| *** dhubbard has quit IRC | 00:01 | |
| *** dhubbard has joined #openstack-dev | 00:02 | |
| *** cp16net is now known as cp16net|away | 00:05 | |
| FatDarrel | is there a good way to install openstack on ubuntu | 00:05 |
|---|---|---|
| *** garyTh has quit IRC | 00:07 | |
| *** terryh has joined #openstack-dev | 00:07 | |
| *** ecarlin has quit IRC | 00:08 | |
| *** joshuamckenty has joined #openstack-dev | 00:08 | |
| vishy | FatDarrel: https://www.devstack.com is a nice way to give it a try | 00:09 |
| openstackgerrit | A change was merged to openstack-infra/devstack-gate: Set backdoor port to a random port. https://review.openstack.org/22603 | 00:09 |
| *** jaybuff has joined #openstack-dev | 00:11 | |
| jaybuff | hi, i'm trying to understand why jenkins isn't running some of the keystoneclient tests | 00:11 |
| *** cp16net|away is now known as cp16net | 00:12 | |
| jaybuff | specifically, tests.test_auth_token_middleware is failing for me using the latest keystoneclient code from github.com | 00:12 |
| *** garyTh has joined #openstack-dev | 00:12 | |
| henrynash | young, dolphm: gyee: young said there was an issue with v2 tokens failing policy checks? | 00:12 |
| jaybuff | it looks like https://review.openstack.org/#/c/20427/ is the commit that got merged that caused my failures to happen | 00:12 |
| openstackgerrit | A change was merged to openstack/nova: Remove function redefinitions https://review.openstack.org/22588 | 00:12 |
| jaybuff | i think the failure is related to bad test code | 00:13 |
| jaybuff | my main concern is why jenkins didn't run that particular test when that review was submitted | 00:13 |
| gyee | henrynash, what's the problem? | 00:13 |
| clarkb | jaybuff: can you link to the review/test that didn't run the test? | 00:13 |
| jaybuff | sure, i'll show the failure | 00:13 |
| henrynash | gyee: not sure, Adam said he thought I broke something :-) but will be back in a bit | 00:14 |
| *** tomoe_ has quit IRC | 00:14 | |
| henrynash | gyee: and I think I might have done…I think assert_admin will fail | 00:14 |
| *** tomoe_ has joined #openstack-dev | 00:15 | |
| jaybuff | clarkb: https://gist.github.com/jaybuff/5009696 | 00:15 |
| clarkb | jaybuff: I won't be much help sorting out why it fails locally but can look at the jenkins logs to figure out why jenkins didn't run the test | 00:15 |
| gyee | henrynash, v2 API doesn't use policies right? | 00:16 |
| *** melwitt has quit IRC | 00:16 | |
| jaybuff | clarkb: http://logs.openstack.org/20427/3/gate/gate-python-keystoneclient-python26/727/console.html.gz shows that jenkins didn't run the test | 00:16 |
| jaybuff | but the test existed in that commit | 00:17 |
| henrynash | gyee: yes, it requires everything to be admin…and asser_admin calls our policy front-end with an "illegal" rule…. | 00:17 |
| jaybuff | clarkb: those tests were run because of https://review.openstack.org/#/c/20427/ | 00:17 |
| clarkb | jaybuff: http://logs.openstack.org/20427/3/gate/gate-python-keystoneclient-python26/727/nose_results.html.gz shows why it did not run | 00:17 |
| *** dolphm has joined #openstack-dev | 00:17 | |
| *** andrewbogott is now known as andrewbogott_afk | 00:17 | |
| henrynash | gyee: since the old one used to force "rule:" on the front of very action | 00:17 |
| clarkb | jaybuff: nose does not fail if setupmodules fail | 00:18 |
| clarkb | it just moves on | 00:18 |
| henrynash | gyee: so actually, I *think* assert admin will fail for v3 and v2 ! | 00:18 |
| *** salv-orlando has quit IRC | 00:18 | |
| jaybuff | clarkb: i see | 00:18 |
| *** winston-d_ has joined #openstack-dev | 00:18 | |
| jaybuff | that nose_results.html complains about examples/pki/cms/auth_token_scoped.pem not existing | 00:19 |
| jaybuff | but that file is checked in | 00:19 |
| *** harlowja has quit IRC | 00:20 | |
| henrynash | gyee: if I'm right, it's a trivial one-line fix to assert_admin in wsgi.py…..just call the enforce with "rules:admin_required" rather than just "admin_required" | 00:20 |
| clarkb | jaybuff: I am not sure why that happened. | 00:22 |
| gyee | henrynash, we don't use assert admin in v3, we use controller.protected | 00:22 |
| henrynash | gyee: ah, right, true, but we would just to prices any v2 apis | 00:23 |
| henrynash | process.... | 00:23 |
| *** sgordon has quit IRC | 00:23 | |
| *** jbr_1 has joined #openstack-dev | 00:23 | |
| *** melwitt has joined #openstack-dev | 00:23 | |
| henrynash | gyee: we really should have a test for that :-) | 00:24 |
| * jaybuff looks to see if it is still happening | 00:24 | |
| ladquin | vishy, ping? | 00:24 |
| *** sgordon has joined #openstack-dev | 00:25 | |
| *** melwitt1 has joined #openstack-dev | 00:25 | |
| *** alszar has quit IRC | 00:26 | |
| *** joshuamckenty has quit IRC | 00:26 | |
| *** harlowja has joined #openstack-dev | 00:26 | |
| openstackgerrit | A change was merged to openstack/nova: Add missing ec2 security group quantum mixin https://review.openstack.org/22623 | 00:26 |
| gyee | henrynash, yeah, good idea | 00:27 |
| openstackgerrit | A change was merged to openstack/keystone: Change the default LDAP mapping for description. https://review.openstack.org/22352 | 00:27 |
| openstackgerrit | A change was merged to openstack-infra/devstack-gate: Add statsd to pip-requires. https://review.openstack.org/21249 | 00:27 |
| *** melwitt has quit IRC | 00:28 | |
| *** sacharya has joined #openstack-dev | 00:28 | |
| *** woodspa has quit IRC | 00:30 | |
| *** woodspa has joined #openstack-dev | 00:30 | |
| *** woodspa has quit IRC | 00:30 | |
| *** brucer_afk has joined #openstack-dev | 00:34 | |
| jaybuff | here's a change that happened today, and keystoneclient py26 is still not running AuthTokenMiddlewareTests (and some others): | 00:34 |
| jaybuff | http://logs.openstack.org/22476/1/gate/gate-python-keystoneclient-python26/757/nose_results.html | 00:34 |
| *** jbr_1 has quit IRC | 00:35 | |
| *** aeperezt has quit IRC | 00:37 | |
| *** networkstatic has joined #openstack-dev | 00:38 | |
| *** brucer has quit IRC | 00:38 | |
| *** aeperezt has joined #openstack-dev | 00:40 | |
| *** troytoman-away is now known as troytoman | 00:40 | |
| clarkb | jaybuff: now that I have looked at your original gist it looks like you an improper setUp() super call for that test class | 00:45 |
| clarkb | it keeps recursively calling itself | 00:45 |
| *** dachary has quit IRC | 00:45 | |
| jaybuff | yes, i agree the code is wrong | 00:45 |
| jaybuff | i am using the latest code | 00:46 |
| jaybuff | with no changes | 00:46 |
| clarkb | ah I see | 00:46 |
| * clarkb looks at existing code | 00:46 | |
| *** roampune1 has quit IRC | 00:46 | |
| jaybuff | the author of that patch was josh harlow, and i looked at it with him | 00:46 |
| jaybuff | and he agreed it was broken | 00:46 |
| jaybuff | my concern is that the patch passed CI | 00:46 |
| clarkb | it passed CI because nose is broken | 00:48 |
| clarkb | and I see the problem now | 00:48 |
| jaybuff | awesome | 00:48 |
| clarkb | there are two super calls in setUp for the base middle ware test class | 00:48 |
| clarkb | only the first one should exist | 00:48 |
| openstackgerrit | A change was merged to openstack-dev/devstack: Adding sysstat monitoring https://review.openstack.org/22467 | 00:49 |
| jaybuff | oh, yeah, i was able to hack around it and get the tests passing | 00:49 |
| jaybuff | josh is going to submit a fix to gerrit | 00:50 |
| openstackgerrit | A change was merged to openstack/nova: Fix baremetal migration skipping. https://review.openstack.org/22619 | 00:50 |
| openstackgerrit | A change was merged to openstack/nova: Migration 135: Fix drop_column statement https://review.openstack.org/22625 | 00:51 |
| clarkb | and the fix for nose is to stop using setUpModule then start using testr :) | 00:51 |
| *** joshuamckenty has joined #openstack-dev | 00:54 | |
| *** tomoe_ has quit IRC | 00:54 | |
| *** zzs has left #openstack-dev | 00:58 | |
| jaybuff | cool, thanks for looking at this clark | 00:59 |
| *** jaybuff has left #openstack-dev | 01:00 | |
| *** hemna has quit IRC | 01:00 | |
| *** hemna has joined #openstack-dev | 01:01 | |
| henrynash | gyee: so not so sure it is broken….since there are tests….but now I can't see how they are passing (unless of course the new policy engine allows you to pass in a rule without labelling it as such) | 01:01 |
| *** colinmcnamara has quit IRC | 01:02 | |
| *** njoy_ has quit IRC | 01:03 | |
| henrynash | gyee: ok, so I can't see how it's broken…you absolutely allow to pass in just the name of the rule on its own as an action (in fact you can't pass in "rule:action")….I'll wait to find out more Adam or others... | 01:09 |
| *** thingee is now known as thingee_zzz | 01:09 | |
| henrynash | ayoung: let me know more about how v2 token & policy is broken (not saying it isn't, just been running some tests and haven't got it to fail) | 01:10 |
| *** FatDarrel has quit IRC | 01:14 | |
| henrynash | gyee: do we allow someone to get use a v2 token to all v3 apis? | 01:15 |
| *** dolphm has quit IRC | 01:15 | |
| *** jog0 has quit IRC | 01:18 | |
| openstackgerrit | A change was merged to openstack/quantum: Imported Translations from Transifex https://review.openstack.org/22644 | 01:19 |
| *** dolphm has joined #openstack-dev | 01:19 | |
| henrynash | ayoung: is the issue that we want to support a user with a v2 token getting access to v3 apis? If so, then, yep, my building of the creeds in the protection wrapper isn't good enough | 01:19 |
| *** soody has quit IRC | 01:22 | |
| *** nati_ueno has joined #openstack-dev | 01:23 | |
| *** boden has joined #openstack-dev | 01:25 | |
| *** FatDarrel has joined #openstack-dev | 01:25 | |
| *** reed has quit IRC | 01:27 | |
| *** joshuamckenty has quit IRC | 01:28 | |
| *** nandX64 has joined #openstack-dev | 01:31 | |
| *** FatDarrel has quit IRC | 01:31 | |
| *** obondarev has quit IRC | 01:32 | |
| *** renier has quit IRC | 01:32 | |
| *** renier_ has joined #openstack-dev | 01:32 | |
| *** bdpayne has quit IRC | 01:33 | |
| gyee | henrynash, we allow v2 to v3 tokens to be use interchangeable | 01:33 |
| *** datsun180b has quit IRC | 01:34 | |
| *** obondarev has joined #openstack-dev | 01:34 | |
| *** FatDarrel has joined #openstack-dev | 01:35 | |
| *** gyee has quit IRC | 01:35 | |
| *** FatDarrel_ has joined #openstack-dev | 01:36 | |
| *** burris has joined #openstack-dev | 01:36 | |
| *** soody has joined #openstack-dev | 01:37 | |
| *** vkmc has quit IRC | 01:39 | |
| ayoung | henrynash, yeah, the building of the creds is v3 specific. The old one was v2 specific. I just pulled them together, not a big deal. But surprised it worked. | 01:40 |
| *** boden has quit IRC | 01:40 | |
| *** danwent has quit IRC | 01:40 | |
| *** danwent has joined #openstack-dev | 01:41 | |
| *** bencherian has quit IRC | 01:41 | |
| *** yaguang has joined #openstack-dev | 01:41 | |
| *** anteaya has left #openstack-dev | 01:43 | |
| *** alexxu has joined #openstack-dev | 01:43 | |
| *** FatDarrel_ has quit IRC | 01:44 | |
| ayoung | dolphm, so...the hold has been that henrynash 's latest change, which sets the policy path, seems to change the policy path for later tests | 01:44 |
| dolphm | ayoung: policy path? | 01:45 |
| ayoung | dolphm, yeah...it is _POLICY_PATH inside the policy.py file, set by... | 01:45 |
| dolphm | policy_file in keystone.conf | 01:45 |
| ayoung | dolphm, test_v3_policy.py | 01:45 |
| dolphm | ayoung: can be overridden in tests by self.opt(policy_file=etc('policy.json')) for example, which is done in keystone.test right now | 01:45 |
| ayoung | dolphm, test_v3_protect | 01:45 |
| dolphm | etcdir* | 01:46 |
| ayoung | yeah, so this in another variable | 01:46 |
| dolphm | ayoung: i believe policy path is derived from keystone.conf | 01:46 |
| ayoung | yeah, he's doing an mktemp | 01:46 |
| *** spzala has quit IRC | 01:46 | |
| *** jimfehlig has quit IRC | 01:46 | |
| ayoung | dolphm, the thing is, it is inside on test, and affecting other tests | 01:46 |
| dolphm | ayoung: it's a global variable | 01:47 |
| dolphm | ayoung: set it through conf instead | 01:47 |
| dolphm | ayoung: bbl | 01:47 |
| *** dolphm has quit IRC | 01:47 | |
| *** alexxu has quit IRC | 01:47 | |
| *** garyTh has quit IRC | 01:48 | |
| *** monst_ has quit IRC | 01:49 | |
| *** esp1 has quit IRC | 01:51 | |
| *** CaptTofu has quit IRC | 01:51 | |
| *** CaptTofu has joined #openstack-dev | 01:51 | |
| *** nandX64 has quit IRC | 01:53 | |
| *** dolphm has joined #openstack-dev | 01:54 | |
| openstackgerrit | A change was merged to openstack/ceilometer: Support different publisher interval https://review.openstack.org/22132 | 01:55 |
| *** CaptTofu has quit IRC | 01:56 | |
| ayoung | dolphm, so...I think the bleed over is happening because we on't call tear down in the base class. | 01:58 |
| dolphm | ayoung: link to patch? | 01:59 |
| *** tomoe_ has joined #openstack-dev | 02:00 | |
| *** bing_bu has joined #openstack-dev | 02:02 | |
| *** CaptTofu has joined #openstack-dev | 02:04 | |
| *** cp16net is now known as cp16net|away | 02:04 | |
| *** Nachi has joined #openstack-dev | 02:05 | |
| *** CaptTofu has quit IRC | 02:06 | |
| *** CaptTofu has joined #openstack-dev | 02:06 | |
| *** alexxu has joined #openstack-dev | 02:08 | |
| *** shang has quit IRC | 02:09 | |
| *** topol has joined #openstack-dev | 02:10 | |
| *** CaptTofu has quit IRC | 02:11 | |
| *** unix has joined #openstack-dev | 02:13 | |
| ayoung | dolphm, still looking, but | 02:14 |
| ayoung | the v3 tests don't call the tear down...I'm still tracking down the root cause.... | 02:15 |
| ayoung | Current hypothesis is that the policy CACHE needs to be flushed between tests | 02:15 |
| ayoung | dolphm, look in v3_test tearDown. It doesn't call super.tearDown, but I am guessing that is intentional | 02:16 |
| *** alexxu has quit IRC | 02:17 | |
| dolphm | ayoung: i doubt that's intentional | 02:18 |
| ayoung | and the base test does CONF.reset...but I suspect now that it is irrelevant, that the policy rules are already loaded | 02:18 |
| *** nati_uen_ has joined #openstack-dev | 02:18 | |
| *** cp16net|away is now known as cp16net | 02:18 | |
| *** nati_uen_ has quit IRC | 02:18 | |
| ayoung | dolphm, wel...you wrote it | 02:19 |
| dolphm | ayoung: i didn't write any of the policy stuff | 02:19 |
| ayoung | class RestfulTestCase(test_content_types.RestfulTestCase): | 02:19 |
| dolphm | ayoung: i revised it to support conf, but that's it | 02:19 |
| *** monst_ has joined #openstack-dev | 02:19 | |
| ayoung | dolphm, no this is in the v3 test code | 02:19 |
| *** maoy has joined #openstack-dev | 02:19 | |
| ayoung | setUp doesn't call super .setup, I assumed because you were trying to avoid the fixtures | 02:19 |
| *** nati_uen_ has joined #openstack-dev | 02:19 | |
| *** almaisan-away has quit IRC | 02:20 | |
| dolphm | ayoung: ah, that's probably true | 02:20 |
| *** kagan has quit IRC | 02:20 | |
| *** terryh has quit IRC | 02:20 | |
| ayoung | dolphm, still, I don't think that is the problem. | 02:20 |
| *** Nachi has quit IRC | 02:21 | |
| ayoung | dolphm, right now I am explicitly resetting the policy path in a teardown, but it is still failing, I assume for the same reason | 02:21 |
| ayoung | basically, my policy rules are not in the cache. | 02:21 |
| *** almaisan-away has joined #openstack-dev | 02:22 | |
| *** ctracey` has quit IRC | 02:22 | |
| *** ctracey` has joined #openstack-dev | 02:22 | |
| *** jcmartin has quit IRC | 02:23 | |
| mattrae | hi folks, when i have multiple compute nodes, instances can not resolve eachother via dns. is there an openstack component that handles this? nova-dns maybe? | 02:24 |
| mattrae | oops, meant to ask in non-dev channel | 02:24 |
| *** shang has joined #openstack-dev | 02:24 | |
| *** danwent has quit IRC | 02:25 | |
| ayoung | dolphm, the test_v3_protection is writing its own rules to the file. I think keystone.common.policy.reset() will do the trick | 02:26 |
| dolphm | ayoung: does that already exist? lol | 02:26 |
| ayoung | dolphm, right, it just isn't called yet | 02:27 |
| *** shang has quit IRC | 02:27 | |
| *** nati_uen_ has quit IRC | 02:27 | |
| ayoung | dolphm, interesting. Now, when I call reset in tearDown, my tests fail consistantly. | 02:28 |
| ayoung | not the first one, but the second and third | 02:29 |
| *** alexxu has joined #openstack-dev | 02:29 | |
| ayoung | dolphm, the cache test does an mtime, which is why it wasn't rereading the file once set back to the original | 02:30 |
| *** Nachi has joined #openstack-dev | 02:30 | |
| *** monst_ has quit IRC | 02:30 | |
| ayoung | since the cache was newer than the original file... | 02:30 |
| *** Nachi has quit IRC | 02:31 | |
| *** Nachi has joined #openstack-dev | 02:32 | |
| *** Nachi has joined #openstack-dev | 02:33 | |
| *** dolphm has quit IRC | 02:34 | |
| *** dolphm has joined #openstack-dev | 02:35 | |
| *** shang has joined #openstack-dev | 02:36 | |
| ayoung | dolphm, henrynash got it | 02:36 |
| ayoung | from keystone.policy.backends import rules | 02:36 |
| ayoung | rules.reset() | 02:36 |
| *** CaptTofu has joined #openstack-dev | 02:37 | |
| dolphm | ayoung: henrynash: cool | 02:37 |
| ayoung | dolphm, I got it, irc just didn't put a comma after henrynash . Don't want him stealing my glory. | 02:37 |
| ayoung | dolphm, so, do you like the latest V3 API changes? I put in 'impersonate' and I think we're good, at least content wise | 02:39 |
| ayoung | probably can always keep cleaning up format etc | 02:39 |
| *** terry7 has quit IRC | 02:39 | |
| ayoung | dolphm, well, except the roles thing | 02:40 |
| dolphm | ayoung: i didn't read the examples section in my last review | 02:40 |
| dolphm | ayoung: did you remove the two lists in a response thing? | 02:40 |
| dolphm | ayoung: and do you like the api better or worse after all this? lol | 02:41 |
| ayoung | dolphm, the one thing is, I was resistant to putting the full roles in, but it is probably too big a focus on performance on my part | 02:41 |
| ayoung | dolphm, API is definitely better | 02:41 |
| dolphm | ayoung: http://teacherslovesmartboards.com/wp-content/uploads/2012/03/eat-grandma.jpg | 02:41 |
| ayoung | so..I probably will do one more revision with the roles objects. | 02:42 |
| dolphm | ayoung: i think the flexibility we get down the line is worth the performance penalty | 02:42 |
| ayoung | yeah, and people hate the uuids | 02:42 |
| ayoung | rightfuly so | 02:42 |
| ayoung | so giving them the role names and the endpoint names is much friendlier | 02:43 |
| *** CaptTofu has quit IRC | 02:43 | |
| ayoung | OK...one more revision it is. | 02:43 |
| *** CaptTofu has joined #openstack-dev | 02:44 | |
| dolphm | ayoung: i've been thinking about making the generated uuid length configurable, and defaulting to 6 like in the docs, and recommended the full 32 for a real deployment | 02:44 |
| ayoung | dolphm, gyee asked if impoersonate was going to be defaulted or if it should be mandatory to specify. I think I like mandatory | 02:44 |
| dolphm | ayoung: would you be in favor? | 02:44 |
| ayoung | 6chars ids? So long as we don't call them uuids | 02:44 |
| *** yaguang has quit IRC | 02:44 | |
| dolphm | ayoung: your call on that one, as long as it's opt-in | 02:44 |
| dolphm | ayoung: i never call them uuid's ;) | 02:44 |
| ayoung | dolphm, read up " making the generated uuid length configurable" | 02:45 |
| ayoung | As Mr Connery learned, Never say Never | 02:45 |
| ayoung | opt in is a definite, I think mandatory | 02:45 |
| ayoung | creating a trust is a rare event, and people should think hard about the permissions they are giving away. Lets not hide anything from them\ | 02:46 |
| dolphm | ayoung: we could base-36 encode them to get even more variety in 6 chars | 02:46 |
| ayoung | dolphm, OK need to clean up pep 8 and then I'll ahve another patch to encode | 02:46 |
| ayoung | dolphm, I wonder if that is true. | 02:46 |
| ayoung | I wouldn't be surprised if that actually decreased the entropy | 02:47 |
| ayoung | somehow | 02:47 |
| *** Nachi has quit IRC | 02:47 | |
| *** CaptTofu has quit IRC | 02:48 | |
| dolphm | ayoung: i don't know how long 32-char base-16 UUID's would be in base-36, but it'd be the same value either way | 02:48 |
| dolphm | ayoung: there's some math in there worth doing | 02:49 |
| ayoung | dolphm, I am sorely tempted to hack in an option to run_tests.sh for short pep8 | 02:49 |
| dolphm | ayoung: a base-36 encoded uuid: 7kq687xxvyruynjf1fz8vuu6q | 02:49 |
| dolphm | short pep8? | 02:50 |
| dolphm | ayoung: oh, just write pep8 compliant code :) | 02:50 |
| ayoung | dolphm, without the explantions | 02:50 |
| dolphm | ayoung: do that on wednesday :) | 02:50 |
| *** giulivo has quit IRC | 02:50 | |
| ayoung | dolphm, yeah... | 02:50 |
| ayoung | Havana is going to be no big ticket itmes for me, just all cleanup | 02:51 |
| ayoung | unit tests and LDAP and performance... | 02:51 |
| ayoung | and a pony and the moon on a stick | 02:51 |
| *** amerine has quit IRC | 02:51 | |
| dolphm | ayoung: you really think it will stay that way? :P | 02:52 |
| *** riskable has quit IRC | 02:53 | |
| ayoung | dolphm, not for a moment | 02:53 |
| ayoung | I have too many dreams | 02:53 |
| ayoung | dolphm, but I sorely want to make HTTPD and all that gets us the default. I do not trust Eventlet, and I want Kerberos, etc... | 02:54 |
| ayoung | X509 in place of tokens is also a nice dream. | 02:54 |
| dolphm | ayoung: pow(36, 6) == 2,176,782,336 -- random enough for reliable small deployments? | 02:55 |
| *** amerine has joined #openstack-dev | 02:55 | |
| ayoung | dolphm, probably. What happens if there is a clash? | 02:55 |
| *** monst_ has joined #openstack-dev | 02:56 | |
| ayoung | It really only matters within the same entity type | 02:56 |
| dolphm | ayoung: unnexpected 409 conflict that isn't the user's fault | 02:56 |
| ayoung | I assume uuids because you want to be able to generate on different machines without deconflicting | 02:56 |
| dolphm | ayoung: yep | 02:56 |
| ayoung | deconflicting on line that is | 02:56 |
| dolphm | ayoung: and provide uniqueness among two different clouds to allow reliable federation | 02:56 |
| ayoung | dolphm, but if two users get the same UUID, from different keystone servers, it could be bad | 02:57 |
| dolphm | ayoung: quite | 02:57 |
| ayoung | same thing with, say glance images and so on | 02:57 |
| *** andrewbogott has joined #openstack-dev | 02:57 | |
| *** andrewbogott has joined #openstack-dev | 02:57 | |
| ayoung | dolphm, something to keep us up at night | 02:57 |
| dolphm | ayoung: that's why it'd have to be configurable for big deployments, and it's also something you can reconfigure as you scale up | 02:58 |
| ayoung | dolphm, it would be nice if you could do a Nosql solution for identifying that duplication has occured. Does't have to be right way, just "eventually" | 03:00 |
| *** radez is now known as radez_g0n3 | 03:00 | |
| ayoung | dolphm, OK, finally got an updated patch. The todo items I know about for it are | 03:01 |
| ayoung | 1. impersonate | 03:01 |
| ayoung | 2. objects for roels and endpoints | 03:01 |
| *** garyk has quit IRC | 03:01 | |
| ayoung | 3. v3 auth integration | 03:01 |
| *** garyk has joined #openstack-dev | 03:01 | |
| ayoung | 4. little bit of policy work | 03:01 |
| ayoung | 5. clean up the unit tests | 03:02 |
| ayoung | I think that is it | 03:02 |
| dolphm | ayoung: mind if i check it out in the morning? | 03:03 |
| ayoung | dolphm, don't mind at all | 03:03 |
| *** giulivo has joined #openstack-dev | 03:03 | |
| ayoung | I'm going to start hackin on the API doc one more time and get the objects in there. | 03:04 |
| ayoung | dolphm, we are just putting in names and ids for roles and for endpoints, right? | 03:04 |
| ayoung | no other attributes? | 03:04 |
| dolphm | ayoung: endpoints don't have names, but otherwise, yes | 03:04 |
| dolphm | ayoung: eventually links should go in there | 03:04 |
| ayoung | dolphm, the URLS? | 03:04 |
| *** unix has quit IRC | 03:04 | |
| *** baba has joined #openstack-dev | 03:04 | |
| ayoung | dolphm, that is what I mean, the links | 03:04 |
| dolphm | ayoung: i mean "links": {"self": "http://keystone:35357/v3/roles/{role_id}"} | 03:05 |
| *** monst_ has quit IRC | 03:05 | |
| ayoung | so endspoint : [ {id:bah, links: [http://blah, http://blah]}]? | 03:05 |
| dolphm | ayoung: the link back to keystone for that entity for restful hateos browseryness | 03:05 |
| ayoung | dolphm, yeah | 03:06 |
| dolphm | ayoung: not the endpoint url | 03:06 |
| *** n0ano has quit IRC | 03:06 | |
| dolphm | ayoung: but if you think the endpoint url should go in there, just stick it in the spec -- i wouldn't be opposed | 03:06 |
| ayoung | dolphm, yeah, I think it would be friendly | 03:07 |
| ayoung | list trusts could get huge | 03:07 |
| dolphm | ayoung: exclude the url's then -- we can always add them later | 03:07 |
| dolphm | ayoung: and we can't take them away later | 03:08 |
| *** KurtMartin has joined #openstack-dev | 03:09 | |
| *** woodspa has joined #openstack-dev | 03:11 | |
| *** Mandell has quit IRC | 03:16 | |
| *** amerine has quit IRC | 03:18 | |
| *** woodspa has quit IRC | 03:20 | |
| *** jcmartin has joined #openstack-dev | 03:20 | |
| *** roampune has joined #openstack-dev | 03:21 | |
| *** roampune has quit IRC | 03:23 | |
| *** roampune has joined #openstack-dev | 03:24 | |
| *** jcmartin has quit IRC | 03:28 | |
| *** yaguang has joined #openstack-dev | 03:28 | |
| *** jcmartin has joined #openstack-dev | 03:29 | |
| *** shang_ has joined #openstack-dev | 03:31 | |
| *** monst_ has joined #openstack-dev | 03:32 | |
| *** shang has quit IRC | 03:33 | |
| *** BalleS_ has joined #openstack-dev | 03:34 | |
| *** susanne-balle has quit IRC | 03:38 | |
| *** aeperezt has quit IRC | 03:39 | |
| *** monst_ has quit IRC | 03:41 | |
| *** melwitt1 has quit IRC | 03:45 | |
| *** aeperezt has joined #openstack-dev | 03:46 | |
| *** roampune has quit IRC | 03:49 | |
| *** nati_ueno_2 has joined #openstack-dev | 03:53 | |
| *** nati_ueno has quit IRC | 03:54 | |
| *** otherwiseguy has quit IRC | 03:58 | |
| *** wwallnrr__ has quit IRC | 03:58 | |
| *** epim has quit IRC | 03:59 | |
| *** martine has joined #openstack-dev | 04:08 | |
| *** todo_stack has quit IRC | 04:08 | |
| *** monst_ has joined #openstack-dev | 04:08 | |
| *** sandywalsh_ has quit IRC | 04:11 | |
| *** bencherian has joined #openstack-dev | 04:12 | |
| *** monst_ has quit IRC | 04:17 | |
| *** otherwiseguy has joined #openstack-dev | 04:20 | |
| *** Tross has joined #openstack-dev | 04:24 | |
| *** jcmartin_ has joined #openstack-dev | 04:24 | |
| *** anniec has joined #openstack-dev | 04:24 | |
| *** jcmartin has quit IRC | 04:27 | |
| *** jcmartin_ is now known as jcmartin | 04:27 | |
| *** henrynash has quit IRC | 04:31 | |
| *** smoser has quit IRC | 04:35 | |
| *** crandquist has quit IRC | 04:36 | |
| *** henrynash has joined #openstack-dev | 04:37 | |
| *** smoser has joined #openstack-dev | 04:39 | |
| *** esp has joined #openstack-dev | 04:40 | |
| *** bencherian has quit IRC | 04:44 | |
| *** monst_ has joined #openstack-dev | 04:44 | |
| *** sandywalsh has joined #openstack-dev | 04:44 | |
| *** martine has quit IRC | 04:46 | |
| *** Mandell has joined #openstack-dev | 04:51 | |
| *** mikal has quit IRC | 04:52 | |
| *** mikal has joined #openstack-dev | 04:53 | |
| *** monst_ has quit IRC | 04:53 | |
| *** mohits has joined #openstack-dev | 04:54 | |
| *** nati_ueno has joined #openstack-dev | 04:55 | |
| *** ek6 has quit IRC | 04:56 | |
| *** andrewbogott has quit IRC | 04:58 | |
| *** dolphm has quit IRC | 05:04 | |
| *** gyee has joined #openstack-dev | 05:06 | |
| *** armaan has joined #openstack-dev | 05:07 | |
| *** armaan has left #openstack-dev | 05:07 | |
| *** topol has quit IRC | 05:13 | |
| *** topol has joined #openstack-dev | 05:14 | |
| *** nunosantos has quit IRC | 05:14 | |
| *** nati_ueno has quit IRC | 05:16 | |
| *** Tross has quit IRC | 05:16 | |
| *** monst_ has joined #openstack-dev | 05:20 | |
| *** zaitcev has quit IRC | 05:25 | |
| *** aeperezt has quit IRC | 05:28 | |
| *** monst_ has quit IRC | 05:29 | |
| *** dolphm has joined #openstack-dev | 05:35 | |
| *** navid has joined #openstack-dev | 05:40 | |
| *** almaisan-away is now known as al-maisan | 05:41 | |
| *** maoy has quit IRC | 05:43 | |
| *** dolphm has quit IRC | 05:43 | |
| *** brucer_afk has quit IRC | 05:45 | |
| *** al-maisan is now known as almaisan-away | 05:46 | |
| *** koolhead17 has joined #openstack-dev | 05:49 | |
| *** monst_ has joined #openstack-dev | 05:56 | |
| *** yaguang has quit IRC | 05:57 | |
| *** nati_ueno_2 has quit IRC | 06:02 | |
| *** nati_ueno has joined #openstack-dev | 06:02 | |
| *** cdub_ has quit IRC | 06:06 | |
| *** cdub_ has joined #openstack-dev | 06:06 | |
| *** monst_ has quit IRC | 06:07 | |
| *** gyee has quit IRC | 06:08 | |
| *** aswadrangnekar has left #openstack-dev | 06:10 | |
| *** markmc has joined #openstack-dev | 06:11 | |
| *** olaph has quit IRC | 06:13 | |
| *** olaph has joined #openstack-dev | 06:14 | |
| openstackgerrit | A change was merged to openstack/quantum: Provide a default api_extensions_path for nvp_plugin https://review.openstack.org/22641 | 06:16 |
| *** esp has quit IRC | 06:21 | |
| *** yjiang5_away has quit IRC | 06:23 | |
| *** almaisan-away is now known as al-maisan | 06:25 | |
| *** Mandell has quit IRC | 06:29 | |
| *** Ryan_Lane has quit IRC | 06:30 | |
| *** KurtMartin has quit IRC | 06:30 | |
| *** sacharya has quit IRC | 06:31 | |
| *** monst_ has joined #openstack-dev | 06:33 | |
| *** anniec has quit IRC | 06:34 | |
| *** boris-42 has joined #openstack-dev | 06:41 | |
| *** alunduil has joined #openstack-dev | 06:43 | |
| *** monst_ has quit IRC | 06:43 | |
| *** topol has quit IRC | 06:44 | |
| *** llu has quit IRC | 06:44 | |
| *** alunduil has quit IRC | 06:45 | |
| *** soody has quit IRC | 06:49 | |
| *** soody has joined #openstack-dev | 06:53 | |
| *** ek6 has joined #openstack-dev | 06:56 | |
| *** PaulCarlton has joined #openstack-dev | 07:02 | |
| *** baba has quit IRC | 07:03 | |
| *** techlife has quit IRC | 07:05 | |
| *** techlife has joined #openstack-dev | 07:05 | |
| *** dedis has quit IRC | 07:07 | |
| *** afrittoli has joined #openstack-dev | 07:09 | |
| *** troytoman is now known as troytoman-away | 07:09 | |
| *** amerine has joined #openstack-dev | 07:09 | |
| *** monst_ has joined #openstack-dev | 07:09 | |
| *** andreaf has quit IRC | 07:10 | |
| *** unix has joined #openstack-dev | 07:12 | |
| *** morganfainberg has left #openstack-dev | 07:14 | |
| openstackgerrit | A change was merged to openstack/nova: import base_dir_name config option into vmwareapi https://review.openstack.org/22577 | 07:18 |
| *** alunduil has joined #openstack-dev | 07:19 | |
| *** yaguang has joined #openstack-dev | 07:19 | |
| *** monst_ has quit IRC | 07:20 | |
| *** yolanda has joined #openstack-dev | 07:22 | |
| *** alunduil has quit IRC | 07:22 | |
| *** afazekas has joined #openstack-dev | 07:27 | |
| *** Ryan_Lane has joined #openstack-dev | 07:28 | |
| *** kbrierly has quit IRC | 07:29 | |
| *** esp has joined #openstack-dev | 07:31 | |
| *** kbrierly has joined #openstack-dev | 07:33 | |
| *** smulcahy has quit IRC | 07:34 | |
| *** llu has joined #openstack-dev | 07:35 | |
| *** smulcahy has joined #openstack-dev | 07:35 | |
| *** mindpixel has joined #openstack-dev | 07:38 | |
| *** alunduil has joined #openstack-dev | 07:43 | |
| *** unix has quit IRC | 07:43 | |
| *** unix has joined #openstack-dev | 07:43 | |
| *** boris-42 has quit IRC | 07:44 | |
| *** alunduil has quit IRC | 07:45 | |
| *** Mandell has joined #openstack-dev | 07:46 | |
| *** masumotok has joined #openstack-dev | 07:46 | |
| *** monst_ has joined #openstack-dev | 07:47 | |
| masumotok | Currently, I did "git-review", and jenkins raise error. It is not related to my patch, but my environment also raise same error. After I checked it, I realize library version problem of Jenkins, because if I try pyzmq==2.2.0.1 environment, it worsk, but pyzmq=13.0.0 environment, it didnt. | 07:49 |
| masumotok | Does anyone know who should I ask about Jenkins? | 07:49 |
| lifeless | masumotok: requires.txt describes the dependencies | 07:49 |
| lifeless | masumotok: you can ask on #openstack-infra if you want details about jenkins | 07:50 |
| *** soody has quit IRC | 07:50 | |
| masumotok | lifeless: thank you. I'll visit openstack-infra. | 07:50 |
| *** esp has left #openstack-dev | 07:51 | |
| *** masumotok has left #openstack-dev | 07:51 | |
| *** monst_ has quit IRC | 07:59 | |
| *** unix has quit IRC | 07:59 | |
| *** unix has joined #openstack-dev | 07:59 | |
| *** pnavarro has joined #openstack-dev | 08:00 | |
| *** tomoe_ has quit IRC | 08:02 | |
| *** tomoe_ has joined #openstack-dev | 08:02 | |
| *** rafaduran has joined #openstack-dev | 08:03 | |
| *** FlorianOtel has joined #openstack-dev | 08:05 | |
| *** xga_ has joined #openstack-dev | 08:05 | |
| *** tomoe_ has quit IRC | 08:07 | |
| *** koolhead17 has quit IRC | 08:09 | |
| *** adjohn has quit IRC | 08:09 | |
| *** FlorianOtel has quit IRC | 08:11 | |
| *** thouveng has joined #openstack-dev | 08:14 | |
| *** ondergetekende has joined #openstack-dev | 08:15 | |
| *** henrynash has quit IRC | 08:18 | |
| garyk | arosen: you around | 08:18 |
| *** mmagr has joined #openstack-dev | 08:22 | |
| *** unix has quit IRC | 08:24 | |
| *** unix has joined #openstack-dev | 08:24 | |
| *** FlorianOtel has joined #openstack-dev | 08:24 | |
| *** monst_ has joined #openstack-dev | 08:24 | |
| *** salv-orlando has joined #openstack-dev | 08:28 | |
| *** zoresvit has joined #openstack-dev | 08:30 | |
| *** gael has joined #openstack-dev | 08:32 | |
| *** davidha has quit IRC | 08:33 | |
| *** psedlak has joined #openstack-dev | 08:34 | |
| *** monst_ has quit IRC | 08:36 | |
| *** zoresvit has quit IRC | 08:37 | |
| *** jgallard has joined #openstack-dev | 08:37 | |
| *** nsatterl_ has joined #openstack-dev | 08:40 | |
| *** adjohn has joined #openstack-dev | 08:40 | |
| *** dachary has joined #openstack-dev | 08:41 | |
| *** zing has joined #openstack-dev | 08:42 | |
| *** shardy_afk is now known as shardy | 08:49 | |
| *** adjohn has quit IRC | 08:49 | |
| *** torandu has quit IRC | 08:52 | |
| *** mattymo has joined #openstack-dev | 08:53 | |
| *** zoresvit has joined #openstack-dev | 08:53 | |
| ttx | nijaba: cutting g3 for ceilometer | 08:54 |
| *** torandu has joined #openstack-dev | 08:54 | |
| *** eglynn has joined #openstack-dev | 08:55 | |
| *** jpich has joined #openstack-dev | 08:55 | |
| *** winston-d_ has quit IRC | 08:57 | |
| *** pixelbeat has joined #openstack-dev | 09:00 | |
| *** darjeeling has quit IRC | 09:01 | |
| zykes- | garyk: question: Did LBaaS land ? | 09:01 |
| Daviey | zykes-: Err, you know G-3 has only just been cut now? | 09:01 |
| zykes- | Daviey: yeh :) | 09:01 |
| zykes- | I was a bit over-eager | 09:02 |
| Daviey | zykes-: then give us a chance :P | 09:02 |
| zykes- | :P | 09:02 |
| * zykes- brings out the whip, mosh! | 09:02 | |
| *** dosaboy has joined #openstack-dev | 09:02 | |
| *** monst_ has joined #openstack-dev | 09:02 | |
| *** jprovazn has joined #openstack-dev | 09:04 | |
| *** ndipanov has joined #openstack-dev | 09:05 | |
| *** dachary has quit IRC | 09:06 | |
| zykes- | Daviey: btw, python-ceilometer takes in mongodb | 09:08 |
| *** boris-42 has joined #openstack-dev | 09:08 | |
| *** koolhead17 has joined #openstack-dev | 09:08 | |
| zykes- | ehm, ceilometer-agent-compute that is | 09:08 |
| *** boden has joined #openstack-dev | 09:09 | |
| Daviey | zykes-: Great! | 09:09 |
| zykes- | Daviey: where would you have me report bugs ? :p | 09:10 |
| ttx | nijaba: https://launchpad.net/ceilometer/grizzly/grizzly-3 <- you can announce it | 09:10 |
| zykes- | ttx: are there some features missing for quantum ? | 09:10 |
| ttx | bah, I'll just do it | 09:10 |
| *** dachary has joined #openstack-dev | 09:11 | |
| ttx | zykes-: for some definition of "missing" yes. | 09:11 |
| ttx | if by missing you mean "added on the last week and did not make it", then yes | 09:11 |
| ttx | alaso a number of features were granted an exception and will land next week | 09:11 |
| *** derekh has joined #openstack-dev | 09:11 | |
| ttx | like lbaas | 09:12 |
| Daviey | zykes-: ubuntu-bug ceilometer .. or by hand, https://bugs.launchpad.net/ubuntu/+source/ceilometer/+filebug | 09:12 |
| Daviey | zykes-: the former also includes som data about package version etc. | 09:12 |
| zykes- | ttx: lbaas will be going in ? | 09:13 |
| koolhead17 | al-maisan: slides of your heat demo needed :) | 09:14 |
| ttx | zykes-: that's what has been promised to me | 09:14 |
| *** Mandell has quit IRC | 09:14 | |
| ttx | zykes-: https://launchpad.net/quantum/+milestone/grizzly-rc1 | 09:14 |
| *** darraghb has joined #openstack-dev | 09:14 | |
| *** monst_ has quit IRC | 09:15 | |
| *** afazekas has quit IRC | 09:15 | |
| zykes- | koolhead17: which demo ? : p | 09:15 |
| *** adjohn has joined #openstack-dev | 09:15 | |
| al-maisan | koolhead17: https://github.com/al-maisan/p/blob/master/2013-02-19-chosug.zh.heat.pdf?raw=true | 09:15 |
| koolhead17 | zykes-: did a nice heat demo at swiss OS meetup | 09:16 |
| zykes- | k | 09:16 |
| koolhead17 | al-maisan: planning to blog with deployment instructions sir :P | 09:17 |
| al-maisan | koolhead17: ah, nice :-) | 09:18 |
| *** jruzicka has joined #openstack-dev | 09:18 | |
| koolhead17 | al-maisan: i was poking you 4 that :P | 09:18 |
| *** bing_bu has quit IRC | 09:19 | |
| al-maisan | no problem :-) when you install/deploy on ubutu 12.04 the following may be handy: https://bugs.launchpad.net/heat/+bug/1127886 | 09:19 |
| uvirtbot | Launchpad bug 1127886 in heat "Fix ubuntu 12.04 installation problems" [Undecided,New] | 09:19 |
| *** adjohn has quit IRC | 09:20 | |
| *** nati_ueno has quit IRC | 09:23 | |
| *** nati_ueno has joined #openstack-dev | 09:24 | |
| *** doude has joined #openstack-dev | 09:27 | |
| *** networkstatic has quit IRC | 09:27 | |
| *** trapni has joined #openstack-dev | 09:28 | |
| *** amerine has quit IRC | 09:28 | |
| *** matelakat has joined #openstack-dev | 09:31 | |
| *** nati_ueno_2 has joined #openstack-dev | 09:32 | |
| *** trapnii has joined #openstack-dev | 09:32 | |
| *** nati_ueno has quit IRC | 09:34 | |
| *** trapni has quit IRC | 09:35 | |
| nijaba | ttx: ack | 09:38 |
| *** egallen has joined #openstack-dev | 09:39 | |
| *** monst_ has joined #openstack-dev | 09:41 | |
| ttx | nijaba: I made a follow-up announcement | 09:42 |
| nijaba | ttx: where? | 09:42 |
| ttx | on the openstack list | 09:43 |
| * nijaba looks | 09:43 | |
| ttx | nijaba: https://lists.launchpad.net/openstack/msg21181.html | 09:43 |
| nijaba | ttx: nice. thanks | 09:44 |
| *** darjeeling has joined #openstack-dev | 09:44 | |
| *** al-maisan is now known as almaisan-away | 09:44 | |
| *** trapnii has quit IRC | 09:47 | |
| *** henrynash has joined #openstack-dev | 09:47 | |
| *** xga_ has quit IRC | 09:48 | |
| *** danpb has joined #openstack-dev | 09:51 | |
| *** nati_ueno_2 has quit IRC | 09:53 | |
| *** monst_ has quit IRC | 09:53 | |
| *** nati_ueno has joined #openstack-dev | 09:54 | |
| *** k4n0 has joined #openstack-dev | 09:54 | |
| *** nati_ueno has quit IRC | 09:58 | |
| *** trapnii has joined #openstack-dev | 10:00 | |
| *** alexxu has quit IRC | 10:10 | |
| *** adjohn has joined #openstack-dev | 10:16 | |
| *** unix has quit IRC | 10:18 | |
| *** unix has joined #openstack-dev | 10:19 | |
| *** monst_ has joined #openstack-dev | 10:19 | |
| *** mattmonkeymagic has joined #openstack-dev | 10:20 | |
| mattmonkeymagic | anyone around for a Filesystem question on compute node?? | 10:20 |
| *** adjohn has quit IRC | 10:20 | |
| mattmonkeymagic | having a prob with compute node FS filling up with snapshots, but there doesn't seem to be anything in OpenStack monitoring if a nodes storage is reaching capacity. Is there a config setting I can change to warn on disk full?? | 10:26 |
| *** jprovazn has quit IRC | 10:27 | |
| *** gray-- has joined #openstack-dev | 10:29 | |
| *** monst_ has quit IRC | 10:32 | |
| *** hughsaunders has quit IRC | 10:35 | |
| mattmonkeymagic | as in the node will still accept requests for new instances to fire up, even tho that node no longer has sufficient capacity to fulfil request | 10:35 |
| *** dachary has quit IRC | 10:39 | |
| *** dachary has joined #openstack-dev | 10:41 | |
| *** darjeeling has quit IRC | 10:43 | |
| *** henrynash has quit IRC | 10:49 | |
| maroh | gongysh : hi | 10:50 |
| *** ndipanov has quit IRC | 10:53 | |
| *** henrynash has joined #openstack-dev | 10:54 | |
| *** monst_ has joined #openstack-dev | 10:58 | |
| *** henrynash has quit IRC | 11:00 | |
| *** unix has quit IRC | 11:01 | |
| *** henrynash has joined #openstack-dev | 11:01 | |
| gongysh | maroh: hi | 11:04 |
| *** unix has joined #openstack-dev | 11:05 | |
| *** mkollaro has joined #openstack-dev | 11:07 | |
| *** monst_ has quit IRC | 11:10 | |
| *** pcm__ has joined #openstack-dev | 11:19 | |
| *** pcm__ has quit IRC | 11:19 | |
| *** henrynash has quit IRC | 11:19 | |
| *** pcm_ has joined #openstack-dev | 11:20 | |
| *** mohits has quit IRC | 11:21 | |
| *** olaph has quit IRC | 11:23 | |
| *** olaph has joined #openstack-dev | 11:25 | |
| *** darjeeling has joined #openstack-dev | 11:26 | |
| *** unix has quit IRC | 11:27 | |
| *** unix has joined #openstack-dev | 11:32 | |
| *** zbitter has joined #openstack-dev | 11:37 | |
| *** zaneb has quit IRC | 11:37 | |
| *** monst_ has joined #openstack-dev | 11:37 | |
| *** ndipanov has joined #openstack-dev | 11:37 | |
| *** davidha has joined #openstack-dev | 11:39 | |
| *** martine has joined #openstack-dev | 11:39 | |
| *** yaguang has quit IRC | 11:40 | |
| *** sthaha has joined #openstack-dev | 11:43 | |
| *** vkmc has joined #openstack-dev | 11:45 | |
| *** adjohn has joined #openstack-dev | 11:47 | |
| *** gray-- has quit IRC | 11:48 | |
| *** monst_ has quit IRC | 11:49 | |
| *** adjohn has quit IRC | 11:51 | |
| *** dachary1 has joined #openstack-dev | 11:52 | |
| *** dachary has quit IRC | 11:55 | |
| openstackgerrit | A change was merged to openstack/tempest: Simplify xml-json inheritance in compute https://review.openstack.org/22112 | 11:59 |
| smulcahy | sdague: thanks for that review | 12:00 |
| sdague | smulcahy: no prob | 12:00 |
| sdague | thanks for that super quick turn around :) | 12:00 |
| openstackgerrit | A change was merged to openstack/tempest: Have all test case to use a single base class https://review.openstack.org/20687 | 12:02 |
| *** arbrandes has joined #openstack-dev | 12:06 | |
| *** dachary1 is now known as dachary | 12:08 | |
| *** monst_ has joined #openstack-dev | 12:16 | |
| openstackgerrit | A change was merged to openstack-dev/devstack: Update cinder to use rootwrap_config https://review.openstack.org/22474 | 12:16 |
| openstackgerrit | A change was merged to openstack-dev/devstack: Fix EXTRA_BAREMETAL_OPTS https://review.openstack.org/22281 | 12:18 |
| *** mkollaro has quit IRC | 12:22 | |
| openstackgerrit | A change was merged to openstack-dev/devstack: replace use of cirros 0.3.0 with 0.3.1 https://review.openstack.org/22452 | 12:22 |
| openstackgerrit | A change was merged to openstack-dev/devstack: Deprecate DATABASE_TYPE and use_database https://review.openstack.org/22635 | 12:26 |
| *** romcheg has joined #openstack-dev | 12:26 | |
| *** monst_ has quit IRC | 12:28 | |
| *** romcheg1 has joined #openstack-dev | 12:28 | |
| *** maoy has joined #openstack-dev | 12:28 | |
| *** adalbas has joined #openstack-dev | 12:30 | |
| openstackgerrit | A change was merged to openstack-dev/devstack: Upload XenServer style ovf directly to glance https://review.openstack.org/21354 | 12:30 |
| *** romcheg has quit IRC | 12:32 | |
| *** mkollaro has joined #openstack-dev | 12:35 | |
| *** unix has quit IRC | 12:35 | |
| *** tomoe_ has joined #openstack-dev | 12:36 | |
| smulcahy | sdague: keen to get it merged before the weekend | 12:36 |
| smulcahy | is jenkins normally that slow verifying changes or is it v busy? | 12:37 |
| sdague | smulcahy: it should go through | 12:37 |
| sdague | what was the review again? | 12:37 |
| smulcahy | sdague: https://review.openstack.org/#/c/22533 | 12:39 |
| sdague | smulcahy: oh, well we need dtroyer to take a look again. even though it's just dean and I doing core reviews, we try to make sure we both +2 things before they go in | 12:39 |
| smulcahy | ah, musta heard me :) | 12:39 |
| sdague | smulcahy: yeh, figure ~1 hr for the verify | 12:39 |
| smulcahy | sdague: ah ok, devstack tests are bigger than cinder ones I guess, makes sense :) | 12:40 |
| sdague | yeh, it brings up a full devstack node | 12:40 |
| sdague | and runs the 700 tempest tests o nit | 12:40 |
| sdague | which means spawning vms, creating volumes, all sorts of things | 12:40 |
| *** johnthetubaguy1 has joined #openstack-dev | 12:41 | |
| *** timello has quit IRC | 12:41 | |
| smulcahy | sure | 12:41 |
| smulcahy | ok, hopefully dean will get a chance to look at it today | 12:41 |
| mordred | sdague: guess what? | 12:43 |
| sdague | mordred: .... oh no | 12:43 |
| sdague | what? | 12:43 |
| mordred | sdague: I think that pyzmq made a new release, and I think that in places that we use it we're not pinned, and I think that in those places it can break us | 12:43 |
| mordred | sdague: http://logs.openstack.org/22523/3/check/gate-oslo-incubator-python27/1113/nose_results.html | 12:43 |
| sdague | well, at least it can only break oslo, right? | 12:44 |
| mordred | dunno - I have no idea what's actually going on here | 12:44 |
| *** CaptTofu has joined #openstack-dev | 12:45 | |
| mordred | other than pyzmq _did_ make a release yesterday, and now I've got a failing test case :( | 12:45 |
| mordred | I'm going to try to dig in further - but figured I should toss out a heads-up | 12:45 |
| sdague | mordred: I'd suggest pinning in oslo, I think that's the only place it will splode | 12:45 |
| mordred | great. I've got a patch for that coming | 12:45 |
| sdague | as far as I know nothing else exercises the pyzmq stuff | 12:46 |
| sdague | ewindisch would be the person to ask as well, I think they are basically the only ones that use zmq for real | 12:46 |
| *** johnthetubaguy1 has left #openstack-dev | 12:47 | |
| *** DuncanT has joined #openstack-dev | 12:47 | |
| *** adjohn has joined #openstack-dev | 12:48 | |
| k4n0 | https://review.openstack.org/#/c/21415/ , please review this guys :) | 12:49 |
| *** rods has joined #openstack-dev | 12:51 | |
| *** adjohn has quit IRC | 12:53 | |
| *** Guest86187 has quit IRC | 12:54 | |
| *** monst_ has joined #openstack-dev | 12:55 | |
| *** markvoelker has joined #openstack-dev | 12:56 | |
| *** yamahata has joined #openstack-dev | 12:59 | |
| openstackgerrit | A change was merged to openstack/nova: Fix launching libvirt instances with swap https://review.openstack.org/22586 | 13:02 |
| *** yolanda has quit IRC | 13:03 | |
| *** davidha has quit IRC | 13:03 | |
| *** timello has joined #openstack-dev | 13:04 | |
| *** davidha has joined #openstack-dev | 13:04 | |
| *** aeperezt has joined #openstack-dev | 13:05 | |
| *** mattmonkeymagic has quit IRC | 13:07 | |
| *** monst_ has quit IRC | 13:08 | |
| *** yolanda has joined #openstack-dev | 13:09 | |
| *** unix has joined #openstack-dev | 13:12 | |
| sandywalsh | can't get a venv going for tests ... anyone seen this: "No distributions at all found for oslo-config from http://tarballs.openstack.org/oslo-config/oslo-config-2013.1b4.tar.gz" | 13:18 |
| *** adjohn has joined #openstack-dev | 13:19 | |
| *** adalbas has quit IRC | 13:21 | |
| *** danwent has joined #openstack-dev | 13:23 | |
| *** Tross has joined #openstack-dev | 13:24 | |
| *** sandywalsh has quit IRC | 13:24 | |
| *** adjohn has quit IRC | 13:24 | |
| mordred | markmc: so, I think this: http://logs.openstack.org/22523/4/check/gate-oslo-incubator-python27/1118/nose_results.html | 13:28 |
| mordred | markmc: is uncovering an actual bug | 13:28 |
| mordred | markmc: but I don't understand the intent of the test | 13:28 |
| mordred | markmc: as best I can see - exception.wrap_exception should always only ever throw an Error - not the wrapped exception | 13:29 |
| mordred | I'm not really sure the previous use of the base Exception class was causing that test to pass - but I'm fairly certain it's a bad test | 13:30 |
| mordred | also - I'd like to take this moment to register my EXTREME objection to catching one exception and then raising a different one | 13:30 |
| mordred | so, I think this chunk of code should diaf anyway - but that's a whole other kettle of fish to be sorted out after fixing the testing of the current code | 13:32 |
| markmc | mordred, well, the current code looks correct - i.e. the test should pass | 13:32 |
| mordred | markmc: really? can you explain why you think that? | 13:32 |
| markmc | mordred, but it does look like a big pile of bogusness more generally | 13:32 |
| mordred | markmc: as far as I can see - if your wrapped function raised an exception based on Error, it will re-raise | 13:33 |
| markmc | mordred, bad function throws exception, wrap exception catches and throws Error, Error is a subclass of Exception, we assertRaises(Exception) | 13:33 |
| *** adalbas has joined #openstack-dev | 13:33 | |
| markmc | mordred, actually, sorry | 13:34 |
| *** monst_ has joined #openstack-dev | 13:34 | |
| mordred | yah. k. that's why it passes - but I don't understand why it's testing that | 13:34 |
| markmc | mordred, bad function throws exception, wrap exception catches it and re-raises it because it's not an Error, we assertRaises(Exception) | 13:34 |
| markmc | mordred, it's testing its behaviour, whether the behaviour is useful ... :) | 13:35 |
| mordred | markmc: but no - wrap_exception doesn't re-raise - it raises a new Error | 13:35 |
| markmc | mordred, I'm all for just deleting the exception module from oslo | 13:35 |
| mordred | I think the test was intending to test that the function re-raises | 13:35 |
| markmc | mordred, only it caught an Error instance | 13:35 |
| markmc | mordred, you're right | 13:35 |
| mordred | but because it used a test Exception that was also the base class of Error... | 13:35 |
| markmc | mordred, if it's not an Error, it raises an Error | 13:35 |
| mordred | yah. which is HORRIBLE, btw | 13:36 |
| markmc | agree | 13:36 |
| mordred | how about I just delete this module :) | 13:36 |
| markmc | looks like wsgi is the only module using exceptions from it | 13:36 |
| markmc | move those exceptions into wsgi.py and that's probably enough to delete it | 13:36 |
| mordred | consider it done | 13:37 |
| markmc | cheers | 13:37 |
| *** anteaya has joined #openstack-dev | 13:37 | |
| *** sandywalsh has joined #openstack-dev | 13:37 | |
| *** Tross has quit IRC | 13:40 | |
| *** radez_g0n3 is now known as radez | 13:43 | |
| *** mlavalle has left #openstack-dev | 13:46 | |
| *** soody has joined #openstack-dev | 13:46 | |
| *** monst_ has quit IRC | 13:46 | |
| *** CaptTofu has quit IRC | 13:48 | |
| *** CaptTofu has joined #openstack-dev | 13:48 | |
| mordred | markmc: https://review.openstack.org/22681 is a stab at it | 13:48 |
| *** adjohn has joined #openstack-dev | 13:50 | |
| *** annegentle has joined #openstack-dev | 13:52 | |
| *** dprince has joined #openstack-dev | 13:52 | |
| mordred | except that I suck. | 13:53 |
| *** CaptTofu has quit IRC | 13:53 | |
| *** k4n0 has left #openstack-dev | 13:54 | |
| *** adjohn has quit IRC | 13:54 | |
| *** pasquier-s has joined #openstack-dev | 13:55 | |
| *** rkukura has joined #openstack-dev | 13:55 | |
| *** maoy has quit IRC | 13:56 | |
| *** kbringard has joined #openstack-dev | 13:58 | |
| *** sandywalsh has quit IRC | 14:00 | |
| *** almaisan-away is now known as al-maisan | 14:03 | |
| *** sandywalsh has joined #openstack-dev | 14:03 | |
| *** Tross has joined #openstack-dev | 14:04 | |
| *** henrynash has joined #openstack-dev | 14:04 | |
| *** martine has quit IRC | 14:06 | |
| *** ondergetekende has quit IRC | 14:08 | |
| *** martine has joined #openstack-dev | 14:09 | |
| *** monst_ has joined #openstack-dev | 14:12 | |
| dachary | jdurgin Do you know if https://wiki.openstack.org/wiki/BlockDeviceConfig has been discussed already ? During meetings / on lists. It looks like a change that will trigger a fair amount of discussion. It recommends adding a number of new fields. I'm very much in favor of giving more control to the user over the allocation of disks because it often frustrated me to not be able to fine tune disks allocation. | 14:12 |
| *** FlorianOtel has quit IRC | 14:14 | |
| *** eharney has joined #openstack-dev | 14:15 | |
| *** al-maisan is now known as almaisan-away | 14:16 | |
| *** renier has joined #openstack-dev | 14:17 | |
| *** bknudson has quit IRC | 14:17 | |
| *** renier_ has quit IRC | 14:17 | |
| ewindisch | mordred: I'll look at the new pyzmq release & compat, but pinning sounds reasonable for now. | 14:19 |
| mordred | ewindisch: ossum | 14:19 |
| *** CaptTofu has joined #openstack-dev | 14:19 | |
| *** CaptTofu has quit IRC | 14:20 | |
| *** CaptTofu has joined #openstack-dev | 14:21 | |
| *** BLZbubba_ has joined #openstack-dev | 14:23 | |
| *** monst_ has quit IRC | 14:24 | |
| ewindisch | also, this got a FFE but hasn't gotten any eyes: https://review.openstack.org/#/c/20434/ | 14:24 |
| *** CaptTofu has quit IRC | 14:25 | |
| *** henrynash has quit IRC | 14:27 | |
| *** henrynash has joined #openstack-dev | 14:29 | |
| *** maoy has joined #openstack-dev | 14:30 | |
| *** martine has quit IRC | 14:30 | |
| *** bknudson has joined #openstack-dev | 14:30 | |
| *** mindpixel has quit IRC | 14:31 | |
| *** davidkranz has joined #openstack-dev | 14:31 | |
| *** unix has quit IRC | 14:32 | |
| *** unix has joined #openstack-dev | 14:33 | |
| openstackgerrit | A change was merged to openstack/oslo-incubator: Fix "Needs to specify pyzmq version" https://review.openstack.org/22672 | 14:33 |
| *** martine has joined #openstack-dev | 14:36 | |
| sandywalsh | am I reading this correctly, that a postgres client is required to run tests now? | 14:39 |
| *** dolphm has joined #openstack-dev | 14:39 | |
| *** markmcclain has joined #openstack-dev | 14:41 | |
| *** CaptTofu has joined #openstack-dev | 14:41 | |
| *** woodspa has joined #openstack-dev | 14:45 | |
| *** armaan has joined #openstack-dev | 14:46 | |
| *** anniec has joined #openstack-dev | 14:46 | |
| *** READ10 has joined #openstack-dev | 14:47 | |
| ewindisch | sandywalsh: where are you reading that? I don't see postgres in pip-requires/test-requires | 14:49 |
| danpb | sandywalsh: yep, hit that the other day - needed to install postgresql-devel to bootstrap | 14:50 |
| *** jruzicka has quit IRC | 14:50 | |
| *** monst_ has joined #openstack-dev | 14:50 | |
| sandywalsh | that's kinda shitty, no? Shouldn't it just pass the tests if it doesn't find the client? | 14:51 |
| *** adjohn has joined #openstack-dev | 14:51 | |
| sandywalsh | and I'm seeing problems with getting from pypi lxml under python 2.6 ... works fine under 2.7 (when setting up a fresh venv) | 14:52 |
| *** mtreinish has joined #openstack-dev | 14:52 | |
| sandywalsh | ewindisch, it's the psycops (?) package that needs it ... Feb 4 merge | 14:52 |
| danpb | ewindisch: tools/test-requires:psycopg2 | 14:52 |
| *** anniec has quit IRC | 14:52 | |
| ewindisch | danpb: I was looking in oslo-incubator… not there | 14:53 |
| danpb | commit 5dcf4dd9b2e996913c8fe4bfa20a6186e9a066ab | 14:53 |
| danpb | Author: Kurt Taylor <krtaylor@us.ibm.com> | 14:53 |
| danpb | Date: Thu Jan 17 15:15:59 2013 -0500 | 14:53 |
| danpb | Added Postgres CI opportunistic test case | 14:53 |
| danpb | its in Nova | 14:53 |
| sdague | sandywalsh: this remains the problem of no test optionals | 14:53 |
| *** anniec has joined #openstack-dev | 14:53 | |
| *** jruzicka has joined #openstack-dev | 14:53 | |
| sandywalsh | sdague, how do the zeromq tests deal with it? | 14:53 |
| sdague | we can revisit again, but there is a desire that the installed set of software for ci is the same as for developers, | 14:54 |
| ewindisch | sandywalsh: I do think that if postgresql is supported at all, it should be in the test-requires at a minimum | 14:54 |
| sdague | sandywalsh: it might be a pure python driver | 14:54 |
| ewindisch | sandywalsh: we now require installing zeromq and pyzmq | 14:54 |
| ewindisch | *libzmq | 14:54 |
| sdague | ewindisch: it is in test-requires | 14:54 |
| sandywalsh | ugh | 14:54 |
| *** armaan has left #openstack-dev | 14:54 | |
| ewindisch | sandywalsh: although the latest pyzmq includes the source to libzmq and compiles it | 14:54 |
| sandywalsh | ewindisch, turning into "download the internet" | 14:55 |
| sdague | sandywalsh: have you built a venv from scratch recently for nova? :) | 14:55 |
| sdague | we're about 1/2 way there already | 14:55 |
| sandywalsh | sdague, that's what I'm trying to do now | 14:55 |
| sdague | you get really good at compiling numpy | 14:55 |
| ewindisch | sandywalsh: no different than having to install mysqlclient. | 14:55 |
| *** adjohn has quit IRC | 14:55 | |
| sandywalsh | sdague, can't get it to work under 2.6 though | 14:56 |
| sdague | sandywalsh: really? | 14:56 |
| sandywalsh | ewindisch, yeah, I guess. Sort of depends on "what is the core deploy?" | 14:56 |
| *** garyTh has joined #openstack-dev | 14:56 | |
| ewindisch | sandywalsh: well, if pyzmq wasn't in test-requires for oslo, that could would break. It *has* broken from not being tested. | 14:57 |
| sdague | sandywalsh: so there is probably a good conversation for summit about test-optional as a 3rd list, now that we have more of these tools | 14:57 |
| ewindisch | I assume that is a lesson that can be applied elsewhere to code that doesn't get tested. | 14:57 |
| ewindisch | sdague: and what is the value? that people will only realize they broke something after it gets to Jenkins? | 14:58 |
| sandywalsh | sdague, +1 | 14:58 |
| sdague | ewindisch: not if jenkins always installs it | 14:58 |
| sandywalsh | ewindisch, I guess the question becomes, when we add code are we responsible for ensuring every optional component still works | 14:58 |
| ewindisch | sdague: I meant, people won't run some tests locally, then they'll hit the wall on jenkins, reevaluate, and run the wider local tests? | 14:59 |
| sandywalsh | ewindisch, for example, I add something to xen, do I need to ensure it works under libvirt even though I don't use it. | 14:59 |
| sdague | ewindisch: yeh | 14:59 |
| sdague | sandywalsh: you need to ensure you don't break libvirt | 14:59 |
| ewindisch | sdague: which might not sound too horrible, but they'll have not had those libraries installed, so they'll now spend time setting up their environment, pissed off at those optional components | 14:59 |
| sandywalsh | sdague, bad example, I'm thinking adding functionality, not changing it. :) | 15:00 |
| sdague | sandywalsh: right, but that's why pg is in the requires list now | 15:00 |
| sdague | I spent 2 months chasing pg regressions :) | 15:00 |
| sdague | now it's in ci, so you can't slide code in that breaks what we've got tested | 15:00 |
| *** mrodden has joined #openstack-dev | 15:01 | |
| *** romcheg has joined #openstack-dev | 15:01 | |
| sandywalsh | sdague, I can see the argument, but I can also see the logical conclusion of all this. | 15:01 |
| *** diogogmt has joined #openstack-dev | 15:01 | |
| danpb | sandywalsh: yes if you add code whether a feature or bug, you are responsible for not breaking other drivers | 15:01 |
| *** cloudchimp has joined #openstack-dev | 15:02 | |
| danpb | i'm not saying you need to implement the feature for all drivers, just make sure you don't cause breakage in any existing work | 15:02 |
| sandywalsh | danpb, I guess it brings up the question of more formal interfaces | 15:02 |
| sdague | sandywalsh: I'd personally be ok with test-optional (given that at least in the db case we are doing skips if the db isn't setup right), and then catch it in check | 15:02 |
| sdague | and force devs to pull test-optional then to make it past jenkins check | 15:03 |
| *** monst_ has quit IRC | 15:03 | |
| sdague | given that about 80% of them probably won't ever need to, but I can also see the argument that everyone should be on the same dev env | 15:03 |
| *** romcheg1 has quit IRC | 15:03 | |
| ewindisch | and if you modify those other drivers attempting to implement it, it should either work, or not break it. Tests should run and catch it. Russell made good *looking* changes to the zeromq driver, but didn't actually run the tests, and the code was broken for a week. Then, for 2 months, the new feature still didn't even *work* until I wrote new tests that caught it. | 15:03 |
| sandywalsh | danpb, I'm cool with that. A stub that says "not implemented" ... but otherwise we'll get to the point we have to be experts in every unique snowflake in the system. | 15:04 |
| ewindisch | that was before the tests were gating in Jenkins, of course. | 15:04 |
| danpb | sdague: IMHO as long as the stuff the dev env is depending on is all opensource, then it isn't unreasonably onerous to have devs install the complete set of deps | 15:04 |
| *** jimfehlig has joined #openstack-dev | 15:04 | |
| sdague | danpb: yep, that's fair. it's status quo | 15:05 |
| sandywalsh | I'd much rather see test jigs on the core interfaces and the drivers have to be compatible with that ... but nova core shouldn't need to worry about the details. | 15:05 |
| ewindisch | and the fact the projects have broken up means that the dependencies are slightly better handled. | 15:05 |
| danpb | sdague: since it is really just a few more KB download / few more seconds build time when creating the .venv - which is counterbalanced by better quality for ethe community as a whole | 15:05 |
| sdague | sandywalsh: I think we'd all like to be there eventually, it's just not clear that we're really ready to nail down those interfaces | 15:05 |
| ewindisch | you only need postgres when testing Nova, or zeromq when testing oslo-incubator. | 15:05 |
| sandywalsh | *until it affects a core interface ... then it's "feats of strength" | 15:05 |
| danpb | sandywalsh: raise NotImplementedError is basically what is already done with compute drivers | 15:06 |
| *** romcheg has quit IRC | 15:06 | |
| *** mikal has quit IRC | 15:06 | |
| *** ron-slc_ has joined #openstack-dev | 15:07 | |
| sandywalsh | danpb, theoretically yes, but should any driver be able to make it into core or should it be an optional install. Just like an incubated project | 15:08 |
| *** sthaha has quit IRC | 15:08 | |
| *** dolphm has quit IRC | 15:08 | |
| *** mikal has joined #openstack-dev | 15:08 | |
| sdague | sandywalsh: we are a long way from that, look at how much change has happened just in this cycle | 15:08 |
| sdague | I'm all for getting us there | 15:08 |
| sandywalsh | sdague, good summit discussion I think. | 15:09 |
| sdague | sandywalsh: absolutely | 15:09 |
| *** navid has left #openstack-dev | 15:09 | |
| danpb | sandywalsh: well for virt drivers, we've basically got all the interesting ones now anyway :-) | 15:09 |
| sdague | especially once you freeze an API, you also need to be able to version it. | 15:09 |
| danpb | i think we're a very long way from being able to freeze the internal driver APIs | 15:10 |
| *** digitalsanctum has joined #openstack-dev | 15:10 | |
| sandywalsh | danpb, true ... and the reality would be that I would be affected by that change in the negative (we're Xen and KVM is the thousand pound gorilla), so we'd be an optional component. | 15:10 |
| sandywalsh | but, if we're going through the effort on Oslo to share code, this is the next logical step. | 15:11 |
| danpb | sandywalsh: in libvirt we've basically accepted support for any hypervisor driver where someone has been prepared to serve as the maintainer | 15:11 |
| sandywalsh | danpb, until they don't | 15:12 |
| danpb | sandywalsh: libvirt devs doing internal refactoring are required to make sure every driver in tree continues to build | 15:12 |
| sdague | danpb: I agree, we're a long way on the virt api. | 15:12 |
| danpb | over the long term it has worked out fine and we've not had to consider freezing our internal APIs at all | 15:12 |
| sandywalsh | danpb, interesting ... certainly the easiest solution for the maintainers. Not so much for the developers. I need to stew on that a little. | 15:13 |
| sandywalsh | *general developers | 15:13 |
| danpb | we prioritize the good of the project as a whole over an easy life for individual developrs | 15:14 |
| *** terryh has joined #openstack-dev | 15:14 | |
| sandywalsh | yep, I can see the rationale. It's a interesting set of tradeoffs. | 15:14 |
| sandywalsh | the thing that sparked it for me was seeing I had to install postgres client and thinking "great, but cruft on my dev laptop (since it's outside of the venv)" | 15:15 |
| sandywalsh | and realizing it's only going to get worse | 15:15 |
| sandywalsh | s/but/more/ | 15:16 |
| sandywalsh | haha, funny slip | 15:16 |
| sdague | :) | 15:16 |
| ewindisch | There is something psychologically broken about test-optional. Instead of making everyone install all the dependencies, it becomes something that one needs to do to fix code they've already relegated as "optional" or unnecessary. I don't know if there is a term for it, but it will likely harbor resentment. | 15:16 |
| ewindisch | instead, people like Sandy can just bemoan the same pains as everyone else. | 15:17 |
| *** wwallnrr__ has joined #openstack-dev | 15:18 | |
| *** sacharya has joined #openstack-dev | 15:18 | |
| sdague | ewindisch: yeh. Being the brunt of the complaints for having introduced real db dependencies, I'm just sensitive to people's pain | 15:18 |
| sdague | even if most of them come from people developing on macs... :) | 15:18 |
| *** rpedde_away is now known as rpedde | 15:20 | |
| sandywalsh | ewindisch, good point. It's a question of letting the interfaces grow organically or making them become a gating/community vote thing early on. | 15:20 |
| sandywalsh | and, where process kills, organic wins | 15:20 |
| ewindisch | sdague: I think you and I are in the same boat. I'm surprised there haven't been more complaints about zeromq being required to test in oslo, but I think that can be attributed to few people actually doing work there. Also, it helped that I did warn the mailing list. | 15:21 |
| sdague | sandywalsh: and realize, every time we've added those external dependencies it was because real bugs were found, and they were part of keeping them from regressing | 15:21 |
| ewindisch | sandy: just wait until I add gpgme... | 15:22 |
| sandywalsh | heh ... I'd even been happy with a distro package of the non-python dev requirements. Something I can just freshen up, rather than having to chase down pip failures | 15:23 |
| *** lucasagomes has joined #openstack-dev | 15:23 | |
| sdague | sandywalsh: yeh, that's a good idea | 15:23 |
| sandywalsh | sdague, might be the first step? | 15:24 |
| *** wwallnrr__ has quit IRC | 15:24 | |
| sdague | sandywalsh: yeh, sure, making install_venv a little smarter | 15:24 |
| *** topol has joined #openstack-dev | 15:25 | |
| *** wwallnrr__ has joined #openstack-dev | 15:25 | |
| *** davidha has quit IRC | 15:26 | |
| ewindisch | btw, I have a draft patch for recursive dependency resolution of oslo requirements, if anyone wants a gander. | 15:26 |
| sdague | sandywalsh: design / hack session on that would be reasonable in portland | 15:27 |
| *** boris-42 has quit IRC | 15:27 | |
| *** cp16net is now known as cp16net|away | 15:27 | |
| ewindisch | another thing is we could do better to link dependencies to the code that actually depends on them. | 15:27 |
| *** sacharya has quit IRC | 15:27 | |
| BLZbubba | argh, is there no longer a way to specify the nova-api ports on the command line? | 15:27 |
| sandywalsh | I was working on a tool called Signature, for checking method signatures between drivers/plugin modules https://github.com/sandywalsh/signature | 15:28 |
| sandywalsh | might be a handy set of tests too | 15:29 |
| *** monst_ has joined #openstack-dev | 15:29 | |
| ayoung | henrynash, http://logs.openstack.org/20289/20/check/gate-keystone-python26/2693/console.html.gz | 15:29 |
| BLZbubba | if not, what is the more up-to-date way to do something like this: http://www.cybera.ca/tech-radar/putting-openstack-production-part-3-load-balancing-and-https | 15:29 |
| henrynash | ayoung: looking at it | 15:31 |
| *** sc68cal has joined #openstack-dev | 15:31 | |
| ayoung | henrynash, same problem...policy is getting messed up. My suggestion is to use the generated link on the page to pull the code as a branch. | 15:32 |
| henrynash | ayoung: on your latest patch, you mean? | 15:32 |
| ayoung | henrynash, yeah | 15:32 |
| ayoung | https://review.openstack.org/#/c/20289/ | 15:33 |
| henrynash | ayoung: will do | 15:33 |
| *** danwent has quit IRC | 15:33 | |
| ayoung | if you look at the section just above the diff, where it is labled Download, there are a range of options. THe oe I found most useful is 'checkout' | 15:33 |
| *** yidclare has quit IRC | 15:36 | |
| *** esheffield has joined #openstack-dev | 15:36 | |
| *** maoy has quit IRC | 15:36 | |
| *** mindpixel has joined #openstack-dev | 15:37 | |
| BLZbubba | ugh do i really have to have a one line config file for each nova-api process instead of being able to specify --osapi_listen_port ? | 15:37 |
| BLZbubba | i hope you added those port options back in folsom | 15:37 |
| *** arbrandes has quit IRC | 15:39 | |
| BLZbubba | or else have it fork off each request | 15:39 |
| *** danwent has joined #openstack-dev | 15:39 | |
| *** danwent has quit IRC | 15:41 | |
| mordred | markmc: got a sec to join in #openstack-infra? | 15:41 |
| *** monst_ has quit IRC | 15:41 | |
| *** mmagr has quit IRC | 15:43 | |
| *** zoresvit has quit IRC | 15:44 | |
| henrynash | ayoung: ok, have it re-produced inside a single test file (you v3_trusts)…by including a fake test class ahead of yours that rewrite the policy file…so will debug now | 15:45 |
| *** alszar has joined #openstack-dev | 15:46 | |
| *** markwash has quit IRC | 15:46 | |
| *** alunduil has joined #openstack-dev | 15:47 | |
| *** dolphm has joined #openstack-dev | 15:47 | |
| *** psedlak has quit IRC | 15:48 | |
| *** alunduil has quit IRC | 15:48 | |
| *** rnirmal has joined #openstack-dev | 15:49 | |
| *** alunduil has joined #openstack-dev | 15:49 | |
| *** radez is now known as radez_g0n3 | 15:51 | |
| *** soody has quit IRC | 15:51 | |
| *** adjohn has joined #openstack-dev | 15:52 | |
| *** soody has joined #openstack-dev | 15:53 | |
| *** gyee has joined #openstack-dev | 15:53 | |
| *** radez_g0n3 is now known as radez | 15:54 | |
| *** ecarlin has joined #openstack-dev | 15:55 | |
| *** adjohn has quit IRC | 15:57 | |
| *** jgallard has quit IRC | 15:57 | |
| *** datsun180b has joined #openstack-dev | 15:57 | |
| *** diogogmt has quit IRC | 15:59 | |
| *** sacharya has joined #openstack-dev | 15:59 | |
| *** jaypipes has quit IRC | 15:59 | |
| henrynash | ayoung: ok, got the fix. | 15:59 |
| *** ecarlin has quit IRC | 16:00 | |
| henrynash | ayoung: it's because self.opt(policy_file=self.tmpfilename) changes the (effective_ CONF.policy file name | 16:00 |
| *** diogogmt has joined #openstack-dev | 16:00 | |
| *** monst_ has joined #openstack-dev | 16:01 | |
| henrynash | ayoung: so teardown has to put it back. Will post a fix shortly | 16:01 |
| *** ecarlin has joined #openstack-dev | 16:01 | |
| *** rnirmal has quit IRC | 16:01 | |
| *** zoresvit has joined #openstack-dev | 16:02 | |
| *** gary_th has joined #openstack-dev | 16:03 | |
| *** cp16net|away is now known as cp16net | 16:04 | |
| ayoung | henrynash, so, can you do it in setUp as well. | 16:04 |
| ayoung | Probably best for each test to be paranoid | 16:04 |
| ayoung | and ot assume anything about the state of the system coming in | 16:04 |
| *** jgallard has joined #openstack-dev | 16:04 | |
| ayoung | and 'not assume' | 16:04 |
| henrynash | just seeing what self.opt() does… | 16:05 |
| *** garyTh has quit IRC | 16:05 | |
| *** gongysh has quit IRC | 16:05 | |
| *** monst_ has quit IRC | 16:05 | |
| henrynash | ayoung: since CONF.Policyfile jas been changed by it…so unless you reload the the conf settings (or save it off and put it back in your teardown, you won;t have it right) | 16:06 |
| ayoung | henrynash, we should reload it | 16:06 |
| henrynash | ayoung: yes, I think so | 16:06 |
| ayoung | henrynash, In setup we specify the conf file to lead | 16:07 |
| ayoung | load | 16:07 |
| *** adalbas has quit IRC | 16:07 | |
| henrynash | ayoung: fyi, here's how I debugged it…added this fake test class ahead of yours in test_v3_trusts…as it is it now works, take out the reseting of policy_file in teardown and it fails. http://fpaste.org/yAAX/ | 16:09 |
| *** zbitter is now known as zaneb | 16:09 | |
| *** CaptTofu has quit IRC | 16:09 | |
| *** CaptTofu has joined #openstack-dev | 16:10 | |
| *** CaptTofu has quit IRC | 16:10 | |
| *** CaptTofu has joined #openstack-dev | 16:11 | |
| *** CaptTofu has quit IRC | 16:11 | |
| *** CaptTofu has joined #openstack-dev | 16:11 | |
| *** CaptTofu has quit IRC | 16:11 | |
| *** CaptTofu has joined #openstack-dev | 16:12 | |
| *** monst_ has joined #openstack-dev | 16:15 | |
| *** esp1 has joined #openstack-dev | 16:15 | |
| *** monst_ has quit IRC | 16:15 | |
| *** jrodom has joined #openstack-dev | 16:15 | |
| topol | YorkikSar, are you around? | 16:16 |
| henrynash | ayoung: so maybe I'll post a patch that does the correct cleanup (since not good to leave things in a bad way)…do you want add the loading part? | 16:16 |
| henrynash | (in setup) | 16:16 |
| *** monst_ has joined #openstack-dev | 16:16 | |
| *** nunosantos has joined #openstack-dev | 16:16 | |
| *** radez is now known as radez_g0n3 | 16:16 | |
| topol | YorikSar, sorry for the wrong spelling. You there? | 16:16 |
| BLZbubba | anyone here know anything about nova-api? | 16:17 |
| *** ecarlin has quit IRC | 16:17 | |
| *** pabelanger has joined #openstack-dev | 16:17 | |
| *** jbr_ has quit IRC | 16:18 | |
| *** radez_g0n3 is now known as radez | 16:19 | |
| *** jaypipes has joined #openstack-dev | 16:19 | |
| BLZbubba | ugh is nova-api configurable at all? it is ignoring the port options in nova.conf and the undocumented flagfile option seems useless too | 16:20 |
| *** flaper87 has joined #openstack-dev | 16:21 | |
| *** thingee_zzz is now known as thingee | 16:21 | |
| *** thingee has quit IRC | 16:22 | |
| *** adjohn has joined #openstack-dev | 16:23 | |
| *** maroh has quit IRC | 16:23 | |
| *** aloga has quit IRC | 16:23 | |
| *** adalbas has joined #openstack-dev | 16:23 | |
| *** yamahata has quit IRC | 16:24 | |
| *** jcmartin has quit IRC | 16:26 | |
| *** doude has left #openstack-dev | 16:26 | |
| ayoung | henrynash, so my bad for that failing. I removed the CONF.reset() prior to pushing, and of course, that was also required | 16:26 |
| henrynash | ayoung: np…double whammy :-) | 16:27 |
| *** adjohn has quit IRC | 16:27 | |
| henrynash | ayoung: so about to post a patch which fixes the tear downs of test_policy and test_v3_protection | 16:27 |
| *** alunduil has quit IRC | 16:28 | |
| *** devoid has joined #openstack-dev | 16:29 | |
| *** thingee has joined #openstack-dev | 16:29 | |
| *** ecarlin has joined #openstack-dev | 16:30 | |
| *** ecarlin has quit IRC | 16:30 | |
| *** dansmith is now known as SteelyDan | 16:30 | |
| *** koolhead17 has quit IRC | 16:30 | |
| *** imsplitbit has joined #openstack-dev | 16:32 | |
| *** esp1 has quit IRC | 16:32 | |
| henrynash | ayoung: https://review.openstack.org/#/c/22705/ | 16:34 |
| ayoung | henrynash, the change you make in v3 protection should be in the base class. I think it should be moved to test_v3 | 16:35 |
| *** networkstatic has joined #openstack-dev | 16:35 | |
| ayoung | henrynash, but, yes, that looks about right | 16:36 |
| henrynash | …but not all v3 child test classes re-write policy files…but I guess it wouldn't harm | 16:36 |
| *** monst_ has quit IRC | 16:36 | |
| ayoung | henrynash, one problem is that you can't tell just from that if it is working....I've incorporated the changed into my own patch for now and can drive on. | 16:36 |
| henrynash | ayoung: yes, agreed | 16:37 |
| ayoung | henrynash, so, yeah, each test should clean up after itself, but each test class should have a certain blank slate to work with. Its like the live DB tests. If you haven't cleaned up the db before run the test, they fail | 16:38 |
| *** adjohn has joined #openstack-dev | 16:38 | |
| ayoung | and that is a pain | 16:38 |
| *** Tross has quit IRC | 16:39 | |
| *** anniec has quit IRC | 16:41 | |
| gyee | ayoung, trust API goes to RC1 then? | 16:42 |
| *** thouveng has quit IRC | 16:42 | |
| gyee | there was an announcement about G3 availability | 16:42 |
| ayoung | gyee, yeah | 16:44 |
| *** CaptTofu has quit IRC | 16:44 | |
| *** opilotte has quit IRC | 16:45 | |
| ayoung | gyee, s'alright. Nothing depends on it yet. I just want it there so people can build on it in the future. | 16:45 |
| *** pnavarro has quit IRC | 16:46 | |
| gyee | ayoung, k, I was confuse about the announcement and FFE | 16:46 |
| ayoung | gyee, are you guys waiting on trusts? | 16:47 |
| *** opilotte has joined #openstack-dev | 16:48 | |
| gyee | ayoung, yeah, we have similar use case as HEAT | 16:49 |
| ayoung | gyee, we'll have it soon. BTW, Can you take a look at the patch, specifically at the token/controller.py changes, and figure out how that should go into v3 auth? | 16:50 |
| *** jbryce has joined #openstack-dev | 16:53 | |
| *** soody has quit IRC | 16:53 | |
| gyee | ayoung, sure | 16:54 |
| *** soody has joined #openstack-dev | 16:55 | |
| *** rnirmal has joined #openstack-dev | 16:56 | |
| *** edmund has joined #openstack-dev | 16:57 | |
| sandywalsh | still getting test failures around test_xml_and_uri_rescue_* ... anyone run into this before? | 16:58 |
| sandywalsh | (nova trunk) | 16:59 |
| *** nati_ueno has joined #openstack-dev | 16:59 | |
| *** nati_ueno has quit IRC | 17:02 | |
| *** monst_ has joined #openstack-dev | 17:03 | |
| *** monst_ has joined #openstack-dev | 17:04 | |
| *** Gordonz has joined #openstack-dev | 17:04 | |
| *** imsplitbit has quit IRC | 17:05 | |
| *** henrynash has quit IRC | 17:05 | |
| *** danwent has joined #openstack-dev | 17:05 | |
| *** nsatterl_ has quit IRC | 17:06 | |
| ewindisch | FFE review ping: https://review.openstack.org/#/c/20434/ | 17:06 |
| *** martine has quit IRC | 17:08 | |
| *** soody has quit IRC | 17:09 | |
| *** CaptTofu has joined #openstack-dev | 17:09 | |
| *** mrodden has quit IRC | 17:10 | |
| *** imsplitbit has joined #openstack-dev | 17:10 | |
| *** dolphm has quit IRC | 17:11 | |
| jdurgin | dachary: danpb created most of that page after some discussion on the list (http://lists.openstack.org/pipermail/openstack-dev/2013-January/thread.html#4842) | 17:11 |
| vishy | garyk: I need to chat more with you about the dhcp issue, but I'm on a bad connection right now. Will you be around in about an hour? | 17:11 |
| openstackgerrit | A change was merged to openstack/python-openstackclient: Correct the version mapping to image service. https://review.openstack.org/22227 | 17:12 |
| *** rafaduran has left #openstack-dev | 17:12 | |
| *** hemna has quit IRC | 17:14 | |
| *** spzala has joined #openstack-dev | 17:15 | |
| *** dolphm has joined #openstack-dev | 17:15 | |
| *** soody has joined #openstack-dev | 17:15 | |
| *** jgallard has quit IRC | 17:16 | |
| *** pnavarro has joined #openstack-dev | 17:17 | |
| *** jpich has quit IRC | 17:17 | |
| *** troytoman-away is now known as troytoman | 17:17 | |
| *** dhellmann has joined #openstack-dev | 17:18 | |
| *** pnavarro has quit IRC | 17:18 | |
| *** esp1 has joined #openstack-dev | 17:18 | |
| danpb | jdurgin think i'm missing some conversation here... most of what page ? | 17:19 |
| *** Gordonz has quit IRC | 17:20 | |
| *** zaitcev has joined #openstack-dev | 17:21 | |
| *** gael has quit IRC | 17:22 | |
| *** flaper87 has quit IRC | 17:23 | |
| arosen | garyk: yup here | 17:24 |
| *** cp16net is now known as cp16net|away | 17:25 | |
| *** mrodden has joined #openstack-dev | 17:26 | |
| *** amerine has joined #openstack-dev | 17:26 | |
| gyee | ayoung, endpoint scoping is not supported yet right? | 17:27 |
| *** cp16net|away is now known as cp16net | 17:27 | |
| *** Gordonz has joined #openstack-dev | 17:28 | |
| *** jruzicka has quit IRC | 17:28 | |
| *** lglenden has joined #openstack-dev | 17:28 | |
| *** amerine has quit IRC | 17:28 | |
| *** mindpixel has quit IRC | 17:29 | |
| *** TheEstablishment has joined #openstack-dev | 17:29 | |
| *** amerine has joined #openstack-dev | 17:29 | |
| nikhil | a doc question for nova: if i write an extension and there is changes in the /servers and /servers/detail response, does that sample response need to exist in doc/api_samples/<extension_name>/sample response | 17:29 |
| *** roampune has joined #openstack-dev | 17:31 | |
| *** henrynash has joined #openstack-dev | 17:31 | |
| *** ndipanov has quit IRC | 17:33 | |
| dachary | danpb: jdurgin answered my question about the origins of https://wiki.openstack.org/wiki/BlockDeviceConfig :-) I updated https://blueprints.launchpad.net/nova/+spec/flexible-block-device-config with a pointer to http://lists.openstack.org/pipermail/openstack-dev/2013-January/thread.html#4842 | 17:33 |
| *** jrodom has quit IRC | 17:33 | |
| *** almaisan-away is now known as al-maisan | 17:33 | |
| *** unix has quit IRC | 17:34 | |
| *** pasquier-s has quit IRC | 17:34 | |
| ayoung | gyee, that is correct. THe trust supports it, but the tokens do not. | 17:35 |
| ayoung | gyee, does v3 auth support endpoint scoping> | 17:35 |
| gyee | ayoung, not right now | 17:35 |
| danpb | dachary: ah, i just updated the wiki page to link back to the blueprint to avoid future confusion | 17:35 |
| gyee | just want to know what's the expectation | 17:35 |
| *** arbrandes has joined #openstack-dev | 17:36 | |
| *** markmc has quit IRC | 17:37 | |
| nikhil | alaski: ^^ | 17:38 |
| *** TheEstablishment has quit IRC | 17:39 | |
| ayoung | gyee, when it comes down to it, it really is up to policy or auth_token middleware to support endpoint scoping. | 17:39 |
| *** dachary has quit IRC | 17:39 | |
| *** jrodom has joined #openstack-dev | 17:39 | |
| gyee | ayoung, what I mean is when we issue a trust token, do we filter the endpoints returned? | 17:40 |
| *** arbrandes has quit IRC | 17:40 | |
| ayoung | gyee, ah, no | 17:41 |
| gyee | why not? | 17:41 |
| *** Mandell has joined #openstack-dev | 17:41 | |
| ayoung | gyee, but...we can | 17:41 |
| gyee | we should right? | 17:41 |
| ayoung | gyee, really? because it wasn't enforced yet. Lazyiness, really | 17:41 |
| *** njoy has joined #openstack-dev | 17:41 | |
| ayoung | yeah, I'll add that in. | 17:41 |
| gyee | if we filter them, that would make them effective | 17:42 |
| *** john5223 has joined #openstack-dev | 17:42 | |
| *** AnilV4 has quit IRC | 17:43 | |
| *** njoy_ has joined #openstack-dev | 17:44 | |
| ayoung | gyee, we don't filter the service catalog based on user yet, so I can't confirm that the trustor has the endpoint in their catalog | 17:44 |
| gyee | ayoung, I have a blueprint to establish the relationship between project and endpoints | 17:45 |
| ayoung | gyee, good stuff, but I don't think that will make it into rc1 | 17:45 |
| gyee | couple that with trust, that would make endpoint scoping effective | 17:45 |
| zykes- | win 40 | 17:45 |
| ayoung | gyee, still needs enforcement | 17:45 |
| ayoung | otherwise, token gets passed to service and it gets through regardless of what is in it | 17:46 |
| gyee | enforcement is in the client | 17:46 |
| ayoung | enforcement needs to be in auth_token middleware. Or in policy | 17:46 |
| gyee | no endpoint in the catalog, user have no access | 17:46 |
| zykes- | ayoung: will trust stuff get into g ? | 17:46 |
| ayoung | zykes-, yes | 17:47 |
| *** njoy has quit IRC | 17:47 | |
| ayoung | gyee, so I need to wrap trust.create's input document in a tope level trust value? | 17:47 |
| ayoung | becaus of XML? | 17:47 |
| gyee | ayoung, sadly yes | 17:48 |
| ayoung | gyee, OK. I think that is fine. dumb, but fine | 17:48 |
| gyee | with XML, only one root element is allowed | 17:48 |
| *** Ryan_Lane has quit IRC | 17:48 | |
| zykes- | ayoung: are there services using it as well ? | 17:48 |
| ayoung | zykes-, there will be soon, but not yet | 17:48 |
| *** Ryan_Lane has joined #openstack-dev | 17:48 | |
| *** mrodden has quit IRC | 17:49 | |
| *** FatDarrel has joined #openstack-dev | 17:49 | |
| gyee | ayoung, I was kind of joking about endpoint enforcement in the client | 17:50 |
| *** devananda has quit IRC | 17:50 | |
| gyee | security 101, never trust the client :) | 17:50 |
| *** danpb has quit IRC | 17:50 | |
| ayoung | gyee, I need to cut scope somewhere. No endpoint enforcement in G | 17:50 |
| *** devananda has joined #openstack-dev | 17:51 | |
| gyee | ayoung, that's fine, lets do this in H | 17:51 |
| *** trapnii has quit IRC | 17:54 | |
| *** matelakat has quit IRC | 17:54 | |
| *** CaptTofu has quit IRC | 17:56 | |
| ayoung | gyee, actually, I thine the API doc was wrong. I thin I already had the top level trust groupoing in there. The controller's require it | 17:56 |
| *** CaptTofu has joined #openstack-dev | 17:56 | |
| ayoung | https://review.openstack.org/#/c/22715/1/tests/test_v3_trust.py | 17:56 |
| ayoung | line 192 | 17:56 |
| *** edmund has quit IRC | 17:57 | |
| *** tomoe_ has quit IRC | 17:59 | |
| *** davidh_ has joined #openstack-dev | 17:59 | |
| *** russellb is now known as rustlebee | 18:00 | |
| *** CaptTofu has quit IRC | 18:01 | |
| gyee | ayoung, k, I'll wait for your next patch then | 18:01 |
| *** lucasagomes has quit IRC | 18:02 | |
| *** yidclare has joined #openstack-dev | 18:02 | |
| ayoung | gyee, please don't wait | 18:02 |
| *** derekh has quit IRC | 18:02 | |
| *** Tross has joined #openstack-dev | 18:03 | |
| ayoung | I'm working on the endpoint and roles for dolphm 's suggestions, which is that they get returned as objects | 18:03 |
| zykes- | ayoung: as in G or ? | 18:03 |
| ayoung | but the API doc is wrong, and I'm fixing that as well, to match {trust:{.... | 18:03 |
| alaski | nikhil: yes. An extensions samples should be in doc/api_samples/<extension_name>/<samples>, which in your case would be a superset of the servers response. And the docs should be generated by templates in nova/tests/integrated/api_samples/<extension>/ | 18:03 |
| gyee | ayoung, sure, I'll take a look at your code | 18:03 |
| ayoung | gyee, just give me a plan for the v3 auth and tokens | 18:03 |
| *** alunduil has joined #openstack-dev | 18:04 | |
| ayoung | gyee, ah, you meant my next patch on API...yes, I'll ping you when that is done, too. | 18:04 |
| ayoung | thanks | 18:04 |
| gyee | ayoung, we'll fix all the bugs in rc1 | 18:04 |
| gyee | for the token inconsistent formats | 18:04 |
| sdague | rustlebee: as markmc isn't around, what's the oslo stance on freeze? are we allowed to start factoring in modules that won't migrate out to projects until havana? | 18:04 |
| openstackgerrit | A change was merged to openstack-dev/devstack: Added cinder-backup service (not enabled by default) https://review.openstack.org/22533 | 18:04 |
| openstackgerrit | A change was merged to openstack-dev/devstack: set dhcp_agent_manager for ovs plugin https://review.openstack.org/22693 | 18:04 |
| *** melwitt has joined #openstack-dev | 18:05 | |
| rustlebee | sdague: same policy, oslo-incubator will have a stable/grizzly branch, so shouldn't be merging features without an exception right now | 18:06 |
| *** mrodden has joined #openstack-dev | 18:06 | |
| sdague | rustlebee: ok | 18:06 |
| *** CaptTofu has joined #openstack-dev | 18:06 | |
| *** markwash has joined #openstack-dev | 18:07 | |
| *** dontalton has joined #openstack-dev | 18:08 | |
| *** zing has quit IRC | 18:09 | |
| *** darraghb has quit IRC | 18:14 | |
| *** olaph has quit IRC | 18:14 | |
| *** dachary has joined #openstack-dev | 18:15 | |
| vishy | jgriffith: ping | 18:16 |
| *** dprince has quit IRC | 18:16 | |
| jgriffith | vishy: pong | 18:18 |
| *** zoresvit has quit IRC | 18:18 | |
| ayoung | dolphm, what attributes do we want in roles and endpoint for the trust? Is it OK if I go with a minimal set at first (role id,nam) endpoint(id) ? | 18:19 |
| ayoung | we can do links etc later? | 18:19 |
| dolphm | ayoung: minimal is ideal | 18:19 |
| ayoung | dolphm, OK, on list, I am thinking of removing most values | 18:20 |
| ayoung | I'll leave trustro and trustee and trust id | 18:20 |
| dolphm | ayoung: wait, on GET /trusts? | 18:20 |
| vishy | jgriffith: so we have an odd test run where c-vol failed to start and doesn't log at all | 18:20 |
| vishy | i was wondering if you have ever seen that | 18:20 |
| *** olaph has joined #openstack-dev | 18:20 | |
| ayoung | dolphm, yeah. You get all the values on GET trusts/(id) | 18:20 |
| jgriffith | vishy: yes | 18:20 |
| ayoung | list does one DB lookup | 18:20 |
| dolphm | ayoung: i'm asking about GET /trusts (the collection) | 18:20 |
| *** trapnii has joined #openstack-dev | 18:20 | |
| jgriffith | vishy: It's not the same one we looked at last week? | 18:20 |
| ayoung | yeah, | 18:20 |
| jgriffith | vishy: the one we thought was screen issue? | 18:20 |
| ayoung | dolphm, on GET trusts I only want to do one DB lookup | 18:21 |
| dolphm | ayoung: all the other collections return full resources | 18:21 |
| ayoung | dolphm, OK...I'll kep it | 18:21 |
| *** andrewbogott_afk is now known as andrewbogott | 18:21 | |
| ayoung | keep | 18:21 |
| dolphm | ayoung: wait, i see what you're saying | 18:21 |
| *** devananda has quit IRC | 18:22 | |
| dolphm | ayoung: what attributes do you want to include? trustor_user_id and trustee_user_id? | 18:22 |
| ayoung | dolphm, yeah, and the trust id | 18:22 |
| *** andrewbogott is now known as andrewbogott_afk | 18:22 | |
| *** andrewbogott_afk is now known as andrewbogott | 18:22 | |
| dolphm | ayoung: how about the impersonation flag? | 18:22 |
| *** andrewbogott is now known as andrewbogott_afk | 18:22 | |
| ayoung | dolphm, not on lit, I think | 18:22 |
| ayoung | list | 18:23 |
| ayoung | dolphm, but if you insist, sure | 18:23 |
| *** andrewbogott_afk is now known as andrewbogott | 18:23 | |
| dolphm | ayoung: hrm, so ... how about splitting up the API a bit ... | 18:23 |
| *** alunduil has quit IRC | 18:23 | |
| dolphm | ayoung: don't know why i didn't htink of this before, but do: | 18:23 |
| dolphm | ayoung: /trusts/{trust_id}/roles and /trusts/{trust_id}/endpoints | 18:23 |
| *** devananda has joined #openstack-dev | 18:23 | |
| dolphm | ayoung: err, but you want immutable... | 18:24 |
| ayoung | dolphm, that works immutable | 18:24 |
| *** sandywalsh_ has joined #openstack-dev | 18:24 | |
| *** sandywalsh has quit IRC | 18:24 | |
| ayoung | you add them all on create. THat api would be just query, but I don;t think it buys muh | 18:24 |
| ayoung | muh | 18:24 |
| ayoung | much | 18:24 |
| ayoung | dolphm, so limiting values on list is OK? | 18:25 |
| dolphm | ayoung: you'll have to implement GET /trusts/{trust_id}/roles (return the collection of full role resources), HEAD /trusts/{trust_id}/roles/{role_id} (check if role is included in trust), and GET /trusts/{trust_id}/roles/{role_id} | 18:25 |
| dolphm | ayoung: /trusts/{trust_id}/projects too | 18:25 |
| dolphm | ayoung: don't even include them on the list | 18:25 |
| ayoung | dolphm, no need for those up front, but they would be OK as RFEs | 18:25 |
| *** trapniii has joined #openstack-dev | 18:25 | |
| jgriffith | vishy: If it's not the same one then the only thing I can think of is to doulbe check the changes we made a few days ago | 18:25 |
| ayoung | dolphm, how about | 18:25 |
| ayoung | each list gives you trustor, truste and a link to the GET | 18:26 |
| dolphm | ayoung: you can keep POST /trusts basically the same, but then make it so that to get those collections back, you have to ask for them specifically | 18:26 |
| ayoung | then follow the link if you want full details | 18:26 |
| jgriffith | vishy: service --> launcher | 18:26 |
| sandywalsh_ | anyone else getting trunk libvirt build failures? http://paste.openstack.org/show/32353/ | 18:26 |
| dolphm | ayoung: trusts should include a "links" object that -- let me make a paste | 18:26 |
| vishy | jgriffith: it is a random failure | 18:27 |
| ayoung | dolphm, yeah, the self link | 18:27 |
| vishy | jgriffith: i thought these kind of failures were due to screen_it using stuff but it is still happening | 18:27 |
| *** trapnii has quit IRC | 18:27 | |
| jgriffith | vishy: K, that's what I was wondering | 18:28 |
| jgriffith | vishy: recent gate logs? | 18:28 |
| *** maoy has joined #openstack-dev | 18:28 | |
| jgriffith | realizing there's nothing there :( | 18:28 |
| dolphm | ayoung: should be able to include stuff like this http://paste.openstack.org/raw/32357/ | 18:29 |
| *** PaulCarlton has quit IRC | 18:29 | |
| vishy | jgriffith: http://logs.openstack.org/22590/4/gate/gate-tempest-devstack-vm-quantum/9667/logs/ | 18:29 |
| *** CaptTofu has quit IRC | 18:29 | |
| ayoung | dolphm, how about just self for now? | 18:30 |
| dolphm | ayoung: sure, and you'll get "self" for free by using the v3 controllers with the wrap_collection / wrap_member class methods | 18:31 |
| ayoung | dolphm, schweet! | 18:31 |
| *** CaptTofu has joined #openstack-dev | 18:31 | |
| *** Ryan_Lane has quit IRC | 18:32 | |
| dolphm | ayoung: make sure you set collection_name = 'trusts' and member_name = 'trust' e.g. https://github.com/openstack/keystone/blob/master/keystone/identity/controllers.py#L652 | 18:32 |
| dolphm | ayoung: and then call TrustV3.wrap_member(context, ref) and TrustV3.wrap_collection(context, refs) as appropriate, e.g. https://github.com/openstack/keystone/blob/master/keystone/identity/controllers.py#L664 | 18:33 |
| *** trapniii has quit IRC | 18:34 | |
| *** nati_ueno has joined #openstack-dev | 18:35 | |
| openstackgerrit | A change was merged to openstack/nova: Migration 144: Fix drop index statement https://review.openstack.org/22628 | 18:35 |
| *** nati_ueno has quit IRC | 18:36 | |
| jgriffith | vishy: Working on a local setup to try something | 18:37 |
| jgriffith | vishy: I seem to recall hitting this once a while back when I enabled swift on my devstack locally | 18:38 |
| jgriffith | vishy: Maybe I can hit it again and get some more info | 18:38 |
| vishy | jgriffith: you think it could be a bad interaction with swift? | 18:38 |
| jgriffith | vishy: That's a super long shot... but it's interesting I've spun up devstack so many times lately to do reviews etc | 18:39 |
| jgriffith | never hit it | 18:39 |
| notmyname | jgriffith: vishy: is something I need to look in to? | 18:39 |
| ayoung | dolphm, adding calls to wrap_member have astrange side effect: Conflict: Conflict occurred attempting to store tenant. Duplicate ID, bar. | 18:39 |
| vishy | jgriffith: infra is doing a search to see if the failure happens to other workers | 18:39 |
| jgriffith | notmyname: don't think so, just trying to figure out how to repro something | 18:39 |
| vishy | notmyname: not yet, we are grasping at straws :) | 18:40 |
| *** gary_th has quit IRC | 18:40 | |
| jgriffith | vishy: cool | 18:40 |
| ayoung | the might befrom wrap_collection actually | 18:40 |
| notmyname | ok, let me know if something comes up | 18:40 |
| jgriffith | vishy: The other idea is if they can setup a trap on it | 18:40 |
| jgriffith | vishy: Then I can log on and try to trace through the startup maybe | 18:40 |
| jgriffith | I'll ask them.. | 18:40 |
| ayoung | dolphm, maybe I misunderstood what it was doing, let me try something a little different | 18:41 |
| dolphm | ayoung: you should call it on the return | 18:41 |
| vishy | garyk: back yet? | 18:42 |
| *** CaptTofu has quit IRC | 18:43 | |
| *** CaptTofu has joined #openstack-dev | 18:43 | |
| openstackgerrit | A change was merged to openstack/nova: BM Migrations 2 & 3: Fix drop_column statements https://review.openstack.org/22630 | 18:44 |
| *** jcmartin has joined #openstack-dev | 18:44 | |
| dolphm | gyee: revised https://review.openstack.org/#/c/22696/ | 18:45 |
| openstackgerrit | A change was merged to openstack/nova: Fix migration snake-walk. https://review.openstack.org/22631 | 18:45 |
| *** dprince has joined #openstack-dev | 18:46 | |
| ayoung | dolphm, Nah, I got it. Code error. | 18:47 |
| openstackgerrit | A change was merged to openstack/nova: Add `post_downgrade` hook for migration tests. https://review.openstack.org/22633 | 18:48 |
| *** nati_ueno has joined #openstack-dev | 18:49 | |
| openstackgerrit | A change was merged to openstack/nova: Migration 146: Execute delete call. https://review.openstack.org/22634 | 18:50 |
| *** kagan has joined #openstack-dev | 18:50 | |
| *** dachary has quit IRC | 18:51 | |
| *** trapniii has joined #openstack-dev | 18:51 | |
| *** pixelbeat has quit IRC | 18:53 | |
| *** koolhead17 has joined #openstack-dev | 18:56 | |
| gyee | dolphm, https://review.openstack.org/#/c/22696/ looks good | 18:57 |
| *** cp16net is now known as cp16net|away | 18:57 | |
| gyee | now I need to adjust the code | 18:57 |
| dolphm | ayoung: your turn https://review.openstack.org/#/c/22696/ | 18:58 |
| dolphm | henrynash: ^ | 18:58 |
| *** alunduil has joined #openstack-dev | 18:59 | |
| dolphm | ayoung: henrynash: see the referenced bug | 18:59 |
| *** cp16net|away is now known as cp16net | 18:59 | |
| ayoung | dolphm, DOes the API doc change have a deadline? | 19:00 |
| *** soody has quit IRC | 19:00 | |
| henrynash | dolphm: looking at it | 19:00 |
| *** capnkooc has joined #openstack-dev | 19:01 | |
| *** troytoman is now known as troytoman-away | 19:02 | |
| *** beagles has quit IRC | 19:03 | |
| *** al-maisan is now known as almaisan-away | 19:04 | |
| *** beagles has joined #openstack-dev | 19:07 | |
| *** networkstatic has quit IRC | 19:07 | |
| henrynash | dolphm, young: I'm good with https://review.openstack.org/#/c/22696/ - but Adam, do you want to have a look before we push the buton | 19:08 |
| henrynash | dolphm, ayoung: I'm good with https://review.openstack.org/#/c/22696/ - but Adam, do you want to have a look before we push the buton | 19:08 |
| ayoung | henrynash, I'll take a look. | 19:09 |
| *** afazekas has joined #openstack-dev | 19:10 | |
| *** riskable has joined #openstack-dev | 19:10 | |
| ayoung | dolphm, is it correct to say tokesn do have ID attributes, or just that they are not embedded? | 19:10 |
| *** novas0x2a|laptop has quit IRC | 19:10 | |
| ayoung | actually, I think that statement is correct from an API perspective | 19:12 |
| ayoung | henrynash, dolphm pulled the trigger | 19:12 |
| *** colinmcnamara has joined #openstack-dev | 19:14 | |
| *** FatDarrel has quit IRC | 19:15 | |
| *** dhellmann is now known as dhellmann-afk | 19:15 | |
| gyee | ayoung, what's the branch for rc1? | 19:16 |
| ayoung | gyee, no idea... | 19:16 |
| topol | dolphm, i there still a fix me on line 1145 i22696. Is that OK? | 19:17 |
| *** soody has joined #openstack-dev | 19:17 | |
| gyee | henrynash, dolphm, any idea? how do I code against rc1? | 19:18 |
| topol | dolphm, err there was still a dolph fix me on line 1145 is what I was trying to say | 19:19 |
| *** morganfainberg has joined #openstack-dev | 19:19 | |
| mtreinish | bcwaldon: I just pushed the rebased migration tests | 19:20 |
| *** alunduil has quit IRC | 19:20 | |
| *** cloudchimp has quit IRC | 19:20 | |
| *** Ryan_Lane has joined #openstack-dev | 19:22 | |
| *** jaypipes has quit IRC | 19:22 | |
| *** FatDarrel has joined #openstack-dev | 19:22 | |
| *** kagan has quit IRC | 19:24 | |
| *** psedlak has joined #openstack-dev | 19:27 | |
| henrynash | gyee: i thought that master was still headed for rc1, no? | 19:28 |
| *** devoid has quit IRC | 19:28 | |
| henrynash | gyee: which sounds dangerous…so that can't be right! | 19:28 |
| *** dubsquared1 has joined #openstack-dev | 19:30 | |
| dubsquared1 | I have a random question about openstack services from folsom to grizzly … does anyone know if nova-network and nova-volume are still present in the grizzly release? | 19:32 |
| jgriffith | dubsquared1: nova-volume is history | 19:32 |
| dubsquared1 | so what was 'left' so volume, is now cinder … and the attach/detach commands are left up to compute, yes? | 19:33 |
| *** dachary has joined #openstack-dev | 19:33 | |
| jgriffith | dubsquared1: correct | 19:33 |
| dubsquared1 | perfect, thanks | 19:33 |
| dolphm | ayoung: as long as we're not adding new features to the API doc -- when havana opens up for dev, we should start considering new api features as v3.1 | 19:33 |
| ayoung | dolphm, sounds good | 19:34 |
| dolphm | ayoung: i also just caught a typo in one of my json changes | 19:34 |
| *** devoid has joined #openstack-dev | 19:35 | |
| dolphm | ayoung: henrynash: gyee: someone approve typo fix? https://review.openstack.org/#/c/22734/ | 19:35 |
| *** dprince has quit IRC | 19:36 | |
| ayoung | dolphm, done | 19:36 |
| sdague | henrynash: master is headed to rc1, the final branching doesn't happen until later. | 19:36 |
| dolphm | ayoung: thanks, sorry about that -- i normally validate json, but these changes were so trivial.. | 19:36 |
| ayoung | dolphm, NP. API docs will get a final scrub before GA | 19:37 |
| gyee | sdague, can you quantify "later"? If I need to make bug fix for rc1 now, which branch should I be using? | 19:38 |
| sdague | master | 19:38 |
| gyee | sdague, cool, thanks for the info | 19:38 |
| sdague | it's up to the cores on each project at this point to enforce freeze on their master | 19:38 |
| *** novas0x2a|laptop has joined #openstack-dev | 19:38 | |
| sdague | and let in what they think is right | 19:38 |
| gyee | make sense | 19:39 |
| ladquin | vishy, got a sec? | 19:39 |
| vishy | ladquin: sup? | 19:40 |
| openstackgerrit | A change was merged to openstack/glance: Document db_auto_create in default config files https://review.openstack.org/22214 | 19:40 |
| sdague | jgriffith: you seeing this error in cinder local tests? - Exception: Versioning for this project requires either an sdist tarball, or access to an upstream git repository. | 19:40 |
| ladquin | vishy, hey, I working on a commit to document this https://review.openstack.org/#/c/17273/ | 19:40 |
| ladquin | vishy, I'm sure the URI there is correct, but I don't understand exactly how it works... is this for the admin pass? | 19:41 |
| jgriffith | sdague: fixed that by bumping the oslo version | 19:43 |
| openstackgerrit | A change was merged to openstack/python-openstackclient: Added compute agent support. https://review.openstack.org/22337 | 19:43 |
| sdague | jgriffith: which oslo version? | 19:43 |
| jgriffith | sdague: https://review.openstack.org/#/c/22450/ | 19:44 |
| jgriffith | b4 | 19:44 |
| sdague | hmmm.... that doesn't fix it for me | 19:44 |
| sdague | I have b4 | 19:44 |
| jgriffith | :( | 19:44 |
| openstackgerrit | A change was merged to openstack/glance: Remove compat cfg wrapper https://review.openstack.org/22712 | 19:44 |
| jgriffith | That was the exact same error IIRC | 19:44 |
| jgriffith | I had numerous problems with the version code however | 19:46 |
| sdague | let me re-re-replicate it on totally fresh repo | 19:46 |
| ayoung | dolphm, I just rebased and merged my trusts API changes. I noticed that in the rebase it looks like a small section of the sample token data was removed. Is that correct? | 19:46 |
| jgriffith | sdague: nova? | 19:46 |
| sdague | cinder | 19:46 |
| jgriffith | sdague: Oh... | 19:46 |
| sdague | that's why I'm asking you :) | 19:47 |
| jgriffith | :) | 19:47 |
| *** soody has quit IRC | 19:47 | |
| jgriffith | Thought maybe it was just broke somewhere else now (hoped) | 19:47 |
| *** kagan has joined #openstack-dev | 19:47 | |
| jgriffith | sdague: I'm not using venv either | 19:50 |
| sdague | jgriffith: ok, let me try a couple more things | 19:50 |
| jgriffith | sdague: that may have something to do with it | 19:50 |
| sdague | get back to you in a minute | 19:50 |
| sdague | yeh, it might... :P | 19:50 |
| jgriffith | k | 19:50 |
| *** esp1 has quit IRC | 19:52 | |
| sdague | jgriffith: you just use tox? | 19:53 |
| jgriffith | sdague: No... I do run_tests.sh outside of venv | 19:54 |
| jgriffith | tox, venv etc just causes me problems on a constant basis | 19:54 |
| *** FatDarrel has quit IRC | 19:54 | |
| jgriffith | sdague: I'm downloading venv and trying your case now | 19:54 |
| *** esp1 has joined #openstack-dev | 19:54 | |
| jgriffith | It should only take 15 minutes | 19:54 |
| jgriffith | :( | 19:54 |
| sdague | so, there is an install_venv_common issue as well, which I'm going to sync and fix | 19:55 |
| *** FatDarrel has joined #openstack-dev | 19:56 | |
| sdague | but then I tripped over this other thing | 19:56 |
| jgriffith | sdague: So we've had a number of issues with the version with the oslo activity the past week or so | 19:57 |
| jgriffith | I thought/hoped that was behind us | 19:57 |
| *** gyee has quit IRC | 19:57 | |
| *** dachary has quit IRC | 20:00 | |
| *** mdomsch has joined #openstack-dev | 20:00 | |
| sdague | jgriffith: yep, still an issue. But let me at least push this install_venv_common fix into the queue | 20:01 |
| jgriffith | sounds good.... | 20:01 |
| *** arbrandes has joined #openstack-dev | 20:01 | |
| jgriffith | I'm still downloading... maybe I'll get lunch and come back | 20:01 |
| jgriffith | Oh crap... meeting! | 20:01 |
| jgriffith | sdague: I'll check in with ya when I get back | 20:02 |
| sdague | cool | 20:02 |
| *** numero8 has joined #openstack-dev | 20:05 | |
| *** cloudchimp has joined #openstack-dev | 20:07 | |
| *** sandywalsh_ has quit IRC | 20:08 | |
| *** sacharya has quit IRC | 20:08 | |
| sdague | jgriffith: it's this bug - https://bugs.launchpad.net/cinder/+bug/1125416 | 20:08 |
| uvirtbot | Launchpad bug 1125416 in cinder "run_tests.sh fails with version_string errors" [Undecided,In progress] | 20:08 |
| *** FatDarrel has quit IRC | 20:09 | |
| *** arbrandes has quit IRC | 20:11 | |
| *** arbrandes has joined #openstack-dev | 20:12 | |
| *** nsatterl_ has joined #openstack-dev | 20:14 | |
| openstackgerrit | A change was merged to openstack-infra/devstack-gate: Enable sysstat on devstack for investigate timeout https://review.openstack.org/22657 | 20:16 |
| *** reed has joined #openstack-dev | 20:17 | |
| *** danwent has quit IRC | 20:19 | |
| *** dolphm has quit IRC | 20:20 | |
| *** n0ano has joined #openstack-dev | 20:21 | |
| *** ron-slc_ has quit IRC | 20:22 | |
| mtreinish | markwash, bcwaldon: So CI found a bug with migration 018. The engine wasn't set to InnoDB when image_locations was created. Is it ok to just update the definition in the migration or should I make a new migration to switch it? | 20:23 |
| *** arbrandes has quit IRC | 20:24 | |
| mtreinish | also I guess this question applies to the status length in https://review.openstack.org/#/c/22731/ too | 20:24 |
| *** psedlak has quit IRC | 20:25 | |
| *** psedlak has joined #openstack-dev | 20:25 | |
| *** arbrandes has joined #openstack-dev | 20:25 | |
| YorikSar | topol: Hi, here I am. | 20:27 |
| YorikSar | We had OpenStack meetup here :) | 20:27 |
| *** devoid has quit IRC | 20:28 | |
| *** ayoung is now known as ayoung-lunch | 20:28 | |
| YorikSar | topol: I should've asked you in the Gerrit. How did you check if my code ran or not? | 20:28 |
| topol | Hi YorikSar, I need to finish something that should take about 15 minutes. I ran your code against a live ldap and tried to test cases. 1st test case was to see if your code would create the member role. It did that test case perfectly | 20:30 |
| *** mkollaro has quit IRC | 20:31 | |
| *** gyee has joined #openstack-dev | 20:31 | |
| topol | YorikSar, I then tried a second test case where I put a user as a member attribute under a created project and the code did not move the user. | 20:32 |
| *** devoid has joined #openstack-dev | 20:32 | |
| *** dolphm has joined #openstack-dev | 20:32 | |
| topol | I put in a comment the ldap ldif file I was using if you want to remove it. It did not appear to me that the migrate routine ran. Did I need to do something special for it to run? In any case I need to finish something but can discuss more in bout 15 minutes | 20:33 |
| topol | s/remove/try | 20:34 |
| *** crandquist has joined #openstack-dev | 20:34 | |
| YorikSar | topol: db_sync runs that migration. At least should. | 20:36 |
| topol | YorikSar, in 15 minutes I can pull up that test branch I had and try and do some debugging | 20:36 |
| YorikSar | topol: Ok, I'll check it again in my local LDAP in the mean time. | 20:37 |
| topol | k | 20:37 |
| *** gyee has quit IRC | 20:38 | |
| *** Mathnerd314 has joined #openstack-dev | 20:38 | |
| *** cp16net is now known as cp16net|away | 20:39 | |
| *** lloydde has joined #openstack-dev | 20:40 | |
| *** zzs has joined #openstack-dev | 20:41 | |
| *** gyee has joined #openstack-dev | 20:41 | |
| *** radez is now known as radez_g0n3 | 20:41 | |
| *** dubsquared1 has quit IRC | 20:41 | |
| *** arbrandes_ has joined #openstack-dev | 20:44 | |
| *** arbrandes has quit IRC | 20:46 | |
| *** cp16net|away is now known as cp16net | 20:46 | |
| *** dubsquared1 has joined #openstack-dev | 20:46 | |
| YorikSar | topol: Looks like I tested one thing then changed it. | 20:48 |
| YorikSar | topol: Fixed that problems. Let me add more tests and push it to Gerrit. | 20:49 |
| zigo_ | https://paste.debian.net/237199/ <--- What's going on with the test suite of Nova??? | 20:51 |
| *** arbrandes_ has quit IRC | 20:52 | |
| *** arbrandes__ has joined #openstack-dev | 20:52 | |
| *** aeperezt has quit IRC | 20:53 | |
| topol | YorikSar, sounds good, When its ready I will pull it down and do some more live ldap tests. | 20:55 |
| *** lglenden has quit IRC | 20:56 | |
| YorikSar | topol: I actually did LDAP tests. But it looks like I did some optimization (smashing and breaking) afterwards | 20:56 |
| ayoung-lunch | dolphm, OK, I forget exactly what I told you last night was on the todo list, but right now, all I can remember that is not done is: 1. v3 auth, 2. clean up unit tests, 3. match the list calls to the API spec. 4. make role an object in the get call (just doing name right now. | 20:56 |
| YorikSar | topol: I mean, live LDAP | 20:56 |
| bcwaldon | mtreinish: we probably want a new migration | 20:57 |
| topol | YorikSar, understand. but always good to have the non code developer test it :-) | 20:57 |
| YorikSar | topol: Sure | 20:57 |
| *** mnaser has quit IRC | 20:57 | |
| *** aeperezt has joined #openstack-dev | 20:58 | |
| clarkb | zigo_ update or venv and or testr | 20:58 |
| *** cloudchimp has quit IRC | 20:58 | |
| zigo_ | clarkb: I have the latest testr from upstream. | 20:59 |
| *** esp1 has quit IRC | 20:59 | |
| mtreinish | bcwaldon: ok, I pushed it | 20:59 |
| zigo_ | clarkb: From the package I just made ... | 20:59 |
| zigo_ | clarkb: I'm using -N, so it shouldn't be using virtualenv ! | 20:59 |
| clarkb | it should install setup.py entrypoints created | 20:59 |
| *** yolanda has quit IRC | 20:59 | |
| clarkb | latest testr that is | 20:59 |
| ayoung-lunch | gyee, as far as V3 auth, I assume I need to 1) add a checkinto the auth controller for the trust field, and 2 extend the token-factory. | 21:00 |
| clarkb | and that entrypoint isnt being found | 21:00 |
| zigo_ | clarkb: I have 0.0.9-1 | 21:00 |
| clarkb | I think there is a newer version | 21:01 |
| *** imsplitbit has quit IRC | 21:01 | |
| *** mnaser has joined #openstack-dev | 21:01 | |
| clarkb | 0.0.14 according to pypi | 21:02 |
| gyee | ayoung-lunch yes, token-factory does all the heavy lifting already | 21:02 |
| zigo_ | clarkb: I'll try, thanks. | 21:03 |
| *** ayoung-lunch is now known as ayoung | 21:03 | |
| ewindisch | dripton: take another look at 20434 - redis matchmaker ? | 21:03 |
| gyee | ayoung-lunch, all you have to do is filter the roles and set the user-id accordingly | 21:03 |
| ayoung | gyee, there any direct unit tests for token-factory, or is it just tested through the other objects? | 21:03 |
| dripton | ewindisch: I will | 21:03 |
| ewindisch | thanks :) | 21:03 |
| gyee | tested through test_v3_auth | 21:04 |
| *** zzs has left #openstack-dev | 21:04 | |
| *** dubsquared1 has quit IRC | 21:04 | |
| *** reed has quit IRC | 21:04 | |
| *** dachary has joined #openstack-dev | 21:06 | |
| *** FatDarrel has joined #openstack-dev | 21:06 | |
| *** monst__ has joined #openstack-dev | 21:08 | |
| *** cp16net is now known as cp16net|away | 21:08 | |
| *** esp1 has joined #openstack-dev | 21:09 | |
| *** martine_ has joined #openstack-dev | 21:09 | |
| *** nsatterl_ has quit IRC | 21:10 | |
| *** danwent has joined #openstack-dev | 21:10 | |
| YorikSar | topol: Here it is. | 21:11 |
| *** monst_ has quit IRC | 21:11 | |
| topol | YorikSar, did you post a new patch? | 21:12 |
| YorikSar | topol: Yep | 21:12 |
| topol | YorikSar, OK | 21:12 |
| *** arbrandes__ has quit IRC | 21:13 | |
| *** cp16net|away is now known as cp16net | 21:16 | |
| YorikSar | ayoung: I remember, you said there is some change request or bug about logging in keystone-manage. Can you point me at one? | 21:17 |
| ayoung | YorikSar, not sure what you meant. What are you trying to do? | 21:17 |
| *** reed has joined #openstack-dev | 21:18 | |
| YorikSar | ayoung: I want keystone-manage to setup logging so that, for example, if --debug is passed, it'll log to console. | 21:18 |
| *** maoy has quit IRC | 21:18 | |
| ayoung | YorikSar, not aware of anything that will do that for you. No bug that I know of. | 21:19 |
| YorikSar | (looks like Russian commas) | 21:19 |
| YorikSar | ayoung: Ok, pushing new change request then. | 21:19 |
| *** annegentle has quit IRC | 21:20 | |
| YorikSar | ayoung: Really simple: https://review.openstack.org/22745 :) | 21:21 |
| *** pabelanger has quit IRC | 21:21 | |
| *** buzztroll_ has joined #openstack-dev | 21:22 | |
| *** FatDarrel has quit IRC | 21:23 | |
| *** hugokuo has joined #openstack-dev | 21:26 | |
| *** vkmc has quit IRC | 21:27 | |
| *** numero8 has quit IRC | 21:27 | |
| *** yidclare has quit IRC | 21:27 | |
| *** Tross has quit IRC | 21:28 | |
| *** eharney has quit IRC | 21:29 | |
| sdague | jgriffith: you about around? | 21:30 |
| *** stevebaker has quit IRC | 21:31 | |
| *** psedlak has quit IRC | 21:31 | |
| *** koolhead17 has quit IRC | 21:31 | |
| *** FatDarrel has joined #openstack-dev | 21:32 | |
| *** iartarisi has joined #openstack-dev | 21:33 | |
| *** yidclare has joined #openstack-dev | 21:35 | |
| *** stevebaker has joined #openstack-dev | 21:38 | |
| *** digitalsanctum has quit IRC | 21:38 | |
| *** vipul is now known as vipul|away | 21:39 | |
| *** adalbas has quit IRC | 21:39 | |
| *** yidclare has quit IRC | 21:43 | |
| *** psedlak has joined #openstack-dev | 21:44 | |
| *** mdomsch has quit IRC | 21:46 | |
| ayoung | YorikSar, that scares me. What does it do? | 21:46 |
| YorikSar | ayoung: It just creates all Handlers for logging infrastructure. | 21:47 |
| YorikSar | ayoung: So that we can see, for example, LDAP iteractions if we want. | 21:47 |
| ayoung | YorikSar, if it changes the default output on a normal run, it will break a lot of tools | 21:47 |
| YorikSar | ayoung: It might change stderr output, but not stdout. | 21:48 |
| ayoung | YorikSar, I'd prefer it if that were not the default behaviour | 21:48 |
| YorikSar | ayoung: If keystone.conf is accessible (not Ubuntu) and has verbose=True or debug=True (not production) | 21:48 |
| YorikSar | ayoung: It is not. | 21:49 |
| ayoung | YorikSar, That will go to the log files. Not the right place | 21:49 |
| YorikSar | ayoung: By default debug and verbose are False. | 21:49 |
| ayoung | YorikSar, it shoulkd be controlled by a command line switch, not from the config files | 21:50 |
| ayoung | and should log to stderr only, I think | 21:50 |
| YorikSar | ayoung: If you pass --debug or --verbose, it will. | 21:50 |
| *** mtreinish has quit IRC | 21:51 | |
| YorikSar | ayoung: In typical logging configuration it'll create /var/log/keystone/keystone-manage.log, I think/ | 21:51 |
| *** vipul|away is now known as vipul | 21:51 | |
| YorikSar | ayoung: Just like nova-manage. | 21:51 |
| ayoung | YorikSar, not sure that is appropriate, but lets let some others chime in. | 21:51 |
| *** digitalsanctum has joined #openstack-dev | 21:52 | |
| ayoung | Added some other reviewers | 21:52 |
| YorikSar | ayoung: And I think, management process should be logged since it's basically some direct poking to backends. | 21:53 |
| ayoung | YorikSar, I'm not saying it is a bad idea, just not ready to commit... | 21:53 |
| YorikSar | ayoung: Oh, 'some' other reviewers :) | 21:54 |
| ayoung | YorikSar, I'm heads down in the trusts work. | 21:54 |
| *** martine_ has quit IRC | 21:55 | |
| openstackgerrit | A change was merged to openstack/nova: Remove Print Statement https://review.openstack.org/22743 | 21:58 |
| *** eharney has joined #openstack-dev | 21:59 | |
| *** eharney has quit IRC | 21:59 | |
| *** eharney has joined #openstack-dev | 21:59 | |
| *** mdomsch has joined #openstack-dev | 21:59 | |
| *** mdomsch has quit IRC | 22:01 | |
| ayoung | gyee, so I want to do auth_response = self.auth_v3_controller.authenticate({}, request_body) but I see that there is auth_context and auth_data and I have to say, that is not necessarily self docuementing. What is the scope of each? | 22:01 |
| ayoung | auth_info, not auth_data | 22:01 |
| *** TheEstablishment has joined #openstack-dev | 22:02 | |
| *** monst__ has quit IRC | 22:02 | |
| ayoung | oh I want authenticate_for_token don;t I | 22:03 |
| *** monst_ has joined #openstack-dev | 22:03 | |
| gyee | ayoung, are you using auth.controllers.Auth | 22:03 |
| markwash | eglynn: I was just taking a look at https://review.openstack.org/#/c/19547/, I think that approach is the right way to go and left a note on the (abandoned) review | 22:03 |
| ayoung | gyee, yep | 22:04 |
| markwash | eglynn: let me know what you think, we can figure out how to move forward for rc1 | 22:04 |
| gyee | yeah | 22:04 |
| *** TheEstablishment has quit IRC | 22:04 | |
| gyee | you need authenticate_for_toke | 22:04 |
| *** dubsquared1 has joined #openstack-dev | 22:05 | |
| gyee | ayoung, I purposely name it that way so we can have authenticate_for_x509 in the future | 22:05 |
| *** jrodom has quit IRC | 22:06 | |
| *** jbryce has quit IRC | 22:07 | |
| openstackgerrit | A change was merged to openstack/tempest: update identity to handle new table attributes https://review.openstack.org/22269 | 22:08 |
| *** mdomsch has joined #openstack-dev | 22:09 | |
| *** arbrandes__ has joined #openstack-dev | 22:09 | |
| *** dubsquared1 has quit IRC | 22:10 | |
| *** soody has joined #openstack-dev | 22:11 | |
| topol | YorikSar, the latest version did not get through basic devstack keystone operations without failing. Here is the keystone log fpaste.org/Q10k/ | 22:11 |
| *** sandywalsh has joined #openstack-dev | 22:11 | |
| *** ayoung has quit IRC | 22:11 | |
| *** sandywalsh has quit IRC | 22:12 | |
| YorikSar | topol: Looks like smth wrong with sql.connection setting | 22:13 |
| topol | YorikSar, you mean something set by devstack or something in your patch I pulled down? | 22:15 |
| YorikSar | topol: I guess, devstack config file has 'connection=' line in [sql] section. | 22:15 |
| *** esp1 has left #openstack-dev | 22:16 | |
| YorikSar | topol: Not devstack config, of course, but Keystone one. | 22:16 |
| openstackgerrit | A change was merged to openstack/tempest: Cleanup of identity/admin/test_users.py https://review.openstack.org/22649 | 22:17 |
| *** john5223 has quit IRC | 22:17 | |
| *** buzztroll_ has quit IRC | 22:17 | |
| topol | YorikSar, Oh, OK. I was testing using devstack with minimal services. I will retest turning on full services | 22:17 |
| topol | YorikSar, I will try and retest later tonight | 22:18 |
| *** boden has quit IRC | 22:18 | |
| *** mdomsch has quit IRC | 22:19 | |
| YorikSar | Wow. Ran tests the first time after refactor and they worked. | 22:19 |
| *** mdomsch has joined #openstack-dev | 22:19 | |
| *** dolphm has quit IRC | 22:22 | |
| *** n0ano has quit IRC | 22:23 | |
| *** crandquist has quit IRC | 22:23 | |
| YorikSar | topol: Look - more LDAP backend cleanup :) https://review.openstack.org/22752 | 22:26 |
| *** topol has quit IRC | 22:29 | |
| *** giulivo has quit IRC | 22:29 | |
| *** dubsquared1 has joined #openstack-dev | 22:31 | |
| *** n0ano has joined #openstack-dev | 22:33 | |
| *** esheffield has quit IRC | 22:33 | |
| *** nati_ueno has quit IRC | 22:35 | |
| *** Nachi has joined #openstack-dev | 22:35 | |
| *** nati_ueno has joined #openstack-dev | 22:35 | |
| *** spzala has quit IRC | 22:37 | |
| *** networkstatic has joined #openstack-dev | 22:37 | |
| *** esp1 has joined #openstack-dev | 22:39 | |
| *** Nachi has quit IRC | 22:39 | |
| *** eharney has quit IRC | 22:42 | |
| *** hugokuo has quit IRC | 22:42 | |
| *** hugokuo has joined #openstack-dev | 22:44 | |
| *** arbrandes__ has quit IRC | 22:44 | |
| *** hugo_kuo has joined #openstack-dev | 22:44 | |
| *** hugo_kuo has left #openstack-dev | 22:45 | |
| *** pcm_ has quit IRC | 22:46 | |
| *** mdomsch has quit IRC | 22:46 | |
| *** Tross has joined #openstack-dev | 22:46 | |
| *** hugo__kuo has joined #openstack-dev | 22:46 | |
| *** soody has quit IRC | 22:47 | |
| *** eglynn has quit IRC | 22:47 | |
| *** soody has joined #openstack-dev | 22:47 | |
| *** iartarisi has quit IRC | 22:48 | |
| *** hugo__kuo has quit IRC | 22:50 | |
| *** dubsquared1 has quit IRC | 22:50 | |
| *** dubsquared1 has joined #openstack-dev | 22:51 | |
| *** pixelbeat has joined #openstack-dev | 22:51 | |
| clarkb | jgriffith: cinder-api fails to start using the sample cinder.conf because it is missing a log_config specification. gate doesn't catch it because i believe we set it on the commandline | 22:51 |
| clarkb | jgriffith: is that something that should be a bug? | 22:52 |
| *** Gordonz has quit IRC | 22:53 | |
| *** jcmartin has quit IRC | 22:54 | |
| clarkb | hmm maybe that is not the problem | 22:55 |
| *** bdpayne has joined #openstack-dev | 22:55 | |
| *** monst__ has joined #openstack-dev | 22:56 | |
| openstackgerrit | A change was merged to openstack/nova: Fix exception handling in baremetal API. https://review.openstack.org/22627 | 22:56 |
| *** alunduil has joined #openstack-dev | 22:57 | |
| *** soody has quit IRC | 22:58 | |
| *** monst_ has quit IRC | 22:59 | |
| *** arbrandes__ has joined #openstack-dev | 23:00 | |
| *** kbringard has quit IRC | 23:00 | |
| *** dachary has quit IRC | 23:01 | |
| *** monst__ has quit IRC | 23:01 | |
| *** rnirmal has quit IRC | 23:01 | |
| bknudson | is there any way to break into the debugger during run_tests.sh? | 23:03 |
| YorikSar | bknudson: Add 'import pdb; pdb.set_trace()' to set a breakpoint in code and runs run_tests.sh with -s arg | 23:04 |
| clarkb | it also depends on your project | 23:04 |
| *** dachary has joined #openstack-dev | 23:04 | |
| bknudson | nova. | 23:05 |
| openstackgerrit | A change was merged to openstack/nova: Remove compat cfg wrapper https://review.openstack.org/22711 | 23:05 |
| clarkb | bknudson: yeah nova is going to be slightly different because it uses testr as a test runner runner | 23:05 |
| openstackgerrit | A change was merged to openstack/nova: Add module prefix to exception types https://review.openstack.org/22591 | 23:05 |
| *** networkstatic has quit IRC | 23:05 | |
| bknudson | I get stdout: {{{ ... (Pdb) }}} so it's hitting the break but not interactive. | 23:06 |
| clarkb | bknudson: add the pdb stuff to your code then source the venv. Then `python -m testtools.run path.to.test` | 23:06 |
| clarkb | bknudson: yeah testr captures stdout and stderr and logs then logs it all into subunit format. to avoid that use testtools.run directly | 23:06 |
| *** jbr_ has joined #openstack-dev | 23:08 | |
| bknudson | clarkb: thanks. worked like a champ | 23:08 |
| *** dolphm has joined #openstack-dev | 23:08 | |
| bknudson | e.g., tools/with_venv.sh python -m testtools.run nova.tests.api.openstack.compute.contrib.test_cells.TestCellsXMLSerializer | 23:08 |
| *** cloudchimp has joined #openstack-dev | 23:09 | |
| clarkb | yeah I should add that as a thing to run_tests | 23:09 |
| bknudson | clarkb: call it -s. | 23:09 |
| *** Ryan_Lane has quit IRC | 23:09 | |
| *** Ryan_Lane has joined #openstack-dev | 23:09 | |
| *** ayoung has joined #openstack-dev | 23:09 | |
| *** FatDarrel has quit IRC | 23:11 | |
| clarkb | will do | 23:11 |
| *** nati_ueno has quit IRC | 23:11 | |
| *** alunduil has quit IRC | 23:11 | |
| *** nati_ueno has joined #openstack-dev | 23:12 | |
| YorikSar | When I don't find a way to pass stdin/out to pdb, I open another tab (tmux pane), run 'tty; sleep 10000' there and change breakpoint code to smth like "f1,f2=open('/dev/pts/XX','r'),open('/dev/pts/XX','w');import pdb; pdb.Pdb(stdin=f1,stdout=f2).set_trace()" | 23:12 |
| clarkb | heh of course run_tests.sh already has -s for something else | 23:12 |
| YorikSar | Then debug in separate pane as usual. | 23:12 |
| *** dubsquared1 has quit IRC | 23:13 | |
| YorikSar | clarkb: You can purge that old option there. | 23:14 |
| YorikSar | clarkb: virtualenv creates isolated environment by default. | 23:14 |
| clarkb | I think I am just giong to use the old flags before the testr switch | 23:15 |
| clarkb | and reviewers can sort out what they want it to be long term | 23:15 |
| bknudson | btw - what does -s do? I didn't find it in the help. | 23:15 |
| bknudson | for keystone or nova | 23:15 |
| clarkb | bknudson: for nova it is --no-site-packages | 23:15 |
| YorikSar | For Nova: -s, --no-site-packages Isolate the virtualenv from the global Python environment | 23:15 |
| bknudson | oh, there it is. | 23:16 |
| YorikSar | bknudson: For all other it's passed to nosetests | 23:16 |
| bknudson | why would I want that for debug? | 23:16 |
| YorikSar | bknudson: And then nosetsts won't intercept stdin/out/err | 23:16 |
| YorikSar | bknudson: And Pdb is happy. | 23:17 |
| clarkb | YorikSar: nova does not use nose anymore | 23:17 |
| YorikSar | clarkb: Why?.. | 23:17 |
| clarkb | because we wanted clean parallel testing and nose does some really weird things | 23:18 |
| *** vipul is now known as vipul|away | 23:18 | |
| bknudson | clarkb: the parallel testing is great. | 23:19 |
| clarkb | like not fail tests if setUpModule() throws an exception | 23:19 |
| clarkb | or passing if 0 tests are run | 23:19 |
| YorikSar | Then I'll say it this way: Why aren't other project switched to testr? :) | 23:19 |
| clarkb | YorikSar: we have been trying to. its not a simple process as all intertest conflicts need to be sorted out | 23:19 |
| clarkb | we have been working on it though. | 23:19 |
| *** jimfehlig has quit IRC | 23:19 | |
| clarkb | I think mordred has most of the python-*clients done but I may be mistaken | 23:20 |
| clarkb | and at this point I think many projects would rather we wait for havana | 23:20 |
| YorikSar | I've seen some very scary monsters in nose source. Always wondered if there is something as powerful, but cleaner and better. | 23:20 |
| clarkb | I really like testr and subunit | 23:20 |
| clarkb | the subunit logs are great I think and was something missing with nose | 23:20 |
| *** davidh_ has quit IRC | 23:21 | |
| clarkb | (testr keeps track of successive runs so taht you can run just failing tests and ask for statistics) | 23:21 |
| bknudson | clarkb: so is there a similar command for keystone? | 23:21 |
| bknudson | to tools/with_venv.sh python -m testtools.run ... | 23:21 |
| clarkb | bknudson: not yet | 23:22 |
| clarkb | it would be run_tests.sh -s as keystone still uses nose | 23:22 |
| YorikSar | bknudson: For keystone you can use -s | 23:22 |
| jgriffith | clarkb: yes, sounds like a but to me | 23:23 |
| jgriffith | clarkb: but it needs added in devstack | 23:23 |
| YorikSar | clarkb: Why is it run as 'setup.py testr', not 'setup.py test' that sounds more common? | 23:23 |
| jgriffith | Or I mean "it is set in devstack" | 23:23 |
| clarkb | YorikSar: because we needed to condense several testr specific commands into one and mordred went that route | 23:24 |
| clarkb | jgriffith: yeah I am not entirely sure what was going on. I think it was swift ran into a problem where we use the --log-config variable on the command line in devstack and oslo config changes affected swift.conf log parameters | 23:24 |
| clarkb | echohead: ^ say hello to jgriffith | 23:25 |
| openstackgerrit | A change was merged to openstack-dev/devstack: Fixes Bug1130377 devstack ldap set values in keystone.conf https://review.openstack.org/22356 | 23:26 |
| *** vipul|away is now known as vipul | 23:27 | |
| *** Tross has quit IRC | 23:27 | |
| jgriffith | clarkb: all that I don't know why this is an issue? | 23:27 |
| *** henrynash has quit IRC | 23:28 | |
| *** datsun180b has quit IRC | 23:29 | |
| *** bdpayne has quit IRC | 23:29 | |
| jgriffith | sdague: back... what's up? | 23:30 |
| *** bdpayne has joined #openstack-dev | 23:31 | |
| clarkb | jgriffith: apparently using the sample cinder.conf and logging.conf (edited slightly for env specific things) tip of master cinder-api fails to start. | 23:31 |
| *** alszar has quit IRC | 23:32 | |
| clarkb | jgriffith: I think swift had a similar problem and it got through the gate because we start services in the gate with additional command line options | 23:32 |
| zykes- | hmm dolphm about ? I upgrade to latest keystone trunk and now I get sqlalchemy.exc.NoSuchTableError: `tenant` | 23:32 |
| jgriffith | clarkb: Maybe I'm not following, the default in all the projects I believe is log_config=<None> isn't it? | 23:32 |
| zykes- | http://pastebin.com/FNPF8Vjy | 23:33 |
| jgriffith | Or are you saying if folks over-ride it it breaks? | 23:33 |
| *** digitalsanctum has quit IRC | 23:33 | |
| clarkb | jgriffith: yeah I think it was overriden. | 23:34 |
| *** dolphm has quit IRC | 23:34 | |
| *** soody has joined #openstack-dev | 23:35 | |
| clarkb | I believe it was complaining about log-format specifically in oslo config | 23:35 |
| *** trapniii has quit IRC | 23:36 | |
| *** lloydde has quit IRC | 23:36 | |
| *** lloydde has joined #openstack-dev | 23:36 | |
| *** dolphm has joined #openstack-dev | 23:37 | |
| echohead | hey jgriffith: i'm here now. | 23:38 |
| jgriffith | clarkb: well that's unfortunate, I think the real problem is cinder-api doesn't even register the FLAGS :( | 23:38 |
| echohead | jgriffith: reading scrollback. | 23:38 |
| jgriffith | echohead: hi there | 23:38 |
| echohead | hi | 23:38 |
| jgriffith | echohead: Assuming you discovered the issue clarkb has mentioned/ | 23:38 |
| jgriffith | ? | 23:38 |
| echohead | right. | 23:38 |
| echohead | one sec and i'll make a paste with my config and stuff. | 23:39 |
| jgriffith | K | 23:39 |
| *** dolphm has quit IRC | 23:39 | |
| jgriffith | although I think it's most likely just because cinder-api is ignoring your Flags anyway | 23:39 |
| jgriffith | It's only reading command line options | 23:39 |
| jgriffith | So as clarkb pointed out, works fine in the gates | 23:40 |
| jgriffith | and devstack | 23:40 |
| echohead | i tried passing --config-file, --log-config, --config-dir via command line, to no avail. | 23:40 |
| *** lloydde has quit IRC | 23:41 | |
| clarkb | bknudson: vishy sdague https://review.openstack.org/#/c/22761/ | 23:41 |
| jgriffith | echohead: oh, even echohead I would've expected --config-file to work | 23:42 |
| *** rpedde is now known as rpedde_away | 23:42 | |
| *** dhellmann-afk has quit IRC | 23:43 | |
| *** psedlak has quit IRC | 23:43 | |
| echohead | jgriffith: http://paste.openstack.org/show/32379/ | 23:44 |
| sdague | jgriffith: just curious if you were able to confirm the bug | 23:44 |
| jgriffith | sdague: no, but found another one possibly | 23:44 |
| echohead | (sorry the line-endings got botched in the paste) | 23:44 |
| sdague | jgriffith: with a clean venv you weren't able to confirm? | 23:45 |
| *** dosaboy has quit IRC | 23:45 | |
| sdague | jgriffith: your source tree isn't a devstack source tree is it? | 23:45 |
| jgriffith | sdague: I can't get dam venv to install | 23:45 |
| jgriffith | no... on my laptop | 23:45 |
| *** dosaboy has joined #openstack-dev | 23:45 | |
| sdague | jgriffith: oh, you need the fix I put in | 23:45 |
| jgriffith | and I just got back so haven't tried it again | 23:45 |
| jgriffith | crap | 23:45 |
| jgriffith | yeah, I knew that | 23:45 |
| jgriffith | dang | 23:45 |
| sdague | clarkb: what's the bug for the VOLUMES fail? | 23:46 |
| jgriffith | I made the set -u but forgot the bootstrapping one | 23:46 |
| sdague | on quantum | 23:46 |
| bknudson | clarkb: we should still support running without -d ;) | 23:46 |
| clarkb | bknudson: yeah my bad | 23:47 |
| sdague | jgriffith: my install_venv_common.sh got gate killed by that same VOLUMES fail in quantum | 23:47 |
| clarkb | sdague: uh I recently restarted firefox and now I don't have it handy | 23:47 |
| jgriffith | sdague: So that's our friendly "instance fails to come up" | 23:47 |
| echohead | jgriffith: clarkb: i have to run. msg me on irc if there's anything else i can do, will respond later. | 23:47 |
| *** annegentle has joined #openstack-dev | 23:47 | |
| sdague | clarkb: have keywords? | 23:47 |
| jgriffith | sdague: Yeah, that's what I wsa pointing out, it's not really "volumes" that failed | 23:48 |
| clarkb | sdague: I think nati updated it | 23:48 |
| sdague | clarkb: which project was it assigned to? | 23:48 |
| *** trapniii has joined #openstack-dev | 23:48 | |
| jgriffith | sdague: ERROR: No server with a name or ID of 'ex-vol-inst' exists. | 23:48 |
| clarkb | sdague: I think it may be under nova | 23:48 |
| jgriffith | oh wait... sorry | 23:48 |
| *** dosaboy has quit IRC | 23:48 | |
| clarkb | because it was there first when nati_ueno was upadting things | 23:49 |
| nati_ueno | > | 23:49 |
| nati_ueno | ? | 23:49 |
| clarkb | nati_ueno: the bug that you added sysstat to debug | 23:50 |
| clarkb | nati_ueno: what is the bug number? | 23:50 |
| sdague | https://bugs.launchpad.net/nova/+bug/1101147 ? or another one | 23:50 |
| uvirtbot | Launchpad bug 1101147 in nova ""Instance didn't become active" in Devstack Gate Exercises" [Critical,Confirmed] | 23:50 |
| nati_ueno | clarkb: This one https://bugs.launchpad.net/nova/+bug/1131030 ? | 23:50 |
| uvirtbot | Launchpad bug 1131030 in nova "echo 'volume didnt become available within 180 seconds'" [Undecided,Invalid] | 23:50 |
| clarkb | yeah | 23:50 |
| clarkb | ty | 23:51 |
| sdague | nati_ueno: cool, thanks | 23:51 |
| sdague | jgriffith: well once the install_venv fix lands, you should be to a replication state | 23:51 |
| nati_ueno | sdague: np. It looks cinder connection problem | 23:51 |
| jgriffith | sdague: I just pulled the change in running now | 23:51 |
| jgriffith | sdague: So this is the c-vol log empty thing again! | 23:52 |
| sdague | I'm about to become a pumpkin, friday night and all. hopefully it's enough to get you running | 23:52 |
| sdague | jgriffith: yeah, though honestly I just want unit tests to run locally :) | 23:52 |
| jgriffith | clarkb: fungi any chance that instance still exists? | 23:52 |
| sdague | so I can finish the db unit tests for cinder | 23:52 |
| clarkb | jgriffith: no we checked after the most recent one | 23:52 |
| jgriffith | http://logs.openstack.org/22739/1/gate/gate-tempest-devstack-vm-quantum/9725/logs/ | 23:52 |
| clarkb | they get cleaned up pretty quickly | 23:52 |
| jgriffith | bummer | 23:53 |
| *** dontalton has quit IRC | 23:53 | |
| clarkb | jgriffith: we will probably need to have jenkins check for specific failures and not delete the instance if they occur | 23:53 |
| clarkb | but that will take a little work to do | 23:53 |
| *** jcmartin has joined #openstack-dev | 23:53 | |
| jgriffith | clarkb: yes please | 23:53 |
| jgriffith | :) | 23:53 |
| *** gongysh has joined #openstack-dev | 23:53 | |
| jgriffith | clarkb: what about just adding a "check service" after screen_it ? | 23:53 |
| *** mrodden has quit IRC | 23:53 | |
| jgriffith | as it seems it could be one of a number of services that hit this | 23:54 |
| *** colinmcnamara has quit IRC | 23:54 | |
| clarkb | jgriffith: and error early with diagnostic info if the service is not up? | 23:54 |
| jgriffith | clarkb: exactly | 23:54 |
| jgriffith | fail fast | 23:54 |
| clarkb | yeah if you want to add that into devstack I would be for it :) | 23:54 |
| jgriffith | and think about what we might be able to collect | 23:54 |
| jgriffith | clarkb: Yeah, I will see about that after I sort the things echohead and sdague are seeing | 23:55 |
| clarkb | I think that adding the ability to capture arbitrary nodes to jenkins is a useful thing though so we should probably do that too | 23:55 |
| jgriffith | clarkb: that one I'll leave to you :) | 23:55 |
| sdague | alright, well I'm out for the rest of the night. But if you guys come up with a devstack fix, I'll look in the morning over coffee. Otherwise monday. | 23:56 |
| sdague | have a good one folks | 23:56 |
| clarkb | you too | 23:56 |
| jgriffith | sdague: take care | 23:56 |
| jgriffith | echohead: still there? | 23:57 |
| *** soody has quit IRC | 23:58 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!
