| vladimir3p | Folks, have anoybody else seen these issues with nova-dhcpbridge? In previous versions there was a problem with hostname. Now it is removed, but it still crashes on unknown var "interface" | 00:04 |
|---|---|---|
| Daviey | mwhooker: bug 827386 seems pretty bitesize, and is kinda importiant. Not currently possible to see who uploaded an image, kinda bad on a shared cloud - as there is no way of gaining confidence of the person that uploaded it, and made it public. | 00:07 |
| uvirtbot | Launchpad bug 827386 in nova "DescribeImages does not fill imageOwnerId" [Undecided,New] https://launchpad.net/bugs/827386 | 00:07 |
| mwhooker | Daviey: Thanks. I'll check it out | 00:07 |
| Daviey | groovy | 00:07 |
| vishy | huh? imageOwnerId is not there? | 00:10 |
| vishy | how did you upload the image daviey? | 00:10 |
| vishy | Daviey: ^^ | 00:10 |
| *** ohnoimdead has quit IRC | 00:14 | |
| Daviey | vishy: euca-upload-bundle i think | 00:18 |
| Daviey | vishy: ah no, i used the wrapper, uec-publish-tarball | 00:20 |
| vishy | ok | 00:21 |
| vishy | we were returrning owner id so maybe it broke recently | 00:21 |
| Daviey | vishy: i'm lazy and use this, http://pb.daviey.com/hkDi/ | 00:21 |
| Daviey | mwhooker: If you are not able to reproduce it, i'll provide an xml dump of what i am getting back. | 00:24 |
| Daviey | but right now, BED.. nn folks. | 00:24 |
| mwhooker | thanks | 00:24 |
| *** anotherjesse has quit IRC | 00:35 | |
| *** anotherjesse has joined #openstack-dev | 00:42 | |
| *** rods has quit IRC | 00:47 | |
| *** jakedahn has quit IRC | 00:50 | |
| *** anotherjesse has quit IRC | 00:53 | |
| *** mattray has joined #openstack-dev | 00:54 | |
| *** mattray has quit IRC | 00:54 | |
| *** mfer has quit IRC | 01:08 | |
| *** tushar_ has quit IRC | 01:13 | |
| *** lorin1 has joined #openstack-dev | 01:21 | |
| *** lorin1 has left #openstack-dev | 01:21 | |
| *** hisaharu has quit IRC | 01:23 | |
| *** vladimir3p has quit IRC | 01:25 | |
| *** jakedahn has joined #openstack-dev | 01:33 | |
| *** dragondm has quit IRC | 01:39 | |
| *** tsuzuki_ has joined #openstack-dev | 01:58 | |
| *** mfer has joined #openstack-dev | 02:16 | |
| *** mfer has quit IRC | 02:16 | |
| *** mnaser has quit IRC | 02:36 | |
| *** deshantm is now known as deshantm_away | 03:09 | |
| *** anotherjesse has joined #openstack-dev | 03:18 | |
| *** bengrue has quit IRC | 03:22 | |
| *** martine has joined #openstack-dev | 03:39 | |
| *** RobertLaptop has quit IRC | 05:14 | |
| *** RobertLaptop has joined #openstack-dev | 05:17 | |
| *** martine has quit IRC | 05:22 | |
| *** RobertLaptop has quit IRC | 05:27 | |
| *** RobertLaptop has joined #openstack-dev | 05:30 | |
| *** anotherjesse has quit IRC | 05:46 | |
| *** reidrac has joined #openstack-dev | 07:00 | |
| *** phil121 has quit IRC | 08:58 | |
| *** mnour has joined #openstack-dev | 09:07 | |
| *** tsuzuki_ has quit IRC | 09:19 | |
| *** rods has joined #openstack-dev | 09:46 | |
| *** YorikSar has joined #openstack-dev | 09:50 | |
| openstackjenkins | Project nova build #1,261: SUCCESS in 3 min 36 sec: https://jenkins.openstack.org/job/nova/1261/ | 10:05 |
| openstackjenkins | Tarmac: Adds the enabled status of a host when XenServer reports its host's capabilities. This allows the scheduler to ignore hosts whose enabled is False when considering where to place a new instance. | 10:05 |
| *** mancdaz has quit IRC | 11:05 | |
| *** markvoelker has quit IRC | 11:40 | |
| *** mnour has quit IRC | 11:45 | |
| *** markvoelker has joined #openstack-dev | 11:48 | |
| *** lorin1 has joined #openstack-dev | 11:51 | |
| *** lorin1 has left #openstack-dev | 11:53 | |
| *** mancdaz has joined #openstack-dev | 11:53 | |
| *** mancdaz has quit IRC | 11:54 | |
| *** mancdaz has joined #openstack-dev | 11:55 | |
| *** mfer has joined #openstack-dev | 12:03 | |
| *** lts has joined #openstack-dev | 12:09 | |
| *** mnour has joined #openstack-dev | 12:10 | |
| *** chipc has joined #openstack-dev | 12:19 | |
| *** chipc has quit IRC | 12:24 | |
| *** bsza has joined #openstack-dev | 12:27 | |
| *** martine has joined #openstack-dev | 12:29 | |
| *** chipc has joined #openstack-dev | 12:34 | |
| *** zul has joined #openstack-dev | 12:44 | |
| *** deshantm_away is now known as deshantm | 12:46 | |
| *** nmistry has joined #openstack-dev | 12:51 | |
| *** jaypipes has quit IRC | 13:00 | |
| *** dprince has joined #openstack-dev | 13:03 | |
| *** zns has joined #openstack-dev | 13:08 | |
| zns | comstud: Could you help us with an eventlet issue? We've had this reported (http://pastie.org/2387012) and also been told you're the eventlet guru. Any quick pointers? | 13:16 |
| *** joesavak has joined #openstack-dev | 13:21 | |
| openstackjenkins | Project nova build #1,262: SUCCESS in 3 min 33 sec: https://jenkins.openstack.org/job/nova/1262/ | 13:25 |
| openstackjenkins | Tarmac: Fixed several logical errors in the scheduling process. Renamed the 'ZoneAwareScheduler' to 'AbstractScheduler', since the zone-specific designation is no longer relevant. Created a BaseScheduler class that has basic filter_hosts() and weigh_hosts() capabilities. Moved the filters out of one large file and into a 'filters' subdirectory of nova/scheduler. | 13:25 |
| openstackjenkins | Since I had to do a lot of editing of the files to change the naming and references, I cleaned up much of the inconsistent formatting. Normally formatting changes don't belong in a code change merge prop, but as I was making large-scale changes to the files, it was simplest to do it all at once. | 13:25 |
| *** amccabe has joined #openstack-dev | 13:30 | |
| *** kbringard has joined #openstack-dev | 13:37 | |
| *** nmistry has quit IRC | 13:37 | |
| *** joesavak has quit IRC | 13:52 | |
| *** nmistry has joined #openstack-dev | 14:03 | |
| *** joesavak has joined #openstack-dev | 14:05 | |
| *** jsavak has joined #openstack-dev | 14:07 | |
| *** joesavak has quit IRC | 14:10 | |
| *** zul has quit IRC | 14:12 | |
| *** zul has joined #openstack-dev | 14:13 | |
| *** timr has joined #openstack-dev | 14:18 | |
| *** mattray has joined #openstack-dev | 14:18 | |
| *** anp has joined #openstack-dev | 14:24 | |
| *** mnaser has joined #openstack-dev | 14:33 | |
| *** bcwaldon has joined #openstack-dev | 14:43 | |
| *** rnirmal has joined #openstack-dev | 14:49 | |
| *** zul has quit IRC | 14:49 | |
| soren | zns: Are you throwing a lot of stuff at it? I.e. is it likely to run into EMFILE? | 14:53 |
| soren | zns: ...or does it rather smell like it's not closing file handles properly? | 14:54 |
| kbringard | hey soren: have you seen this before? | 14:55 |
| kbringard | http://paste.openstack.org/show/2201/ | 14:55 |
| kbringard | it happens when I launch a VM in a security group that has a different security group set as a --source-group | 14:55 |
| kbringard | so like, the web group has -o default | 14:55 |
| kbringard | then I launch a VM in web, and it craps itself | 14:55 |
| kbringard | I didn't see a bug, but I wanted to ask if you'd seen that before before I start digging into it | 14:56 |
| *** dragondm has joined #openstack-dev | 15:00 | |
| *** dragondm has quit IRC | 15:01 | |
| *** heckj has joined #openstack-dev | 15:01 | |
| *** mnour has quit IRC | 15:02 | |
| *** _cerberus_ has joined #openstack-dev | 15:05 | |
| *** Gordonz has quit IRC | 15:05 | |
| *** ChanServ sets mode: +v _cerberus_ | 15:06 | |
| *** reidrac has quit IRC | 15:06 | |
| *** alekibango has quit IRC | 15:11 | |
| soren | kbringard: This might be what vishy talked about a couple of days ago. | 15:11 |
| *** dragondm has joined #openstack-dev | 15:11 | |
| soren | kbringard: Which version of euca2ools are you using? | 15:12 |
| *** Gordonz has joined #openstack-dev | 15:12 | |
| kbringard | I've tried it with 1.2 and 1.3 | 15:14 |
| kbringard | the 1.2 has the source-port = None default thing, right? | 15:14 |
| kbringard | it doesn't seem to be doing it anymore on the newest trunk, but it seems that the source groups have gone back to not working (at least using euca2ools 1.2) | 15:16 |
| *** yogirackspace has joined #openstack-dev | 15:18 | |
| soren | ERr... I didn't even realise the source groups stuff had been merged. | 15:18 |
| soren | Heh :) | 15:19 |
| soren | I guess it has. | 15:19 |
| kbringard | I didn't think it had either… I just happened to notice that when I tried to launch VMs in a specific group, they all failed :-) | 15:19 |
| kbringard | ah, yea, same thing when I add the rule with euca2ools 1.3 | 15:20 |
| soren | Ok. | 15:21 |
| soren | Can you file a bug about it? I'm about to call it a day. | 15:21 |
| kbringard | sure thing, I'm in meetings all day but I'll file the bug now and try to find some time to look at it myself as well | 15:23 |
| kbringard | thanks, soren. Have a good night, sir! | 15:23 |
| *** kbringard1 has joined #openstack-dev | 15:39 | |
| *** kbringard has quit IRC | 15:39 | |
| *** kbringard1 has quit IRC | 15:40 | |
| *** vladimir3p has joined #openstack-dev | 15:45 | |
| *** jsavak has quit IRC | 15:46 | |
| *** heckj has quit IRC | 15:49 | |
| *** heckj has joined #openstack-dev | 15:49 | |
| *** heckj has quit IRC | 15:50 | |
| *** jsavak has joined #openstack-dev | 15:53 | |
| *** mattfarina has joined #openstack-dev | 15:56 | |
| *** mfer has quit IRC | 15:56 | |
| *** mattfarina is now known as mfer | 15:56 | |
| *** mnaser has quit IRC | 16:05 | |
| *** mfer has quit IRC | 16:06 | |
| *** ohnoimdead has joined #openstack-dev | 16:10 | |
| zns | soren: frankly, not sure how to replicate or why it is happening. The pastie is from anotherjesse. But it does look like an EMFILE ("[Errno 24] Too many open files"... | 16:24 |
| zns | anotherjesse: how can we replicate this? http://pastie.org/2387012 | 16:25 |
| *** jakedahn has quit IRC | 16:25 | |
| *** mnaser has joined #openstack-dev | 16:27 | |
| zns | soren: is there cleanup we should be doing somewhere? Or a flag we can turn on to debug deeper? | 16:28 |
| zns | What we do is pretty basic: | 16:28 |
| zns | server = wsgi.Server() | 16:28 |
| zns | server.start(app, int(conf['service_port']), conf['service_host']) | 16:28 |
| anp | when I install Openstack-dashboard from repo it installs openstack-compute-2.0a1-1.noarch.rpm | 16:32 |
| anp | but I already have openstack-nova-compute-2011.3-0.20110727.1143.10.el6.noarch installed | 16:33 |
| anp | will this conflict? | 16:33 |
| anp | Hi | 16:43 |
| anp | I installed openstack-dashboard | 16:43 |
| anp | I am able to relog | 16:43 |
| anp | but I get this error after log in | 16:43 |
| anp | Error | 16:43 |
| anp | Unable to get service info: This server could not verify that you are authorized to access the document you requested. Either you supplied the wrong credentials (e.g., bad password), or your browser does not understand how to supply the credentials required. | 16:43 |
| *** heckj has joined #openstack-dev | 16:46 | |
| anp | Also I dont see any options to launch a new instance from there | 16:49 |
| anp | but I can see that services are up | 16:50 |
| anp | never mind I found the option by logging in a normal user | 16:52 |
| anp | for all the users it shows: | 16:53 |
| anp | ErrorUnable to get usage info: This server could not verify that you are authorized to access the document you requested. Either you supplied the wrong credentials (e.g., bad password), or your browser does not understand how to supply the credentials required. | 16:53 |
| anp | even if I am logged in succesfully | 16:53 |
| heckj | ttx: ping? | 16:56 |
| *** nmistry has quit IRC | 16:56 | |
| *** bcwaldon_ has joined #openstack-dev | 17:01 | |
| *** bcwaldon has quit IRC | 17:03 | |
| *** alekibango has joined #openstack-dev | 17:03 | |
| *** zns has quit IRC | 17:04 | |
| *** mfer has joined #openstack-dev | 17:05 | |
| *** jakedahn has joined #openstack-dev | 17:07 | |
| *** jsavak has quit IRC | 17:10 | |
| *** zns has joined #openstack-dev | 17:10 | |
| *** jakedahn has quit IRC | 17:13 | |
| *** dprince has quit IRC | 17:14 | |
| *** bcwaldon_ has quit IRC | 17:17 | |
| *** joesavak has joined #openstack-dev | 17:19 | |
| *** mattray1 has joined #openstack-dev | 17:20 | |
| *** mattray has quit IRC | 17:22 | |
| vishy | soren, kbringard: haven't seen that one | 17:23 |
| vishy | blamar_: ping | 17:24 |
| blamar_ | vishy: hey hey | 17:24 |
| *** zul has joined #openstack-dev | 17:24 | |
| vishy | blamar_: wanted to chat with you about some of the stuff in your review | 17:24 |
| openstackjenkins | Project nova build #1,263: SUCCESS in 3 min 32 sec: https://jenkins.openstack.org/job/nova/1263/ | 17:25 |
| openstackjenkins | Tarmac: Fix to return 413 for over limit exceptions with instances, metadata and personality | 17:25 |
| vishy | blamar_: gettext thing is clear, I guess it just wasn't removed from the other binaries, I'll pull it from all of them | 17:25 |
| vishy | blamar_: interested in how we would use entrypoints for the binary loading | 17:25 |
| vishy | blamar_: in the mean time, should I switch it to explicitly pass in all of the parameters to Service() ? | 17:26 |
| *** jsavak has joined #openstack-dev | 17:27 | |
| blamar_ | vishy: incremental changes here are fine/better IMO if you want | 17:27 |
| vishy | blamar_: also, what exactly is the problem you were trying to solve with multiprocessing? Efficient use of cores? Can't we just start multiple copies of the api if we want to use the cores? | 17:27 |
| blamar_ | vishy: ec2 and osapi are two services, why are they sharing a file when every other service gets to run by itself? | 17:28 |
| vishy | blamar_: I get your point there, no problem separating them. I think it makes sense to include a combined one as well for compatibility with existing scripts | 17:29 |
| *** zul has quit IRC | 17:29 | |
| vishy | blamar_: I was just trying to understand if there is some hidden value of multiprocessing that I'm missing | 17:30 |
| blamar_ | vishy: sure, so thats the one point, the other point is kind of processor-related, it's a lot easier to debug an eventlet API for performance if it's running in it's own process | 17:30 |
| *** joesavak has quit IRC | 17:31 | |
| vishy | blamar_: so that is solved by having the separate binary files? | 17:32 |
| blamar_ | vishy: yeah | 17:32 |
| vishy | blamar_: ok I will add the separate files then | 17:33 |
| vishy | I spent about a week early in the process debugging an issue with twisted + multiprocessing | 17:33 |
| blamar_ | vishy: ugh, twisted, sorry you had to do that | 17:34 |
| vishy | and it seems like odd issues like this one keep coming up so I prefer not to use event loops and multiprocessing together | 17:34 |
| blamar_ | vishy: I swear I've never had any issues in the past using gevent + multiprocessing to solve problems related to cpu-utilization so I'll admit to jumping to that solution a little too quickly | 17:35 |
| vishy | blamar_: well it looks like it is actually a twisted bug | 17:35 |
| vishy | s/twisted/eventlet | 17:35 |
| vishy | blamar_: wrt signal handling | 17:35 |
| blamar_ | vishy: and I'm looking into how we might be able to use entry_points in the future, not a good answer for the "how" on that one right now, but I know it's possible | 17:35 |
| *** troytoman-away is now known as troytoman | 17:35 | |
| vishy | blamar_: my entry point experience is kind of limited | 17:36 |
| vishy | blamar_: so there is no need to have a custome signal handler at this point? | 17:36 |
| blamar_ | vishy: you should be able to trash all signal handling | 17:37 |
| vishy | k | 17:40 |
| *** zul has joined #openstack-dev | 17:43 | |
| *** zul has quit IRC | 17:43 | |
| *** zul has joined #openstack-dev | 17:44 | |
| *** yogirackspace has quit IRC | 17:48 | |
| *** nickon has joined #openstack-dev | 17:54 | |
| *** jakedahn has joined #openstack-dev | 17:56 | |
| creiht | entry points++ | 17:57 |
| creiht | :) | 17:57 |
| *** zns has quit IRC | 18:01 | |
| *** yogirackspace has joined #openstack-dev | 18:02 | |
| *** jsavak has quit IRC | 18:02 | |
| *** zns has joined #openstack-dev | 18:02 | |
| *** jtanner has joined #openstack-dev | 18:07 | |
| *** joesavak has joined #openstack-dev | 18:07 | |
| *** zul has quit IRC | 18:15 | |
| *** rods has quit IRC | 18:17 | |
| *** bcwaldon has joined #openstack-dev | 18:20 | |
| vishy | blamar_: check the new version of the code I just pushed? | 18:22 |
| blamar_ | vishy: checking | 18:22 |
| vishy | blamar_: I tried to address your comments and clean up things further | 18:22 |
| vishy | vladimir3p: ping | 18:23 |
| vladimir3p | vishy: pong | 18:23 |
| vishy | so my partner here and I are doing volumes all day today | 18:23 |
| vishy | how should we split this up | 18:23 |
| vladimir3p | vishy: are you doing actual code work / planning ? | 18:24 |
| vishy | actual code | 18:25 |
| vladimir3p | vishy: great... let me know if there are any potions of the code that you would like me to take. I'm prety much available today | 18:25 |
| blamar_ | vishy: think the nova-network Server line might be wrong (line 327) | 18:27 |
| vladimir3p | vishy: also, we can probably do VolumeTypes + ExtraSpecs now and leave scheduler for the future ... | 18:27 |
| vishy | blamar_ nice one | 18:27 |
| blamar_ | also nova-scheduler | 18:28 |
| vishy | oops | 18:28 |
| vishy | didn't push the last version | 18:28 |
| vishy | rev 1457 is coming up give it a sec | 18:28 |
| vishy | :) | 18:28 |
| vishy | * 1456 | 18:29 |
| *** zns has quit IRC | 18:29 | |
| *** mattray1 is now known as mattray | 18:36 | |
| *** mattray has joined #openstack-dev | 18:36 | |
| ironcamel | vishy: do you know about the "attrs" field of a servers response in the api? | 18:40 |
| ironcamel | a stack i built via nova.sh is adding attrs to the servers response | 18:40 |
| ironcamel | http://paste.openstack.org/show/2204/ | 18:49 |
| vishy | ironcamel: not sure | 18:51 |
| ironcamel | vishy: do you see what i'm talking about? | 18:51 |
| vishy | ironcamel: but i don't usually dump the response to see what is in there | 18:51 |
| vishy | yeah | 18:51 |
| ironcamel | i'm wondering if that is some magic cloudbuilders added | 18:51 |
| ironcamel | because that is not in the spec | 18:51 |
| ironcamel | and another stack i deployed using opentstack_vpc does not have the attrs stuff | 18:51 |
| vishy | ironcamel: could be an addition via openstackx | 18:52 |
| ironcamel | i'm not using bindings | 18:52 |
| ironcamel | i'm basically using straight curl | 18:52 |
| Daviey | ironcamel: Hey, did you get any further with kombu? | 18:52 |
| ironcamel | i have not | 18:52 |
| Daviey | ironcamel: ok. | 18:52 |
| ironcamel | Daviey: how about you? | 18:52 |
| vishy | ironcamel: yes i think it is to support one of the extensions | 18:53 |
| Daviey | ironcamel: nope. | 18:53 |
| ironcamel | vishy: openstackx is injecting that attrs stuff? | 18:53 |
| Daviey | ironcamel: not touched it since i last spoke about it | 18:53 |
| vishy | ironcamel: yeah, there is a bunch of stuff missing in the response to make dashboard work, we are trying to get all of the stuff added as extensions | 18:54 |
| vishy | but first pass was just throw the needed data into a attrs as a workaround | 18:54 |
| Daviey | is python-novaclient not using jenkins? Or at least testsuite on merge? | 18:56 |
| ironcamel | vishy: when you add stuff via extensions, aren't they supposed to be in their own namespace? | 19:01 |
| vishy | extra attributes aren't | 19:02 |
| ironcamel | you guys just plopped that stuff right into the top level. so dirty | 19:02 |
| vishy | extra endpoints | 19:02 |
| vishy | :) | 19:02 |
| * ironcamel shakes his head in disapproval | 19:02 | |
| vishy | ironcamel: yes, quite dirty. All of this stuff needs to be added to the spec though | 19:02 |
| vishy | that is why we didn't propose it as-is :) | 19:03 |
| *** dprince has joined #openstack-dev | 19:04 | |
| *** bengrue has joined #openstack-dev | 19:06 | |
| vladimir3p | vishy: pls let me know if I can help with anything re volume types/extra-data | 19:07 |
| vishy | vladimir3p: we're working on cleanup first | 19:12 |
| vishy | vladimir3p: so if you want to start on the extra data types stuff | 19:12 |
| vishy | it will take us at least a couple hours to get to that part | 19:13 |
| *** dolphm has joined #openstack-dev | 19:13 | |
| vladimir3p | vishy: ok. will take extra-specs | 19:14 |
| vladimir3p | vishy: could you prepare DB schemas for types & extra-specs? | 19:14 |
| vishy | vladimir3p: just copy from instance-types | 19:16 |
| vishy | there are currently no extra fields | 19:16 |
| vishy | so it is just id and name at the moment | 19:16 |
| mtaylor | Daviey: no, python-novaclient is not being managed currently - it's on my list to bring up with someone at some point :) | 19:21 |
| vladimir3p | ok, will add the extraspecs a-la instance extraspecs | 19:21 |
| Daviey | mtaylor: might be a good idea, current trunk seems to fail the test suite. | 19:22 |
| *** alekibango has quit IRC | 19:23 | |
| mtaylor | Daviey: oh, it's definitely a good idea. :) | 19:25 |
| mtaylor | dolphm: I get this: http://paste.openstack.org/show/2205/ everytime I run the keystone tests locally ... any ideas? | 19:25 |
| dolphm | mtaylor: it looks like you're running without --with-coverage? | 19:26 |
| mtaylor | dolphm: yes | 19:26 |
| dolphm | mtaylor: then you probably have all kinds of failures... | 19:27 |
| mtaylor | really? | 19:27 |
| mtaylor | we don't run --with-coverage on the main jenkins job for merges | 19:27 |
| dolphm | mtaylor: running tests *without* --with-coverage causes it to fail-fast... so you only see one arbitrary failure | 19:27 |
| mtaylor | ./run_tests.sh -N is all we do there | 19:27 |
| mtaylor | REALLY - that's rather odd - how are merges happening then | 19:27 |
| dolphm | mtaylor: that's fine | 19:27 |
| dolphm | ./run_tests.sh -N # aborts on first failure | 19:28 |
| dolphm | ./run_tests.sh -N --with-coverage # runs through all failures | 19:28 |
| mtaylor | oh. gotcha | 19:29 |
| dolphm | sort of makes sense to me, but i didn't come up with it... | 19:29 |
| mtaylor | sorry, I thought you were saying something different :) | 19:29 |
| dolphm | i guess it's convenient considering stdout is suppressed (grr) | 19:29 |
| dolphm | anyway, you probably have bigger problems than that unit test | 19:29 |
| mtaylor | ok. I'll turn on coverage and see what happens | 19:29 |
| dolphm | paste your git diff if you want ;) | 19:33 |
| mtaylor | dolphm: oh, this is me working on packaging, so now that I can see more things failing - I think it has to do with versions of dependencies | 19:33 |
| * Daviey perks up | 19:38 | |
| dolphm | mtaylor: that would make sense | 19:39 |
| dolphm | mtaylor: anything in keystone.sql.log? | 19:39 |
| mtaylor | dolphm: TypeError: 'ItemNotFoundFault' object is not callable | 19:44 |
| dolphm | where's that being thrown? | 19:46 |
| dolphm | mtaylor: ^ | 19:46 |
| vladimir3p | vishy: ping | 19:47 |
| vishy | vladimir3p: sup? | 19:47 |
| vladimir3p | vishy: prior to starting on volume types could you pls let me know. it is quite problematic to test extra-specs alone, so I added some volume types staff as well | 19:47 |
| vladimir3p | vishy: probably you could take my tree when you will start your part | 19:48 |
| vladimir3p | (or we can merge them...) | 19:48 |
| mtaylor | dolphm: wait - new error (I can fresh so I could track where it was starting to go nuts) | 19:48 |
| mtaylor | dolphm: raise TemplateError('Template %s not found.' % repr(name)) | 19:48 |
| mtaylor | TypeError: __init__() takes at least 3 arguments (2 given) | 19:48 |
| vishy | vladimir3p: you can do types as well at least in the db | 19:48 |
| mtaylor | dolphm: keystone/common/template.py line 95 | 19:48 |
| vladimir3p | vishy: yep, ok | 19:48 |
| dolphm | mtaylor: that's fixable... looking into it | 19:49 |
| dolphm | mtaylor: try changing it to... | 19:55 |
| dolphm | raise TemplateError('Template %s not found.' % repr(name), sys.exc_info()) | 19:55 |
| dolphm | mtaylor: just a wild guess... i get errors just trying to import paste.util.template, so i can't even poke at TemplateError | 19:55 |
| mtaylor | dolphm: aaaah. I may have found a thing ... | 20:03 |
| soren | I'm staring at the XenAPI driver. I'm having trouble grokking the rescue code. | 20:04 |
| soren | I can't for the life of me figure out where the rescue disk image gets attached. | 20:05 |
| soren | I mean... Of course it gets attached as the first disk, but which part of the code makes that happen? | 20:06 |
| soren | The fact that spawn_rescue just calls spawn with identical arguments really weirds me out. | 20:06 |
| soren | I vaguely remember commenting on a merge proposal about that, but I forget what the rationale was :( | 20:07 |
| *** mnour has joined #openstack-dev | 20:07 | |
| mtaylor | dolphm: heh. I'm uncovering some issues because I'm running 2.6 on this maverick box - and we're doing testing on natty with 2.7 right now. | 20:11 |
| * mtaylor really needs to get both up and going... | 20:11 | |
| *** jsavak has joined #openstack-dev | 20:12 | |
| mtaylor | dolphm: keystone/test/functional/test_ext_rs_key_auth.py line 24: SyntaxError: expected path separator ([) | 20:13 |
| soren | jk0: You wrote some of that code. Can you lend me a hand, perhaps? | 20:14 |
| jk0 | sure | 20:14 |
| soren | \o/ | 20:14 |
| mtaylor | dolphm: http://paste.openstack.org/show/2208/ fwiw | 20:15 |
| mtaylor | dolphm: we're getting MUCH further now though | 20:15 |
| jk0 | so bascially a rescue instance has the same attributes as the original instance. it just attached the original instance's disk to it | 20:15 |
| soren | jk0: Right. | 20:15 |
| soren | jk0: What I'm wondering is where it gets the rescue image. | 20:15 |
| *** joesavak has quit IRC | 20:16 | |
| jk0 | soren: it uses the image the original uses | 20:17 |
| jk0 | so if the original was ubuntu, the rescue image would use the same ubuntu base image | 20:17 |
| soren | jk0: Oh, so not a special rescue image? | 20:17 |
| jk0 | doing it this was got us around having to prep and maintain a specific rescue image for each distro | 20:17 |
| jk0 | nope | 20:17 |
| openstackgerrit | Monty Taylor proposed a change to openstack/keystone: Cleaned up come issues with python2.6. https://review.openstack.org/270 | 20:17 |
| jk0 | s/was/way/ | 20:18 |
| soren | jk0: Ok, gotcha. That explains why I couldn't find this magic rescue image :) | 20:18 |
| jk0 | haha, yeah that would do it :) | 20:18 |
| soren | jk0: Great, thanks a lot! | 20:18 |
| jk0 | no problem | 20:18 |
| openstackgerrit | Yogeshwar Srikrishnan proposed a change to openstack/keystone: Changes to allow test to work on python 2.6.*. https://review.openstack.org/271 | 20:23 |
| mwhooker | smoser: hi. I'm working on bug # 827386 https://bugs.launchpad.net/nova/+bug/827386. what image service were you using? | 20:23 |
| uvirtbot | Launchpad bug 827386 in nova "DescribeImages does not fill imageOwnerId" [Undecided,New] | 20:23 |
| mtaylor | dolphm: woot! I found all the issues... updated patch coming... and haha, yogi seemed to have been working on the same thing :) | 20:23 |
| mwhooker | smoser: Looking at the code, glance + euca2ools w/ boto should display the owner id | 20:23 |
| Daviey | mwhooker: He was using nova-objectstore | 20:23 |
| mwhooker | thanks Daviey | 20:24 |
| Daviey | mwhooker: But you don't think it's an issue with glance? | 20:24 |
| *** jsavak has quit IRC | 20:24 | |
| mwhooker | I've only done a quick scan of the code, but I see the owner_id from the database being sent | 20:24 |
| Daviey | mwhooker: I don't have complete confidence about that string being empty. | 20:26 |
| mwhooker | good point | 20:26 |
| mwhooker | it would be helpful if I could see some http sessions | 20:26 |
| vladimir3p | vishy: First cut on volume types & extra-data - lp:~vladimir.p/nova/volume_type_extradata (primarely DB schemas, DB APIs, migration) | 20:27 |
| soren | Daviey: Even if you use nova-objectstore, the image lands in glance. | 20:27 |
| vishy | vladimir3p: coolness | 20:28 |
| vishy | checking | 20:28 |
| soren | Daviey: nova-objectstore is where you push it to (if you're using cloud-utils). Once you call RegisterImage, it goes and grabs the image, deciphers it, stitches it together and pushes it to glance. | 20:28 |
| openstackgerrit | A change was merged to openstack/openstack-ci-puppet: Cleanup. https://review.openstack.org/267 | 20:29 |
| mwhooker | hmm, actually api/ec2/cloud.py._format_image seems to be trying to get 'owner_id', but in glance the field is simply "owner" | 20:29 |
| vishy | vladimir3p: looking good | 20:30 |
| mwhooker | let me see if I can test that | 20:30 |
| vishy | Daviey, mwhooker: it was using a custom owner_id set in properties, because glance didn't have an owner field until recently | 20:31 |
| vladimir3p | vishy: I will add tests for extra-specs & types ... (at least on DB level) | 20:31 |
| vladimir3p | vishy: next steps might be APIs for volume types, Openstack APIs for everything, integration with volume code | 20:32 |
| vishy | agreed | 20:34 |
| mwhooker | vishy: grepping for "owner_id" in nova/glance turns up little. can you point me in the right direction? | 20:34 |
| *** YorikSar has quit IRC | 20:35 | |
| *** dprince has quit IRC | 20:35 | |
| vishy | mwhooker: what do you need pointed at? | 20:35 |
| *** jsavak has joined #openstack-dev | 20:35 | |
| *** YorikSar has joined #openstack-dev | 20:35 | |
| mwhooker | where in the code the "customer owner_id" is set in properties | 20:36 |
| *** zns has joined #openstack-dev | 20:36 | |
| vishy | mwhooker: it is set in properties in nova/image/s3.py i believe | 20:36 |
| vishy | and also in nova-manage register commands | 20:36 |
| vishy | s/register/image register/ | 20:37 |
| mwhooker | looks like owner is being set as "project_id", not "owner_id" | 20:40 |
| openstackgerrit | James E. Blair proposed a change to openstack/openstack-ci: Updates to blueprint/topics. https://review.openstack.org/272 | 20:42 |
| *** zns has quit IRC | 20:42 | |
| Daviey | mwhooker: that seems most likely based on the output, oneiric-20110813212842/oneiric-server-cloudimg-amd64.img | 20:43 |
| Daviey | mwhooker: I uploaded the image using, http://pb.daviey.com/hkDi/ . so the project_id == bucket? | 20:44 |
| Daviey | the project_id certainly was not the same value as bucket btw. | 20:44 |
| mwhooker | interesting. i have to research before I know what that means | 20:45 |
| Daviey | urgh, i might be being daft | 20:45 |
| mwhooker | so it seems like perhaps 2 things need to change to resolve this: 1. make sure owner is set to owner_id. 2. check for 'owner' from glance, and not 'owner_id' | 20:46 |
| *** jsavak has quit IRC | 20:46 | |
| Daviey | mwhooker: I am being daft, it looks like it should have been directly following manifest.xml.. in human form. | 20:47 |
| mwhooker | interesting | 20:47 |
| openstackgerrit | A change was merged to openstack/openstack-ci: Updates to blueprint/topics. https://review.openstack.org/272 | 20:49 |
| openstackjenkins | Project nova build #1,264: SUCCESS in 3 min 50 sec: https://jenkins.openstack.org/job/nova/1264/ | 20:50 |
| openstackjenkins | Tarmac: Remove instances.admin_pass column. | 20:50 |
| openstackgerrit | James E. Blair proposed a change to openstack/openstack-ci-puppet: Update rfc.sh. https://review.openstack.org/273 | 20:51 |
| mwhooker | afk for lunch. See anyone at tonight's openstack meet-up in SF? | 20:51 |
| *** jsavak has joined #openstack-dev | 20:52 | |
| *** joesavak has joined #openstack-dev | 20:56 | |
| *** jsavak has quit IRC | 20:56 | |
| *** mfer has quit IRC | 20:59 | |
| *** dolphm_ has joined #openstack-dev | 21:02 | |
| *** yogirackspace has left #openstack-dev | 21:04 | |
| openstackjenkins | Project nova build #1,265: SUCCESS in 3 min 33 sec: https://jenkins.openstack.org/job/nova/1265/ | 21:20 |
| openstackjenkins | Tarmac: Cleanup the '_base' directory in libvirt tests. | 21:20 |
| *** Tushar has joined #openstack-dev | 21:22 | |
| *** troytoman is now known as troytoman-away | 21:22 | |
| soren | vishy: The doc for a number of the methods in nova.virt.fake say that they're meant to return "a task". Any idea what sort of task that is? | 21:24 |
| vishy | huh? | 21:24 |
| *** jsavak has joined #openstack-dev | 21:24 | |
| vishy | how did that get in | 21:25 |
| vishy | soren: ah, ewan added those docstrings | 21:25 |
| *** lts has quit IRC | 21:25 | |
| vishy | probably was never implemented | 21:25 |
| soren | Didn't think so. All those things are synchronous, are they not? | 21:26 |
| *** joesavak has quit IRC | 21:27 | |
| soren | vishy: I seem to remember some discussion about putting the docs in the fake rather then the abstract driver class. I don't recall why we ended up putting them in the fake. It seems odd. | 21:29 |
| *** chipc has quit IRC | 21:30 | |
| vishy | soren: there wasn't initially an abstract driver | 21:30 |
| vishy | justinsb added it, used to be just a fake | 21:31 |
| vishy | so fake was the deinition | 21:31 |
| soren | vishy: That would explain it :) | 21:31 |
| vishy | * definition | 21:31 |
| *** martine has quit IRC | 21:31 | |
| soren | The discussion must have been whether to add an abstract class just for the docs. | 21:31 |
| soren | Meh. | 21:31 |
| * soren transplants the docs from the fake into the abstract | 21:32 | |
| jk0 | bcwaldon: responded | 21:33 |
| bcwaldon | jk0: RE: testing - I was just thinking you could check that a 'rescue' action mapped to a compute api call | 21:34 |
| bcwaldon | jk0: and that error handling was working | 21:34 |
| jk0 | I think that's what the extensions suite looks at now | 21:35 |
| bcwaldon | jk0: maybe I just missed that point, let me look real quick | 21:35 |
| jk0 | I looked around, there are no unit tests for specific extensions | 21:35 |
| jk0 | other than the generic extension test suite | 21:35 |
| bcwaldon | jk0: nova/tests/api/openstack/contrib/ | 21:36 |
| jk0 | ahah | 21:36 |
| jk0 | woops | 21:36 |
| bcwaldon | jk0: yep, I'm 100% behind you once I see a few tests ;) | 21:36 |
| jk0 | nice catch, I'm not sure how I didn't see that | 21:37 |
| bcwaldon | they're hidden back there, no biggie | 21:37 |
| bcwaldon | Trust me, I always look for tests. We wouldn't have this many extensions with no test coverage | 21:37 |
| jk0 | ok, hang on | 21:37 |
| zykes- | is there any hints on when resource accounting will be in ? | 21:39 |
| vladimir3p | vishy: seems like I need your help... | 21:40 |
| vladimir3p | vishy: unittests for volume types are failing on dictionary comparisons (strings vs unicode strings). What I don't understand is how test_instance_types_extra_specs performing exactly same thing and succeeding there ... | 21:42 |
| *** amccabe has quit IRC | 21:42 | |
| vishy | interesting is it in your current code push? | 21:42 |
| vladimir3p | pushed it a min ago ... | 21:43 |
| vladimir3p | added logs there and printed dicts in both test cases - same string vs unistring compare, but mine is failing | 21:43 |
| *** jsavak has quit IRC | 21:44 | |
| bcwaldon | s1rp: is lp:~jtran/nova/lp824008 good to go in? Noticed you were the second core approve | 21:44 |
| s1rp | bcwaldon: yep, think it's good to go | 21:47 |
| bcwaldon | s1rp: Ok, I can push the button if you're busy. Otherwise, I'll just let you do it | 21:48 |
| s1rp | bcwaldon: yeah, ill go ahead and do that | 21:49 |
| bcwaldon | jk0: gotta head out for a bit. I'll check back later on your rescue branch | 21:49 |
| bcwaldon | s1rp: thank you, sir | 21:49 |
| jk0 | bcwaldon: thanks man | 21:49 |
| bcwaldon | jk0: no problem :) | 21:50 |
| *** bsza has quit IRC | 21:51 | |
| vladimir3p | vishy: what do you think about adding also volume extra_data (not type_extra_data, but volume). or we can call it meta-data in analogue to instance metadata | 21:54 |
| vishy | i think volume metadata is good too | 21:55 |
| vladimir3p | ok, will add it as well. This one we definitely require | 21:56 |
| *** bcwaldon has quit IRC | 21:59 | |
| *** rnirmal has quit IRC | 21:59 | |
| openstackjenkins | Project nova build #1,266: SUCCESS in 3 min 36 sec: https://jenkins.openstack.org/job/nova/1266/ | 22:02 |
| openstackjenkins | * Tarmac: Fixed bug in which DescribeInstances was returning deleted instances. Added tests for pertinent api methods. | 22:02 |
| openstackjenkins | * Tarmac: Accept binary user_data in radix-64 format when you launch a new server using OSAPI. This user_data would be stored along with the other server properties in the database. Once the VM instance boots you can query for the user-data to do any custom installation of applications/servers or do some specific job like setting up networking route table. | 22:02 |
| openstackjenkins | You can query for user-data using curl with the URL:- | 22:02 |
| openstackjenkins | curl http://169.254.169.254/latest/user-data | 22:02 |
| heckj | jaypipes: ping | 22:02 |
| *** rods has joined #openstack-dev | 22:06 | |
| vishy | vladimir3p: looking at your code now | 22:13 |
| *** Guest77784 has quit IRC | 22:15 | |
| vladimir3p | vishy: let me know if you will find a reason of test failure ... I gave up | 22:17 |
| vladimir3p | vishy: in 10-15 min I will push another version with VolumeMetadata | 22:18 |
| vishy | vladimir3p: name of the tests file? | 22:18 |
| vladimir3p | vishy: nova.tests.test_volume_types_extra_specs vs nova.tests.test_instance_types_extra_specs | 22:19 |
| *** zns has joined #openstack-dev | 22:19 | |
| openstackjenkins | Project nova build #1,267: SUCCESS in 3 min 32 sec: https://jenkins.openstack.org/job/nova/1267/ | 22:19 |
| openstackjenkins | Tarmac: Fixes lp828207 | 22:19 |
| openstackjenkins | The Host API reboot action was broken due to addition of a parameter to a utility method in compute/api.py | 22:19 |
| vishy | vladimir3p: it is int vs string | 22:20 |
| vishy | vladimir3p: you have to add specs as a string | 22:22 |
| vladimir3p | vishy: yeah... let me check | 22:22 |
| *** nickon has quit IRC | 22:22 | |
| *** markvoelker has quit IRC | 22:25 | |
| vladimir3p | vishy: ok, seems right. Can't test it right now. will do it a bit later | 22:26 |
| creiht | vishy: quick question about volumes | 22:29 |
| creiht | Is it the desired behavior that you have to restart the instance when you attach a volume to it for it to show up on that instance? | 22:30 |
| *** zns has quit IRC | 22:30 | |
| *** Gordonz has quit IRC | 22:37 | |
| *** dragondm has quit IRC | 22:37 | |
| *** mattray has quit IRC | 22:38 | |
| vladimir3p | vishy: need to go. just added code for metadata to the same tree (lp:~vladimir.p/nova/volume_type_extradata). ttyl | 22:46 |
| vishy | vladimir3p: awesome | 22:46 |
| openstackjenkins | Project swift build #309: SUCCESS in 29 sec: https://jenkins.openstack.org/job/swift/309/ | 22:47 |
| openstackjenkins | Tarmac: Add support for pulling quarantine stats. | 22:47 |
| *** vladimir3p has quit IRC | 22:47 | |
| *** zns has joined #openstack-dev | 22:50 | |
| *** nmistry has joined #openstack-dev | 22:54 | |
| openstackgerrit | James E. Blair proposed a change to openstack/openstack-ci: Update rfc.sh to assert divergence. https://review.openstack.org/274 | 22:55 |
| openstackgerrit | James E. Blair proposed a change to openstack/openstack-ci-puppet: Update rfc.sh to assert divergence. https://review.openstack.org/275 | 22:56 |
| openstackgerrit | Dolph Mathews proposed a change to openstack/keystone: Refactored manage.py to be both testable and useful for testing https://review.openstack.org/276 | 23:05 |
| *** nmistry has quit IRC | 23:10 | |
| *** mwhooker has quit IRC | 23:11 | |
| *** heckj has quit IRC | 23:11 | |
| openstackjenkins | Project nova build #1,268: SUCCESS in 3 min 25 sec: https://jenkins.openstack.org/job/nova/1268/ | 23:11 |
| openstackjenkins | * Tarmac: Updated docs for the recent scheduler class changes. | 23:11 |
| openstackjenkins | * Tarmac: Passes empty string instead of None to MySQLdb driver if the DB password isn't set. | 23:11 |
| *** mfer has joined #openstack-dev | 23:15 | |
| *** ohnoimdead has quit IRC | 23:22 | |
| *** mwhooker has joined #openstack-dev | 23:22 | |
| *** mwhooker has quit IRC | 23:44 | |
| openstackjenkins | Project nova build #1,269: SUCCESS in 3 min 24 sec: https://jenkins.openstack.org/job/nova/1269/ | 23:50 |
| openstackjenkins | Tarmac: Fixes issue where ServersXMLSerializer was missing a method for update actions | 23:50 |
| *** bcwaldon has joined #openstack-dev | 23:54 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!