined #openstack | 22:16 | |
| *** adiantum has joined #openstack | 22:21 | |
| *** drico has joined #openstack | 22:28 | |
| *** mdomsch has quit IRC | 22:29 | |
| *** adiantum has quit IRC | 22:30 | |
| xtoddx | creiht: also, "waiting on tests to run" works in that circumstance | 22:31 |
| openstackhudson | Project swift build #158: SUCCESS in 26 sec: http://hudson.openstack.org/job/swift/158/ | 22:31 |
| openstackhudson | Tarmac: Updates client.py to green HTTPS connections if eventlet is available | 22:31 |
| *** ctennis has quit IRC | 22:36 | |
| *** burris has quit IRC | 22:36 | |
| *** burris has joined #openstack | 22:38 | |
| *** adiantum has joined #openstack | 22:42 | |
| *** spectorclan has quit IRC | 22:45 | |
| creiht | xtoddx: indeed :) | 22:47 |
| *** lvaughn has quit IRC | 22:50 | |
| *** drico_ has joined #openstack | 22:58 | |
| *** DubLo71 has joined #openstack | 22:58 | |
| *** drico has quit IRC | 23:01 | |
| *** DubLo7 has quit IRC | 23:01 | |
| *** adiantum has quit IRC | 23:01 | |
| *** adiantum has joined #openstack | 23:07 | |
| uvirtbot | New bug: #704196 in swift "swift3 middlware errors if content has unicode text" [High,Triaged] https://launchpad.net/bugs/704196 | 23:12 |
| openstackhudson | Project swift build #159: SUCCESS in 26 sec: http://hudson.openstack.org/job/swift/159/ | 23:16 |
| openstackhudson | Tarmac: s3api: fix AWSAccessKeyId | 23:16 |
| openstackhudson | We use cfaccount as AWSAccessKeyId (something like | 23:16 |
| openstackhudson | AUTH_89308df71f274e33af17779606f08fa0). However, users with the same | 23:16 |
| openstackhudson | account use the same cfaccount. In such case, we can't know which | 23:16 |
| openstackhudson | password should be used as a secret key to calculate the HMAC. | 23:16 |
| openstackhudson | This changes AWSAccessKeyId to the combination of account and user: | 23:16 |
| openstackhudson | Authorization: AWS test:tester:xQE0diMbLRepdf3YB+FIEXAMPLE= | 23:16 |
| openstackhudson | The auth validates the HMAC and sends a cfaccount back to the | 23:16 |
| openstackhudson | proxy. The proxy rewrites the path with the cfaccount. | 23:16 |
| *** adiantum has quit IRC | 23:26 | |
| *** adiantum has joined #openstack | 23:27 | |
| *** ctennis has joined #openstack | 23:31 | |
| *** adiantum has quit IRC | 23:32 | |
| *** EdwinGrubbs has quit IRC | 23:35 | |
| *** EdwinGrubbs has joined #openstack | 23:35 | |
| *** JuanPerez has quit IRC | 23:39 | |
| *** adiantum has joined #openstack | 23:44 | |
| *** pvo is now known as pvo_away | 23:51 | |
| *** pharkmillups has joined #openstack | 23:53 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!