| *** vladimir3p has quit IRC | 00:28 | |
| *** Binbin has joined #openstack-dev | 00:40 | |
| *** Tv has quit IRC | 00:41 | |
| *** rnirmal has joined #openstack-dev | 00:44 | |
| *** Tv has joined #openstack-dev | 01:02 | |
| *** pmullaney has quit IRC | 01:18 | |
| *** mattray has joined #openstack-dev | 01:30 | |
| *** cloudgroups has joined #openstack-dev | 01:31 | |
| *** cloudgroups has left #openstack-dev | 01:31 | |
| *** rnirmal has quit IRC | 01:36 | |
| *** mattray has quit IRC | 02:07 | |
| *** Vek has quit IRC | 02:11 | |
| *** Daviey has quit IRC | 02:11 | |
| *** letterj has quit IRC | 02:11 | |
| *** jeremyb has quit IRC | 02:11 | |
| *** johan___ has quit IRC | 02:11 | |
| *** comstud has quit IRC | 02:11 | |
| *** Vek has joined #openstack-dev | 02:11 | |
| *** letterj has joined #openstack-dev | 02:11 | |
| *** comstud has joined #openstack-dev | 02:11 | |
| *** jeremyb has joined #openstack-dev | 02:11 | |
| *** ChanServ sets mode: +v comstud | 02:11 | |
| *** rnirmal has joined #openstack-dev | 02:12 | |
| *** johan___ has joined #openstack-dev | 02:12 | |
| *** jeremyb has quit IRC | 02:13 | |
| *** jeremyb has joined #openstack-dev | 02:13 | |
| *** Daviey has joined #openstack-dev | 02:14 | |
| *** rnirmal has quit IRC | 02:18 | |
| *** glenc_ has joined #openstack-dev | 03:15 | |
| *** glenc has quit IRC | 03:17 | |
| *** Zangetsue has joined #openstack-dev | 03:45 | |
| *** openpercept_ has joined #openstack-dev | 03:59 | |
| *** openpercept_ has quit IRC | 04:31 | |
| *** openpercept_ has joined #openstack-dev | 04:33 | |
| *** openpercept_ is now known as openpercept | 04:34 | |
| *** openpercept has quit IRC | 04:34 | |
| *** openpercept has joined #openstack-dev | 04:34 | |
| *** RobertLaptop has quit IRC | 04:55 | |
| *** RobertLaptop has joined #openstack-dev | 04:56 | |
| *** RobertLaptop has quit IRC | 05:07 | |
| *** RobertLaptop has joined #openstack-dev | 05:08 | |
| *** RobertLaptop has quit IRC | 05:17 | |
| *** byeager has joined #openstack-dev | 05:20 | |
| *** byeager has quit IRC | 05:24 | |
| *** RobertLaptop has joined #openstack-dev | 05:24 | |
| *** Rubberneck has quit IRC | 05:55 | |
| *** Rubberneck has joined #openstack-dev | 05:56 | |
| *** mgius has quit IRC | 06:55 | |
| *** reidrac has joined #openstack-dev | 07:06 | |
| *** zaitcev has quit IRC | 07:10 | |
| *** Tv has quit IRC | 07:37 | |
| *** ChanServ sets mode: +v letterj | 08:56 | |
| *** gholt has left #openstack-dev | 09:13 | |
| *** Binbin has quit IRC | 10:17 | |
| *** markvoelker has joined #openstack-dev | 11:12 | |
| *** dprince has joined #openstack-dev | 13:02 | |
| *** sandywalsh has quit IRC | 13:06 | |
| *** ameade has joined #openstack-dev | 13:12 | |
| *** Zangetsue has quit IRC | 13:24 | |
| *** Zangetsue has joined #openstack-dev | 13:24 | |
| *** pmullaney has joined #openstack-dev | 13:24 | |
| *** sandywalsh has joined #openstack-dev | 13:36 | |
| *** Zangetsue has quit IRC | 13:46 | |
| *** glenc_ is now known as glenc | 13:58 | |
| *** jkoelker has joined #openstack-dev | 14:24 | |
| *** reidrac has quit IRC | 15:17 | |
| *** dragondm has joined #openstack-dev | 15:18 | |
| annegentle | when I try to build the nova docs, I see "DistUtilsExtra required to use i18n builders" - how can I install python-distutils-extra on Mac OSX? Or is it some link/config problem? | 15:26 |
|---|---|---|
| *** openpercept has quit IRC | 15:27 | |
| *** zaitcev has joined #openstack-dev | 15:43 | |
| annegentle | more info with the whole message at http://paste.openstack.org/show/1586/ | 15:45 |
| annegentle | **crickets** | 15:50 |
| creiht | annegentle: that is a package that is used by setup.py in nova to help with the translations stuff | 15:54 |
| creiht | https://launchpad.net/python-distutils-extra | 15:54 |
| creiht | unfortunatly it isn't really common, so there are not a lot of packages for it | 15:55 |
| creiht | so you will have to download the source tarball and install it manually | 15:55 |
| annegentle | ok, I guess I could also see if macports has it | 15:56 |
| annegentle | thanks creiht! | 15:56 |
| *** RobertLaptop has quit IRC | 15:57 | |
| *** rnirmal has joined #openstack-dev | 16:00 | |
| *** mattray has joined #openstack-dev | 16:08 | |
| *** vladimir3p has joined #openstack-dev | 16:18 | |
| *** openpercept_ has joined #openstack-dev | 16:27 | |
| *** openpercept_ is now known as Guest33895 | 16:27 | |
| *** blamar has quit IRC | 16:35 | |
| *** byeager has joined #openstack-dev | 16:45 | |
| *** RobertLaptop has joined #openstack-dev | 16:48 | |
| *** Guest33895 has left #openstack-dev | 16:49 | |
| *** jdurgin has joined #openstack-dev | 16:50 | |
| *** blamar has joined #openstack-dev | 16:51 | |
| *** gholt has joined #openstack-dev | 16:54 | |
| *** Tv has joined #openstack-dev | 16:54 | |
| jaypipes | slow day on #openstack-dev today... | 17:06 |
| dprince | jaypipes: you figure out the glance S3 issue? | 17:07 |
| *** alekibango has quit IRC | 17:07 | |
| jaypipes | dprince: heh, still logged in and working on it... :( | 17:11 |
| jaypipes | dprince: different versions of webob behave differently, unfortunately... 0.9.8 (installed in Ubuntu 10.10) is quite different from 1.0.2 through 1.0.7. And 1.0.8 is different from that even... | 17:12 |
| dprince | jaypipes: Sure. Sorry. Didn't mean to sound like a meany. | 17:12 |
| dprince | jaypipes: So can you reproduce it locally (on your workstation) if you install the same webob version the Maverick Cloud Server uses? | 17:13 |
| jaypipes | dprince: didn't take your comment meanly :) no worries | 17:13 |
| dprince | jaypipes: Great. Just reread it and thought I sounded like a jerk. :) | 17:14 |
| jaypipes | dprince: and yes, my local workstation is actually on 0.9.8 too (since I use Ubuntu 10.10). And in the virtualenv, I've installed updated versions of 1.0.2 and 1.0.8 and they behave differently... | 17:14 |
| jaypipes | dprince: basically, it seems the workaround the webob guys gave me for 1.0.2-1.0.7 breaks 0.9.8 functionality | 17:15 |
| jaypipes | dprince: still trying to figure out a workaround that works for everything, so I don't have to do messy stuff like switch on versions of webob in the code base... | 17:15 |
| openstackjenkins | Project nova build #988: SUCCESS in 2 min 53 sec: http://jenkins.openstack.org/job/nova/988/ | 17:44 |
| openstackjenkins | * Tarmac: Fix for bug#794239 to allow pep8 in run_tests.sh to use the virtual environment | 17:44 |
| openstackjenkins | * Tarmac: adding Authorizer key for ImportPublicKey | 17:44 |
| *** ohnoimdead has joined #openstack-dev | 18:02 | |
| sandywalsh | jaypipes s1rp cerberus had to resubmit since hudson chokes on dependent branches https://code.launchpad.net/~sandy-walsh/nova/dist-sched-3/+merge/63905 | 18:10 |
| openstackjenkins | Project nova build #989: SUCCESS in 2 min 51 sec: http://jenkins.openstack.org/job/nova/989/ | 18:14 |
| openstackjenkins | Tarmac: Cleaned up some pylint errors | 18:14 |
| openstackjenkins | Project nova build #990: SUCCESS in 2 min 51 sec: http://jenkins.openstack.org/job/nova/990/ | 18:24 |
| openstackjenkins | Tarmac: Support for header "X-Auth-Project-Id" in osapi. | 18:24 |
| jaypipes | dprince: ping | 18:30 |
| dprince | jaypipes: ack | 18:31 |
| jaypipes | dprince: ok, so after digging through the code in webob.Request, it looks like there really will be no other way than nasty, ugly hacks for each version of webob... so, do you think it's better for us to say "use webob of this version" or do the nasty hacks? | 18:32 |
| dprince | jaypipes: I think its acceptable to require a specific version. | 18:33 |
| jaypipes | dprince: and which version? | 18:33 |
| dprince | jaypipes: Can we at least detect a version mismatch and blow up if the requirement isn't met. | 18:33 |
| jaypipes | they're all kinda broken... it's just figuring out which one we want to put a workaround in for.. | 18:33 |
| dprince | jaypipes: I'm not sure I'm the one to answer that. Perhaps we should shoot for whatever Ubuntu's current version is? | 18:33 |
| jaypipes | dprince: hmm... for 11.04 or 10.10? | 18:34 |
| dprince | jaypipes: I'm using 10.10. I could go either way. 11.04 is LTS right? | 18:35 |
| jaypipes | dprince: not sure | 18:36 |
| dprince | jaypipes: Maybe not. | 18:36 |
| dprince | jaypipes: Do 10.04 and 10.10 have matching version of webob? | 18:37 |
| dprince | jaypipes: 10.04 is LTS. | 18:37 |
| jaypipes | dprince: I'm really not sure. how would I check that? | 18:38 |
| dprince | jaypipes: Check the mirrors? Fire up Cloud Servers? | 18:39 |
| dprince | jaypipes: Maybe its best to just make it work on what we are testing on? 10.10. | 18:39 |
| jaypipes | dprince: gah, this sucks... | 18:39 |
| jaypipes | dprince: I'm almost tempted to say "use the patched one we put in our PPA" :( | 18:40 |
| dprince | jaypipes: To be honest this is kind of a niche feature/thing anyway. | 18:40 |
| jaypipes | dprince: it shouldn't be,. | 18:40 |
| mattray | 11.04 is definitely not LTS. | 18:41 |
| dprince | jaypipes: Seriously? How many people plan on using Glance w/ S3 as a backend? | 18:42 |
| jaypipes | dprince: I believe this same issue would happen with Swift if I kept the code that makes the body seekable in there... | 18:44 |
| dprince | jaypipes: Ah. Well then. Now thats pretty important then isn't it. | 18:45 |
| jaypipes | dprince: yes. and I checked the Swift code in swift/obj/server.py. It would indeed break on PUT when called by Glance in the Swift backend, because the body_file attribute of the webob.Request is passed as-is to the swift client. It is that body_file attribute that is messed up when you call make_body_seekable() in webob 0.9.8 (the version in Ubuntu 10.10). Unfortunately, you MUST call make_body_seekable() before pass | 18:47 |
| jaypipes | ing body_file to the boto S3 PUT call otherwise you will get an AttributeError for the body_file not having a seek attribute... | 18:47 |
| notmyname | jaypipes: we had to jump from 0.9.8 to 1.0.7 | 18:47 |
| jaypipes | notmyname: heh, yeah... | 18:47 |
| jaypipes | notmyname: looking at the code, I wondered how you could get around that issue.. | 18:48 |
| jaypipes | notmyname: since obviously you can't force users to not use chunked transfer encoding. | 18:48 |
| notmyname | gholt complained loudly to them that they broke it and I thought they reverted it back to .9.8 behavior with the new functionality on new methods or attributes or something | 18:48 |
| dprince | jaypipes: There is your answer. | 18:49 |
| notmyname | but I don't remember the details on it :-) | 18:49 |
| jaypipes | notmyname: https://bitbucket.org/ianb/webob/issue/12/fix-for-issue-6-broke-chunked-transfer | 18:49 |
| dprince | jaypipes: Also, I'd prefer to close the Glance S3 ticket. The issue we are chasing here is a different problem that what you initially hit. Perhaps we should really file a new ticket for this since that thread is already a bit long. | 18:49 |
| jaypipes | dprince: well, the code in the merge proposal is still for adding POST functionality to the S3 driver. ;) | 18:50 |
| jaypipes | dprince: and it's that MP that is hung up on this issue... | 18:50 |
| jaypipes | notmyname: the "fix" that sergey talks about in that issue only fixes 1.0.7+. It doesn't work for 0.9.8 or 1.0.2 through 1.0.6 :( | 18:51 |
| dprince | jaypipes: Right. That code is basically working now besides this issue. If we file a new ticket we can close it out. Would help clear things up a bit. | 18:52 |
| jaypipes | dprince: yeah, but I can't merge that code until this is resolved. But, I suppose you are right about creating a new ticket. I'll do that. | 18:52 |
| notmyname | jaypipes: http://groups.google.com/group/paste-users/browse_thread/thread/10157caa893a82e7?pli=1 | 18:53 |
| notmyname | so swift will require either webob 0.9.8 or a much newer one | 18:54 |
| jaypipes | notmyname: cheers. thx for that link. | 18:55 |
| jaypipes | notmyname: unfortunately, in Glance, we have the unfortunate need to pass a *seekable* file to the S3 backend, and the webob.Request.make_body_seekable() method is broken for chunked transfer requests in webob.0.9.8. | 18:56 |
| notmyname | ah | 18:56 |
| jaypipes | notmyname: so, looks like we either do a hack for the body_seekable thing by simply wrapping the body_file in a cStringIO type of thing (likely this is best solution), or we require webob 1.0.7+ | 18:57 |
| jaypipes | notmyname: damn that boto. :) | 18:57 |
| notmyname | if it were my choice, I'd recommend the new version of webob | 18:57 |
| notmyname | we've got packages for it in the swift PPA, I think | 18:58 |
| notmyname | since we require it too | 18:58 |
| jaypipes | notmyname: versus the hack specifically in the S3 driver? | 18:58 |
| notmyname | ya. require non-buggy version >> hacks in your code (IMO) | 18:58 |
| jaypipes | :) | 18:59 |
| openstackjenkins | Project nova build #991: SUCCESS in 2 min 50 sec: http://jenkins.openstack.org/job/nova/991/ | 19:09 |
| openstackjenkins | Tarmac: MySQL database tables are currently using the MyISAM engine. Created migration script nova/db/sqlalchemy/migrate_repo/versions/021_set_engine_mysql_innodb.py to change all current tables to InnoDB. | 19:09 |
| openstackjenkins | Project nova build #992: SUCCESS in 2 min 52 sec: http://jenkins.openstack.org/job/nova/992/ | 19:16 |
| openstackjenkins | Tarmac: Now forwards create instance requests to child zones. | 19:16 |
| openstackjenkins | Refactored nova.compute.api.create() to support deferred db entry creation. | 19:16 |
| openstackjenkins | NOTE: created instances using the ZoneAwareScheduler result in two db entries: one in the parent zone, one in the child zone. This will be fixed in the next branch which will deal with reservation ID's and a new POST /zone/server OS API command. All the other schedulers work as advertised. | 19:16 |
| openstackjenkins | Project nova-tarball-bzr-delta build #239: FAILURE in 11 sec: http://jenkins.openstack.org/job/nova-tarball-bzr-delta/239/ | 19:16 |
| openstackjenkins | * Tarmac: Now forwards create instance requests to child zones. | 19:16 |
| openstackjenkins | Refactored nova.compute.api.create() to support deferred db entry creation. | 19:16 |
| openstackjenkins | NOTE: created instances using the ZoneAwareScheduler result in two db entries: one in the parent zone, one in the child zone. This will be fixed in the next branch which will deal with reservation ID's and a new POST /zone/server OS API command. All the other schedulers work as advertised. | 19:16 |
| openstackjenkins | * Tarmac: MySQL database tables are currently using the MyISAM engine. Created migration script nova/db/sqlalchemy/migrate_repo/versions/021_set_engine_mysql_innodb.py to change all current tables to InnoDB. | 19:16 |
| *** bcwaldon has joined #openstack-dev | 19:17 | |
| *** dprince has quit IRC | 19:19 | |
| *** bcwaldon has quit IRC | 19:29 | |
| *** bcwaldon has joined #openstack-dev | 19:32 | |
| jaypipes | dprince, bcwaldon: OK, tried to explain the webob and S3 problem to the best of my ability here: https://bugs.launchpad.net/glance/+bug/794718. (notmyname, just FYI...) | 19:37 |
| uvirtbot | Launchpad bug 794718 in glance "S3 requires seekable file. webob versions 0.9.8 through 1.0.6 make_body_seekable() method broken for chunked transfer requests" [Low,Triaged] | 19:37 |
| *** jtran has joined #openstack-dev | 19:51 | |
| *** jtran has quit IRC | 19:54 | |
| *** jtran has joined #openstack-dev | 19:54 | |
| *** bcwaldon has quit IRC | 20:27 | |
| *** ameade has quit IRC | 21:11 | |
| *** bcwaldon has joined #openstack-dev | 21:21 | |
| annegentle | I managed to install python-distutils-extra, feeling pretty proud. But. | 21:32 |
| annegentle | still, when I build_sphinx, I am getting this: http://paste.openstack.org/show/1590/ - something about import lockfile "No module named lockfile". | 21:33 |
| annegentle | my day would be complete if I could just build RST :) | 21:33 |
| creiht | annegentle: http://pypi.python.org/pypi/lockfile | 21:36 |
| creiht | to build the docs, you have to have all the nova depends installed | 21:36 |
| *** bcwaldon has quit IRC | 21:37 | |
| annegentle | yeah I thought venv would give me all the dependencies? | 21:39 |
| annegentle | but apparently that's not the case for now | 21:39 |
| creiht | ahh... that part I do not know | 21:39 |
| annegentle | creiht: it used to, I swear :) | 21:40 |
| annegentle | creiht: appreciate it :) | 21:40 |
| annegentle | creiht: one of these days, thanks to you, I'll submit some python code. One day. | 21:41 |
| creiht | lol | 21:41 |
| creiht | I haven't used venv or setup nova yet, so I'm not sure how all that works together | 21:42 |
| *** pmullaney has quit IRC | 21:52 | |
| *** markvoelker has quit IRC | 22:09 | |
| vishy | annegentle: lockfile should be in the venv, but you will have to build the docs with the venv active... | 22:12 |
| *** jtran has quit IRC | 22:24 | |
| *** jtran_ has joined #openstack-dev | 22:24 | |
| *** mattray has quit IRC | 22:40 | |
| *** jkoelker has quit IRC | 22:55 | |
| *** rnirmal has quit IRC | 23:04 | |
| *** dragondm has quit IRC | 23:24 | |
| *** pyhole has quit IRC | 23:28 | |
| *** pyhole has joined #openstack-dev | 23:29 | |
| jk0 | .w | 23:29 |
| pyhole | Hazel Green, WI: 89F/32C Humidity: 48% Wind: S at 18 mph Sunny | 23:29 |
| jk0 | .help | 23:30 |
| pyhole | Learn how to use active commands (ex: .help <command>) | 23:30 |
| pyhole | Active Commands: cnn, deop, digg, g, google, grouphug, help, imdb, join, lastnight, nick, op, part, reddit, reload, roll, say, title, twitter, urban, version, w, weather, wikipedia, youtube | 23:30 |
| pyhole | Active Keywords: | 23:30 |
| *** pyhole has quit IRC | 23:35 | |
| *** pyhole has joined #openstack-dev | 23:35 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!