| openstackgerrit | Fei Long Wang proposed openstack/zaqar: Fix metadata update issue for v1.1 and v2 https://review.openstack.org/280941 | 00:26 |
|---|---|---|
| wanghao | flwang: ping | 00:41 |
| wanghao | flwang: hi, I noticed this patch(https://review.openstack.org/#/c/276603/) merged fail since "Additional properties are not allowed (u'pool_group' was unexpected)", but zaqar server's change depends on this. | 00:45 |
| wanghao | flwang: do you have any suggestion about this situation? | 00:45 |
| *** rcernin has quit IRC | 00:46 | |
| flwang | wanghao: yep, i saw that. and it's a little bit tricky | 00:56 |
| flwang | since our server side is using the client functional test as gate job | 00:56 |
| wanghao | flwang: yes, and client also need server to run gate job too.... | 00:58 |
| flwang | wanghao: is it possible to resolve it on server side? like in Mitaka, we can accept both pool and pool_group | 00:58 |
| flwang | and totally remove the support of 'poo' name in Newton | 00:58 |
| wanghao | flwang: well, this is a way I think we can try it. | 00:58 |
| wanghao | flwang: sounds good. | 00:59 |
| flwang | that can resolve current issue and we can also give end user some time to migrate | 00:59 |
| wanghao | flwang: sure, will update the patch for this. | 00:59 |
| flwang | wanghao: i think you just need change a little bit at the api level | 01:00 |
| wanghao | flwang: emm, so the client side depends on server side. | 01:01 |
| flwang | wanghao: for sure | 01:02 |
| wanghao | flwang: okay :) | 01:02 |
| *** achanda has quit IRC | 01:03 | |
| flwang | wanghao: thanks for working on this | 01:06 |
| wanghao | flwang: you're welcome :) | 01:17 |
| *** achanda has joined #openstack-zaqar | 01:20 | |
| *** kgriffs is now known as kgriffs|afk | 01:43 | |
| *** wxy__ has quit IRC | 02:06 | |
| *** wxy has joined #openstack-zaqar | 02:10 | |
| *** godlike has quit IRC | 02:24 | |
| openstackgerrit | Fei Long Wang proposed openstack/zaqar: Fix the TTL issue of subscriptions for MongoDB https://review.openstack.org/270464 | 02:26 |
| *** godlike has joined #openstack-zaqar | 02:29 | |
| *** godlike has joined #openstack-zaqar | 02:29 | |
| *** david_cu has joined #openstack-zaqar | 03:15 | |
| *** flwang1 has quit IRC | 03:56 | |
| *** pt_15 has quit IRC | 05:55 | |
| *** GB21 has joined #openstack-zaqar | 06:26 | |
| wxy | Our client's Jenkins check doesn't contain functional test which is included in Jenkins gate. Does anyone know why? | 06:40 |
| wxy | this one: gate-zaqarclient-dsvm-functional | 06:41 |
| wxy | So we can't get the functional error at once until send the patch to gate. | 06:42 |
| *** achanda has quit IRC | 07:14 | |
| *** achanda has joined #openstack-zaqar | 07:17 | |
| *** GB21 has quit IRC | 07:40 | |
| *** rcernin has joined #openstack-zaqar | 07:50 | |
| *** pcaruana has joined #openstack-zaqar | 08:05 | |
| *** GB21 has joined #openstack-zaqar | 08:06 | |
| *** jtomasek|afk is now known as jtomasek | 08:15 | |
| *** dynarro has joined #openstack-zaqar | 08:29 | |
| *** achanda has quit IRC | 09:25 | |
| *** achanda has joined #openstack-zaqar | 09:28 | |
| *** achanda has quit IRC | 09:29 | |
| *** achanda has joined #openstack-zaqar | 09:46 | |
| *** openstackgerrit has quit IRC | 10:02 | |
| *** openstackgerrit has joined #openstack-zaqar | 10:03 | |
| *** achanda has quit IRC | 10:12 | |
| *** achanda has joined #openstack-zaqar | 10:14 | |
| *** achanda has quit IRC | 10:16 | |
| *** dynarro has quit IRC | 10:25 | |
| openstackgerrit | wanghao proposed openstack/zaqar: "group" word should be used instead "pool" https://review.openstack.org/273931 | 10:53 |
| openstackgerrit | wanghao proposed openstack/zaqar: Update operation should return updated result https://review.openstack.org/280563 | 11:03 |
| *** achanda has joined #openstack-zaqar | 11:17 | |
| *** mdnadeem has quit IRC | 11:22 | |
| *** achanda has quit IRC | 11:24 | |
| openstackgerrit | wanghao proposed openstack/zaqar: "group" word should be used instead "pool" https://review.openstack.org/273931 | 11:38 |
| *** mdnadeem has joined #openstack-zaqar | 11:42 | |
| openstackgerrit | wanghao proposed openstack/python-zaqarclient: "pool_group" word should be used instead "pool" https://review.openstack.org/276603 | 12:14 |
| *** GB21 has quit IRC | 12:14 | |
| openstackgerrit | wanghao proposed openstack/zaqar: Update operation should return updated result https://review.openstack.org/280563 | 12:28 |
| EmilienM | flwang: when you around, I need your help | 12:53 |
| vkmc | EmilienM, 'sup | 12:53 |
| vkmc | :P | 12:53 |
| EmilienM | hey | 12:53 |
| EmilienM | I'm trying to start zaqar | 12:54 |
| EmilienM | but I got this http://logs.openstack.org/51/279351/23/check/gate-puppet-zaqar-puppet-beaker-rspec-dsvm-centos7/62fe4f5/logs/zaqar/zaqar.txt.gz | 12:54 |
| EmilienM | while I'm using replicaset | 12:54 |
| EmilienM | you can see the conf http://logs.openstack.org/51/279351/23/check/gate-puppet-zaqar-puppet-beaker-rspec-dsvm-centos7/62fe4f5/logs/etc/zaqar/zaqar.conf.txt.gz | 12:54 |
| openstackgerrit | wanghao proposed openstack/zaqar: "group" word should be used instead "pool" https://review.openstack.org/273931 | 12:55 |
| vkmc | let me see | 12:55 |
| vkmc | EmilienM, paste the conf file please | 12:55 |
| vkmc | sorry, you did | 12:56 |
| EmilienM | vkmc: you have it ^ | 12:56 |
| vkmc | yeah | 12:56 |
| EmilienM | and the mongo RS is created : http://logs.openstack.org/51/279351/23/check/gate-puppet-zaqar-puppet-beaker-rspec-dsvm-centos7/62fe4f5/console.html#_2016-02-18_05_41_33_463 | 12:56 |
| EmilienM | *before* starting zaqar | 12:56 |
| EmilienM | but looking at zaqar/storage/mongodb/driver.py it seems I need more than 1 conn.node or a mongos | 12:57 |
| EmilienM | a mongos is a mongo shard? | 12:57 |
| vkmc | I see yeah... that should work | 12:58 |
| vkmc | maybe it's a problem with the replica set | 12:58 |
| EmilienM | can I just use uri = mongodb://127.0.0.1:27017 ? | 12:59 |
| vkmc | the mongos is a service between the app and the mongodb instance | 12:59 |
| EmilienM | yeah, it's for sharding right? | 13:00 |
| vkmc | that helps retrieving data from the shards | 13:00 |
| vkmc | yeah | 13:00 |
| vkmc | ehm, changing the uri may work | 13:00 |
| vkmc | try that | 13:00 |
| EmilienM | ok | 13:00 |
| EmilienM | vkmc: do you run mongodb in devstack? | 13:01 |
| EmilienM | yes you do, I see in plugin.sh | 13:01 |
| EmilienM | you use mongodb://localhost:27017/zaqar for URL | 13:01 |
| EmilienM | what if I don't specify a dbname? | 13:02 |
| vkmc | EmilienM, it creates a default one | 13:08 |
| vkmc | EmilienM, you should be able to specify the uri without specifying the db for it | 13:09 |
| vkmc | EmilienM, are you working on the puppet modules for it? | 13:10 |
| EmilienM | vkmc: yes, puppet-zaqar is done | 13:15 |
| EmilienM | https://github.com/openstack/puppet-zaqar | 13:16 |
| *** achanda has joined #openstack-zaqar | 13:22 | |
| EmilienM | vkmc: still failing, logs are coming | 13:23 |
| vkmc | EmilienM, k k | 13:23 |
| EmilienM | I think I'll just try to reproduce on my vm and debug in there | 13:23 |
| *** achanda has quit IRC | 13:27 | |
| therve | EmilienM, Do you use replication or sharding? | 13:28 |
| EmilienM | therve: just rs, but on a localhost | 13:28 |
| EmilienM | therve: http://logs.openstack.org/51/279351/24/check/gate-puppet-zaqar-puppet-beaker-rspec-dsvm-centos7/378bd84/logs/ | 13:29 |
| EmilienM | Either a replica set or a mongos is required to guarantee message delivery | 13:29 |
| EmilienM | I probably missed something... trying to reproduce | 13:30 |
| therve | Yeah I guess it's an issue on how mongo is configured | 13:31 |
| EmilienM | it's weird, I created the Replicaset | 13:31 |
| EmilienM | http://logs.openstack.org/51/279351/23/check/gate-puppet-zaqar-puppet-beaker-rspec-dsvm-centos7/62fe4f5/console.html#_2016-02-18_05_41_33_463 | 13:31 |
| therve | Yeah that's just puppet output? What does it really do? | 13:32 |
| therve | EmilienM, http://logs.openstack.org/51/279351/24/check/gate-puppet-zaqar-puppet-beaker-rspec-dsvm-centos7/378bd84/logs/ps.txt.gz there is only one process here | 13:32 |
| EmilienM | it create a replicaset with https://github.com/puppetlabs/puppetlabs-mongodb/blob/master/lib/puppet/provider/mongodb_replset/mongo.rb | 13:32 |
| EmilienM | therve: I don't run mongos, just RS so no extra process, isn't? | 13:33 |
| therve | EmilienM, You need several nodes in your set? | 13:33 |
| EmilienM | why? | 13:34 |
| therve | Because that's what zaqar wants | 13:34 |
| EmilienM | I'm following http://docs.openstack.org/developer/zaqar/devref/development.environment.html | 13:34 |
| EmilienM | I run single node | 13:34 |
| therve | EmilienM, "This URI points to single mongodb node which of course is not reliable," | 13:35 |
| therve | You set unreliable to False | 13:35 |
| therve | Set it to true if you don't catre | 13:35 |
| therve | care | 13:35 |
| EmilienM | ok so if I want to run single node, I need to use unreliable | 13:35 |
| therve | Or run 2 mongo processes | 13:36 |
| therve | You can run 2 processes on the a single node | 13:36 |
| EmilienM | I would like to save resources (memory) | 13:36 |
| therve | Then yeah set unreliable :) | 13:36 |
| EmilienM | therve: look https://github.com/openstack/puppet-openstack-integration#description | 13:36 |
| EmilienM | we already have lot of services | 13:36 |
| EmilienM | therve, vkmc: thx for your help -- testing again with unreliable now | 13:39 |
| *** GB21 has joined #openstack-zaqar | 13:43 | |
| vkmc | EmilienM++ | 13:43 |
| *** amitgandhinz has joined #openstack-zaqar | 13:58 | |
| *** sriram has joined #openstack-zaqar | 14:09 | |
| EmilienM | vkmc, therve: it worked! thanks again | 14:13 |
| vkmc | EmilienM, cool! | 14:14 |
| *** dynarro has joined #openstack-zaqar | 14:32 | |
| *** jtomasek has quit IRC | 14:39 | |
| *** jtomasek has joined #openstack-zaqar | 14:44 | |
| EmilienM | vkmc, therve: tempest fail though http://logs.openstack.org/15/281515/7/check/gate-puppet-openstack-integration-scenario002-tempest-dsvm-centos7/6e52338/console.html#_2016-02-18_14_26_43_958 -- but I think it's related to the logs I see in: http://logs.openstack.org/15/281515/7/check/gate-puppet-openstack-integration-scenario002-tempest-dsvm-centos7/6e52338/logs/zaqar/zaqar.txt.gz | 14:46 |
| EmilienM | Stage <zaqar.storage.mongodb.messages.MessageQueueHandler object at 0x359fb90> does not implement create | 14:46 |
| EmilienM | should I worry about that ^ ? | 14:46 |
| therve | No I don't think that warning matters | 14:47 |
| EmilienM | only some tests are failing: test_release_claim, test_delete_multiple_messages, test_get_multiple_messages and test_post_messages | 14:47 |
| therve | I don't think zaqar tempest tests are maintained though | 14:47 |
| EmilienM | should they work with mongodb? | 14:48 |
| EmilienM | therve: why tempest tests would not be maintained? | 14:48 |
| EmilienM | you don't gate with tempest jobs? | 14:48 |
| vkmc | EmilienM, we don't | 14:49 |
| EmilienM | O_O | 14:49 |
| therve | Yeah not anymore | 14:49 |
| vkmc | EmilienM, we don't have enough bandwidth to maintain those | 14:49 |
| EmilienM | vkmc: is it enough if I run api.messaging.test_queues ? (they all pass) | 14:50 |
| EmilienM | vkmc: http://logs.openstack.org/15/281515/7/check/gate-puppet-openstack-integration-scenario002-tempest-dsvm-centos7/6e52338/console.html#_2016-02-18_14_22_57_325 | 14:50 |
| therve | Ideally you'd run the functional tests instead | 14:50 |
| EmilienM | therve: well, you have to understand we run a single project for testing (tempest) and do go through each project to find how they run testing | 14:51 |
| EmilienM | we basically run one command that test all projects: https://github.com/openstack/puppet-openstack-integration/blob/master/run_tests.sh#L104 | 14:51 |
| EmilienM | I understand why you don't use tempest, you don't have enough people in your team | 14:53 |
| EmilienM | but for people like us who like single interface to run functional tests, it's painful to have multiple tools to run testings | 14:53 |
| EmilienM | so we'll just run api.messaging.test_queues, I hope it's still useful | 14:54 |
| therve | Not using tempest is not a matter of people, rather something to do with being in control of our tests | 14:57 |
| therve | That said, the tests there should be fixed | 14:57 |
| ryansb | EmilienM: also, there was a migration away from "all the anciliary projects use tempest" | 14:58 |
| ryansb | because it forced out-of-tree functional tests | 14:58 |
| EmilienM | therve: no problem, I'm just giving feedback. | 14:59 |
| EmilienM | therve: gnocchi also does not run tempest | 14:59 |
| EmilienM | aodh, same | 14:59 |
| EmilienM | etc | 14:59 |
| EmilienM | for projects like puppet, were we have an integrated CI, we need a single tool to run our tests | 15:00 |
| EmilienM | and until now, tempest is good for us. | 15:00 |
| therve | EmilienM, So yeah that's why I don't get it. You need a single tool, except when you don't? | 15:01 |
| therve | What do you do for gnocchi? | 15:01 |
| ryansb | yeah, tempest is nice for you guys. Unfortunately it's tough for project maintainers since trees can be out of sync | 15:01 |
| ryansb | and it was hard to get reviews on new tempest tests, which made the tests worse | 15:02 |
| therve | ryansb, Does tempest-lib help with that? IE having tempest tests but in our tree? | 15:03 |
| therve | Always wanted to check what it does precisely | 15:03 |
| ryansb | I don't know, actually. It's probably intended to address that case, but it's not something I've used yet | 15:05 |
| EmilienM | therve: right, we try to run a single tool (tempest now) - for gnocchi, we don't test it directly, but we test ceilometer which use gnocchi dispatcher. | 15:07 |
| EmilienM | no problem, we'll just run api.messaging.test_queues | 15:07 |
| Eva-i | ryansb: therve: hello, do you think it's a bug? https://bugs.launchpad.net/zaqar/+bug/1544328 | 15:17 |
| openstack | Launchpad bug 1544328 in zaqar "Zaqar doesn't require X-PROJECT-ID header in requests (noauth)" [Undecided,New] | 15:17 |
| *** GB21 has quit IRC | 15:19 | |
| *** kgriffs|afk is now known as kgriffs | 15:22 | |
| *** kgriffs is now known as kgriffs|afk | 15:22 | |
| ryansb | I guess it depends if we want it to be a SHOULD or a MUST | 15:24 |
| ryansb | based on our docs (must) then yeah, it's a bug | 15:24 |
| *** achanda has joined #openstack-zaqar | 15:25 | |
| *** kgriffs|afk is now known as kgriffs | 15:25 | |
| *** achanda has quit IRC | 15:30 | |
| Eva-i | ryansb: thank you, Fei Long thinks it's a bug too. But no matter what our docs are saying, fixing it will break backward compatibility. Changes like this seem to be not considered as acceptable http://specs.openstack.org/openstack/api-wg/guidelines/evaluating_api_changes.html#guidance. Maybe it's acceptable for us, because we don't have much users, like Fei Long says. But if we continue breaking backward compatibility, we will never have much | 15:39 |
| therve | Frankly I don't even know why we have a noauth mode | 15:48 |
| Eva-i | therve: Maybe noauth can be useful for systems which are not exposed to the public where keystone isn't needed. I know that without keystone, Zaqar performs much faster. | 15:51 |
| ryansb | hm, tough spot then. | 15:57 |
| *** zigo has quit IRC | 16:03 | |
| *** zigo has joined #openstack-zaqar | 16:05 | |
| *** pcaruana has quit IRC | 16:15 | |
| *** dynarro has quit IRC | 16:15 | |
| *** rcernin has quit IRC | 16:17 | |
| *** pt_15 has joined #openstack-zaqar | 16:43 | |
| Eva-i | wow, this time I found really great bug https://bugs.launchpad.net/python-zaqarclient/+bug/1547100 | 16:45 |
| openstack | Launchpad bug 1547100 in Python client library for Zaqar "Zaqar freezes on message POST request without Content-type header" [Undecided,New] - Assigned to Eva Balycheva (ubershy) | 16:45 |
| Eva-i | Feel free to confirm bugs you like https://bugs.launchpad.net/~ubershy | 16:52 |
| *** exploreshaifali has joined #openstack-zaqar | 16:55 | |
| ryansb | Eva-i: cool, I'll plan to do some triage later today | 17:09 |
| ryansb | got some heat stuff to work on first | 17:09 |
| Eva-i | ryansb: oki =) | 17:10 |
| *** achanda has joined #openstack-zaqar | 17:28 | |
| *** achanda has quit IRC | 17:33 | |
| *** rcernin has joined #openstack-zaqar | 17:36 | |
| *** achanda has joined #openstack-zaqar | 18:03 | |
| *** achanda has quit IRC | 18:26 | |
| *** achanda has joined #openstack-zaqar | 18:37 | |
| *** kgriffs is now known as kgriffs|afk | 18:56 | |
| *** achanda has quit IRC | 19:03 | |
| *** kgriffs|afk is now known as kgriffs | 19:08 | |
| *** achanda has joined #openstack-zaqar | 19:17 | |
| EmilienM | hey zaqar folks, I reported https://bugs.launchpad.net/zaqar/+bug/1547188 | 19:35 |
| openstack | Launchpad bug 1547188 in zaqar "some tempest tests are broken" [Undecided,New] | 19:35 |
| *** exploreshaifali has quit IRC | 19:39 | |
| vkmc | thanks EmilienM | 19:50 |
| *** achanda has quit IRC | 19:51 | |
| *** rcernin has quit IRC | 19:55 | |
| *** achanda has joined #openstack-zaqar | 19:58 | |
| EmilienM | vkmc: do you know why all zaqar tests are run while I'm trying to run only some of them on https://review.openstack.org/#/c/281515/10/run_tests.sh ? | 20:01 |
| vkmc | EmilienM, nope | 20:06 |
| *** boris-42 has quit IRC | 20:16 | |
| *** boris-42 has joined #openstack-zaqar | 20:18 | |
| *** boris-42 has quit IRC | 20:24 | |
| EmilienM | vkmc: I figured | 20:24 |
| vkmc | EmilienM, what's up? | 20:24 |
| EmilienM | all zaqar tests are tagged as smoke | 20:24 |
| EmilienM | and that's bad. | 20:24 |
| EmilienM | https://github.com/openstack/tempest/blob/master/tempest/api/messaging/test_claims.py#L54 | 20:25 |
| EmilienM | you tag all tests as smoke and tests are broken | 20:25 |
| EmilienM | so if someone (me here) runs smoke with zaqar enabled | 20:25 |
| EmilienM | tempest fails. | 20:25 |
| EmilienM | also, it would be great if zaqar PTL could find a QA liaison: https://wiki.openstack.org/wiki/CrossProjectLiaisons#QA | 20:26 |
| vkmc | I see | 20:26 |
| vkmc | all right, we will check that out | 20:27 |
| vkmc | thanks EmilienM | 20:27 |
| *** mtreinish has joined #openstack-zaqar | 20:30 | |
| EmilienM | vkmc: https://review.openstack.org/282048 | 20:30 |
| vkmc | EmilienM, that was quick, thx | 20:30 |
| mtreinish | EmilienM: that's probably a good segway for me to raise moving those tests into a tempest plugin | 20:31 |
| mtreinish | http://docs.openstack.org/developer/tempest/plugin.html | 20:32 |
| EmilienM | true, I'm concerned about all those projects not using tempest | 20:32 |
| Eva-i | flwang: hello, are you here? | 20:35 |
| flwang | Eva-i: yep, reading the log | 20:36 |
| flwang | EmilienM: so test_messages and test_claim are still failing, right? | 20:37 |
| EmilienM | all the time | 20:37 |
| EmilienM | and only some of them | 20:37 |
| EmilienM | you can see bug report in https://bugs.launchpad.net/zaqar/+bug/1547188/ | 20:38 |
| openstack | Launchpad bug 1547188 in zaqar "some tempest tests are broken" [Undecided,New] | 20:38 |
| EmilienM | flwang: also fyi https://review.openstack.org/282048 | 20:38 |
| flwang | EmilienM: got, i will fix those failing tests asap | 20:40 |
| EmilienM | if you need help for testing let me know | 20:40 |
| EmilienM | I'm putting our CI system in place so it might help you | 20:41 |
| flwang | EmilienM: awesome, thanks | 20:41 |
| flwang | EmilienM: btw, why the puppet-zaqar has to depend on openstacklib? | 20:41 |
| EmilienM | flwang: but I tested puppet-zaqar and it works fine I guess, except if you see something wrong in configuration. | 20:41 |
| EmilienM | flwang: openstacklib is a puppet library we created, like Oslo in OpenStack | 20:41 |
| EmilienM | we have common code in there | 20:42 |
| flwang | oh, ok | 20:42 |
| Eva-i | flwang: remember yesterday we talked about making Zaqar's error responses provide info in bodies? It's possible, however we'll be not able to make all 404 responses from Zaqar provide info - we can't control 404 responses when the user tries unexisting endpoints - this is handled by falcon library. | 20:42 |
| flwang | EmilienM: i got this comment from our operation " xavpaice: flwang1: that zaqar puppet module depends on puppet-openstacklib, which we currently haven't updated due to mass module breakage everywhere else" | 20:43 |
| flwang | Eva-i: really? Even adding the 'description' for the error? | 20:43 |
| EmilienM | flwang: I suggest him/her to jump on IRC / ML and talk with us. | 20:44 |
| flwang | EmilienM: cool | 20:44 |
| EmilienM | flwang: "mass module breakage everywhere else" > usually because you forked our modules for your own needs and now you have hard time to catch-up with upstream. | 20:45 |
| EmilienM | I bet on that | 20:45 |
| Eva-i | flwang: we can add descriptions to the error responses in our endpoints code. If the user tries to make request to the endpoint which does not exist in Zaqar, falcon library automatically sends 404 response. | 20:48 |
| Eva-i | flwang: I mean this: http://paste.openstack.org/show/uOjpajzYIhhZIYhRvZ9C/ | 20:53 |
| flwang | EmilienM: hah, i will forward the words ;) | 20:53 |
| flwang | Eva-i: no | 20:54 |
| flwang | that's not what we want to fix | 20:54 |
| flwang | or i would day it's totally different scenarios | 20:54 |
| flwang | the problem we were discussing is a non-existing pool, flaovr and queue | 20:55 |
| flwang | not a non-existing endpoint, see my point? | 20:55 |
| *** flwang1 has joined #openstack-zaqar | 20:57 | |
| Eva-i | flwang: I think I see your point. You would like to make 404 responses only decriptive for these situations. It's okay that Zaqar sometimes will produce 404 responses without descriptions, right? | 21:01 |
| flwang | Eva-i: yes, and i think the later case you mentioned above, only happens when user access URL directly instead of using zaqar client | 21:02 |
| Eva-i | flwang: yeah | 21:02 |
| Eva-i | flwang: though you only mentioned non-existing pool, flavor and queue. But there are few more situations when python-zaqarclient can receive 404 response code. You can find all situations by searching "falcon.HTTPNotFound" and "falcon.HTTP_404" in our Zaqar repo. | 21:05 |
| Eva-i | *in zaqar.transport.wsgi | 21:06 |
| flwang | oh, yes, i can imagine | 21:07 |
| *** amitgandhinz has quit IRC | 21:17 | |
| *** amitgandhinz has joined #openstack-zaqar | 21:18 | |
| openstackgerrit | Fei Long Wang proposed openstack/zaqar: Fix metadata update issue for v1.1 and v2 https://review.openstack.org/280941 | 21:20 |
| flwang | vkmc: ping | 21:24 |
| vkmc | flwang, pong | 21:26 |
| flwang | vkmc: https://review.openstack.org/#/c/280941/ | 21:26 |
| flwang | i think you can start to work in the websocket part | 21:26 |
| vkmc | flwang, great! | 21:26 |
| vkmc | will do that then | 21:26 |
| flwang | vkmc: awesome, thanks | 21:27 |
| mtreinish | flwang: did you see my comment about moving the zaqar tests into a tempest plugin? I'm trying to make sure we get tempest in-tree tests trimmed down to scope we defined in yvr before the end of the cycle | 21:28 |
| mtreinish | it'd be good to know who was owning it on the zaqar side (so I can pester them :) ) | 21:29 |
| flwang | mtreinish: i don't think there is an owner, TBH | 21:32 |
| flwang | mtreinish: you can blame me :) | 21:33 |
| flwang | mtreinish: seriously, anything i can do to make everyone happy? | 21:33 |
| mtreinish | flwang: heh, no worries. I'm just raising it now really. I can help I'd just like to make sure it's on the project's radar | 21:34 |
| mtreinish | it should be quite simple to do the plugin migration, and I can help with it | 21:34 |
| flwang | mtreinish: awesome, and TBH, personally, i would like to migrate to use tempest as one of gate job | 21:36 |
| flwang | mtreinish: i'm a tempest fans | 21:36 |
| flwang | but you know, we're a small team, we don't have much bandwidth | 21:36 |
| flwang | but it's on my list of Newton | 21:37 |
| mtreinish | sure, I understand that | 21:37 |
| flwang | and if there is any volunteer from tempest side for zaqar, it would be awesome | 21:37 |
| flwang | otherwise, it will happen in Newton early stage | 21:38 |
| *** sriram has quit IRC | 21:48 | |
| openstackgerrit | Fei Long Wang proposed openstack/zaqar: Fix the TTL issue of subscriptions for MongoDB https://review.openstack.org/270464 | 21:49 |
| *** amitgandhinz has quit IRC | 21:59 | |
| Eva-i | flwang: from now on our Rally DSVM gate should pass. Rally team released fix for the bug I reported: https://review.openstack.org/#/c/281935/ | 21:59 |
| flwang | Eva-i: awesome, thanks for working on this | 22:03 |
| openstackgerrit | Matthew Treinish proposed openstack/zaqar: Add zaqar tempest plugin https://review.openstack.org/282074 | 22:06 |
| mtreinish | flwang: ^^^ | 22:06 |
| mtreinish | I haven't tested that at all or anything | 22:06 |
| mtreinish | but if someone could take that to completion I'd appreciate it | 22:06 |
| mtreinish | (including the tempest removal patch once that's really ready) | 22:07 |
| flwang | mtreinish: cool, i will take over, thanks a lot | 22:08 |
| flwang | but i will surely bug you for questions | 22:08 |
| mtreinish | heh, sure | 22:09 |
| Eva-i | flwang: how do you think, is it possible to run python-zaqarclient CLI commands with noauth? | 22:13 |
| Eva-i | flwang: I mean to make the client not get auth token from keystone. | 22:14 |
| flwang | i don't really understand the question | 22:29 |
| flwang | why do you want to do that? | 22:29 |
| flwang | IMHO, noauth is the lowest priority | 22:30 |
| flwang | i don't think any user want to use that | 22:30 |
| Eva-i | flwang: I wonder if I can get 401 response from Zaqar when using CLI | 22:30 |
| Eva-i | flwang: okay | 22:30 |
| flwang | Eva-i: so what's the background? | 22:31 |
| Eva-i | flwang: I just need to decide should I make python-zaqarclient's transport exceptions provide default error messages, so the CLI user will see the error even if Zaqar doesn't provide description in error response, nevermind | 22:38 |
| Eva-i | flwang: seems like no need to make default error messages | 22:39 |
| *** jtomasek has quit IRC | 23:05 | |
| *** openstackgerrit has quit IRC | 23:47 | |
| *** openstackgerrit_ is now known as openstackgerrit | 23:47 | |
| *** openstackgerrit_ has joined #openstack-zaqar | 23:47 | |
| *** openstackgerrit_ is now known as openstackgerrit | 23:48 | |
| *** openstackgerrit_ has joined #openstack-zaqar | 23:48 | |
| *** openstackgerrit_ has quit IRC | 23:55 | |
| *** openstackgerrit_ has joined #openstack-zaqar | 23:56 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!