| *** achanda has quit IRC | 00:01 | |
| *** akanksha has joined #openstack-zaqar | 01:00 | |
| *** itisha has joined #openstack-zaqar | 01:42 | |
| *** achanda has joined #openstack-zaqar | 02:27 | |
| *** kgriffs|afk has quit IRC | 03:27 | |
| *** VeggieMeat has joined #openstack-zaqar | 03:28 | |
| *** VeggieMeat_ has quit IRC | 03:29 | |
| *** kgriffs|afk has joined #openstack-zaqar | 03:30 | |
| *** kgriffs|afk is now known as kgriffs | 03:30 | |
| *** kgriffs has quit IRC | 03:39 | |
| *** nate_gone is now known as njohnston | 03:42 | |
| *** david-lyle has joined #openstack-zaqar | 03:50 | |
| *** achanda has quit IRC | 03:52 | |
| *** kgriffs|afk has joined #openstack-zaqar | 04:04 | |
| *** kgriffs|afk is now known as kgriffs | 04:04 | |
| *** achanda has joined #openstack-zaqar | 04:05 | |
| *** achanda has quit IRC | 04:11 | |
| *** achanda has joined #openstack-zaqar | 04:21 | |
| *** achanda has quit IRC | 04:27 | |
| openstackgerrit | Fei Long Wang proposed openstack/zaqar: Fix duplicated queues on multi pools https://review.openstack.org/221142 | 04:34 |
|---|---|---|
| openstackgerrit | Fei Long Wang proposed openstack/zaqar: Rename pool to pool_group for flavor https://review.openstack.org/220876 | 04:46 |
| *** diga has joined #openstack-zaqar | 04:55 | |
| *** akanksha has quit IRC | 05:08 | |
| *** akanksha has joined #openstack-zaqar | 05:17 | |
| *** mdnadeem has joined #openstack-zaqar | 05:17 | |
| *** itisha has quit IRC | 05:29 | |
| mdnadeem | Hello Team, I am very new to open stack, face some error while starting the Zaqar server | 05:39 |
| mdnadeem | http://paste.openstack.org/show/462427/ | 05:39 |
| mdnadeem | I have follow the below link for installing zaqar: http://docs.openstack.org/developer/zaqar/development-environment.html | 05:40 |
| *** exploreshaifali has joined #openstack-zaqar | 07:05 | |
| *** ryansb_ has joined #openstack-zaqar | 07:05 | |
| *** ryansb_ has quit IRC | 07:05 | |
| *** ryansb_ has joined #openstack-zaqar | 07:05 | |
| *** achanda has joined #openstack-zaqar | 07:06 | |
| *** ryansb has quit IRC | 07:08 | |
| *** ryansb_ is now known as ryansb | 07:08 | |
| *** boris-42 has quit IRC | 07:10 | |
| zigo | flaper87: jasondotstar: I still have the issue that I can't run unit tests of Zaqar, as setting the port for mongodb through system env doesn't work for me. | 07:17 |
| zigo | Fix that, and then I can upload Zaqar *now*.l | 07:17 |
| zigo | :) | 07:17 |
| *** achanda has quit IRC | 07:22 | |
| *** achanda has joined #openstack-zaqar | 07:25 | |
| mdnadeem | Any Idea friends, regarding error : http://paste.openstack.org/show/462427/ | 07:28 |
| mdnadeem | ^^ ? | 07:28 |
| therve | mdnadeem, Try setting pooling to False in the configuration | 07:34 |
| mdnadeem | therve: Thank you for reply | 07:37 |
| mdnadeem | Here is my zakar.conf file: http://paste.openstack.org/show/462501/ | 07:37 |
| mdnadeem | therve: Please have a look | 07:38 |
| mdnadeem | I think pooling is by default set to False | 07:40 |
| *** achanda has quit IRC | 07:41 | |
| therve | Ah you're right, it should be "unreliable" that you need to set to False | 07:44 |
| mdnadeem | therve:I have doubt that on installation of mongodb, | 07:47 |
| mdnadeem | when i run command sudo service mongod status --> output shows mongod running | 07:47 |
| mdnadeem | however when i execute command sudo service mongodb status --> output shows mongod stop/waiting | 07:48 |
| mdnadeem | any idea? about mongod and mongodb processes? | 07:49 |
| therve | only mongod matters | 07:50 |
| mdnadeem | ok, that means it installed successfully. | 07:51 |
| *** mdnadeem has quit IRC | 07:56 | |
| *** mdnadeem has joined #openstack-zaqar | 08:10 | |
| flaper87 | zigo: mmh, weird. I thought therve had implemented that | 08:24 |
| therve | I did | 08:28 |
| therve | I thought last time we looked unit tests didn't work at all for zigo | 08:29 |
| therve | We never got into the real issues | 08:29 |
| therve | Also it was kilo | 08:29 |
| zigo | It didn't indeed. | 08:29 |
| zigo | I had a look in Liberty, same problem. | 08:29 |
| zigo | It can't connect to Mongo. | 08:29 |
| exploreshaifali | flaper87, I am trying to implement "flavor update" cli | 08:31 |
| exploreshaifali | but for that first I will need to fetch the flavor which needs to be updated | 08:31 |
| exploreshaifali | can you help in that | 08:32 |
| exploreshaifali | ? | 08:32 |
| therve | zigo, If you have some time to debug it we can look at it | 08:32 |
| therve | It did work when I committed the patch | 08:32 |
| zigo | therve: How may I help? Do you need the build log? | 08:34 |
| flaper87 | therve: zigo it'd be really cool if you can figure this out sooner rather than later :D | 08:34 |
| flaper87 | thanks to both | 08:34 |
| therve | zigo, Sure that'd be a start | 08:34 |
| flaper87 | exploreshaifali: fetching the flavor is actually quite simple, just create an instance of flavor.Flavor() | 08:35 |
| zigo | http://sid.gplhost.com/zaqar/zaqar_1.0.0~b3-1_amd64.build | 08:35 |
| zigo | There you are. | 08:35 |
| exploreshaifali | flaper87, does glance or heat does not use "cliff" library to implement cl commands? | 08:36 |
| zigo | The thing is, it does connect... | 08:36 |
| exploreshaifali | I tried to look at their clients | 08:36 |
| zigo | It says "connection closed" not "connection refused"> | 08:36 |
| therve | zigo, What's pymongo version? | 08:37 |
| zigo | therve: 3.0.3-1 | 08:38 |
| zigo | (the latest in Sid) | 08:38 |
| therve | zigo, And mongodb? | 08:38 |
| zigo | 2.4.14 | 08:38 |
| zigo | Could it be that I didn't create the db prior starting the tests? | 08:39 |
| zigo | Or does the testing framework do it for me? | 08:39 |
| therve | It should create it | 08:39 |
| therve | 2.4 is a bit on the old side | 08:39 |
| flaper87 | exploreshaifali: glance doesn't use python-openstackclient so no, they don't use cliff | 08:40 |
| zigo | Right, it should be at least one week old, so therefore, old ... :P | 08:40 |
| therve | We do have some regressions since I made the patch though | 08:40 |
| flaper87 | we use cliff becasue we use python-openstackclient | 08:40 |
| flaper87 | zigo: therve I haven't read the backlog but I wanted to say we should merge whatever fix will make building the package possible | 08:40 |
| therve | zigo, Our builds use 2.6 I believe | 08:40 |
| exploreshaifali | flaper87, okay! Thanks :) | 08:40 |
| flaper87 | just wanted to say it's also high-prio | 08:40 |
| flaper87 | and let me know if I can help | 08:41 |
| therve | And last release is 3.0 | 08:41 |
| * flaper87 goes back to emails | 08:41 | |
| therve | So yeah old | 08:41 |
| zigo | https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=748490 | 08:41 |
| openstack | Debian bug 748490 in mongodb "mongodb: New upstream version is available (2.6.1)" [Normal,Open] | 08:41 |
| zigo | No answer from Laszlo, the package maintainer. | 08:42 |
| zigo | IMO, Laszlo does a way too many things in Debian, so he can't take care of it well enough. | 08:42 |
| * zigo sends a ping in the bug report | 08:42 | |
| mdnadeem | therve: Now i was trying to install zaqar using devstack, it shows me error : "Error: couldn't connect to server 127.0.0.1:27017 at src/mongo/shell/mongo.js:145". Any Idea sir? | 08:43 |
| zigo | Anyway, supporting mongodb 2.4 would be nice, so we don't have to rely on a backport. | 08:44 |
| therve | mdnadeem, Not really. Verify mongo is running properly, check listening port | 08:44 |
| zigo | And also, I don't think that's the issue we're having here. | 08:44 |
| therve | zigo, Possibly not | 08:44 |
| flaper87 | zigo: supporting mongodb 2.4 might be possible except that probably the version of pymongo we use can't talk to it (but I'm prob wrong about this) | 08:45 |
| flaper87 | the thing we care about is ttl collections and we have those in mongodb 2.4 | 08:46 |
| flaper87 | FYI | 08:46 |
| mdnadeem | therve: ok, Thanks | 08:46 |
| zigo | Ok, so what's the way forward? | 08:48 |
| zigo | What can I do / test? | 08:48 |
| zigo | FYI, I'm doing this to launch tests: PYTHONPATH=$(CURDIR) PYTHON=python$$i bash -x ./debian/setup-test-env-mongodb.sh testr-python$$PYMAJOR run --subunit | 08:49 |
| zigo | and the wrapper is here: http://anonscm.debian.org/cgit/openstack/zaqar.git/tree/debian/setup-test-env-mongodb.sh?h=debian/liberty | 08:49 |
| zigo | The wrapper basically starts mongo on a non-standard port, and do export ZAQAR_TEST_MONGODB_URL=mongodb://localhost:${MONGO_PORT} | 08:50 |
| therve | zigo, We do have a regression | 08:52 |
| therve | So... wait for the fix I guess | 08:52 |
| therve | flaper87, We need to have a test build setting the URL | 08:55 |
| flaper87 | therve: as in the setUp phase ? | 09:01 |
| therve | flaper87, As in jenkins | 09:01 |
| flaper87 | therve: mmh, devkstack is pointing to localhost, isn't it ? | 09:02 |
| therve | flaper87, devstack is always using the standard 27017 port. We should change mongo port so that it doesn't break every other patch | 09:03 |
| *** exploreshaifali has quit IRC | 09:04 | |
| flaper87 | therve: mmh, I'm a bit lost, tbh. | 09:06 |
| flaper87 | What's the current issue? | 09:06 |
| flaper87 | is 2017 causing issues? | 09:06 |
| therve | flaper87, If mongodb is not running on that port, yes | 09:08 |
| flaper87 | therve: but why wouldnt mongodb be running on that port? | 09:11 |
| flaper87 | does the gate change it? | 09:11 |
| therve | flaper87, Debian builds do | 09:11 |
| therve | That's the whole problem :) | 09:11 |
| flaper87 | ah yeah but you said that we needed to do it in Jenkins and I thought you were talking about Zaqar's gate | 09:11 |
| * flaper87 confused | 09:11 | |
| flaper87 | or you want to use the env var in the gate to make sure it works | 09:12 |
| flaper87 | ? | 09:12 |
| flaper87 | sorry, I feel dumb | 09:12 |
| flaper87 | hahaha | 09:12 |
| therve | flaper87, Yes, that one | 09:12 |
| therve | I want to use the env var in the gate | 09:12 |
| flaper87 | therve: coolio, now I got it | 09:13 |
| flaper87 | sounds great to me | 09:13 |
| flaper87 | I guess you can just hack gate_hook.sh to set it | 09:13 |
| flaper87 | and the devstack plugin | 09:13 |
| therve | Yeah | 09:15 |
| therve | Fixing the tests first | 09:15 |
| * flaper87 gives therve a morning beer | 09:16 | |
| therve | Not sure it'll do me any good :) | 09:17 |
| therve | That would be less painful if the timeout was so long | 09:17 |
| therve | Urggg | 09:17 |
| zigo | flaper87: therve: I'd suggest that you copy my wrapper script to start mongodb on a non-standard port. That's what Ceilometer does (even before I started doing so in the packages). | 09:26 |
| *** diga has quit IRC | 09:26 | |
| therve | I don't know what I fixed last time but I couldn't possibly left that many tests over :/ | 09:39 |
| therve | flaper87, https://github.com/openstack/zaqar/blob/master/zaqar/tests/unit/storage/test_pool_catalog.py#L52 Good lord what can I do with that | 09:39 |
| flaper87 | therve: replace it with whatever the value of the env var is | 09:43 |
| flaper87 | one thing to keep in mind is that you can't create 2 pools w/ the same URL | 09:43 |
| flaper87 | :D | 09:43 |
| therve | flaper87, The localhost/127.0.0.1 hack is particularly nasty | 09:44 |
| therve | Also it doesn't work when try to do that | 09:44 |
| flaper87 | therve: :D | 09:46 |
| *** bradjones|away has quit IRC | 09:47 | |
| flaper87 | It's nasty and it's part of the technical debt we have in Zaqar's unittests | 09:47 |
| flaper87 | :( | 09:47 |
| *** bradjones has joined #openstack-zaqar | 09:49 | |
| *** bradjones has joined #openstack-zaqar | 09:49 | |
| *** flwang1 has joined #openstack-zaqar | 10:14 | |
| flwang1 | morning | 10:14 |
| *** flwang has quit IRC | 10:31 | |
| *** flwang has joined #openstack-zaqar | 10:32 | |
| *** diga has joined #openstack-zaqar | 10:34 | |
| flwang1 | vkmc: ping | 10:40 |
| *** diga has quit IRC | 10:40 | |
| flaper87 | flwang1:morning :D | 10:47 |
| *** akanksha has quit IRC | 10:48 | |
| flwang1 | flaper87: https://review.openstack.org/221122 need your bless | 10:48 |
| flwang1 | and this one https://review.openstack.org/221142 | 10:49 |
| * flaper87 clicks | 10:49 | |
| flwang1 | we're still called 'Queue service' on http://docs.openstack.org/developer/openstack-projects.html | 10:56 |
| *** akanksha has joined #openstack-zaqar | 10:57 | |
| flaper87 | flwang1: that's fixable :d | 11:02 |
| flaper87 | :D | 11:02 |
| flaper87 | flwang1: https://review.openstack.org/#/c/221142/2/zaqar/storage/pooling.py,cm | 11:06 |
| flaper87 | mmh, not entirely sure that's correct | 11:06 |
| flwang1 | any question? | 11:07 |
| flaper87 | I don't even think we need that controller at all | 11:07 |
| flaper87 | since we don't use pooling for the management plane | 11:07 |
| flaper87 | The queue controller is just 1 and it should point to the management storage | 11:07 |
| flwang1 | flaper87: yes, i can see your point | 11:07 |
| flaper87 | rather than using pooling to store the data | 11:07 |
| flwang1 | flaper87: but that's a big change | 11:07 |
| flwang1 | flaper87: it would be more safe to do that in M | 11:08 |
| flwang1 | how do you think | 11:08 |
| flaper87 | sure, although I still think there's something weird there | 11:08 |
| flaper87 | why don't you need to go through the whole catalogue ? | 11:09 |
| flwang1 | because any queue controller can get the queue list | 11:10 |
| flwang1 | if we iterate all pools, and merge the queues list, we will get duplicated queues | 11:10 |
| flwang1 | is that your question? | 11:10 |
| flaper87 | ish. I'm wondering why that's the case. | 11:12 |
| flaper87 | a queue should exist in a single pool | 11:12 |
| flaper87 | ah because the queue controller is the same | 11:13 |
| flwang1 | yep | 11:13 |
| flaper87 | mmh, that's just complicated | 11:13 |
| flaper87 | could you write a comment there? | 11:13 |
| flaper87 | and file a bug | 11:13 |
| flaper87 | this needs to be refactored | 11:14 |
| flwang1 | you mean add a TODO comment and ref a bug? | 11:14 |
| flaper87 | flwang1: here's another idea | 11:14 |
| flwang1 | listening... | 11:15 |
| flaper87 | Can we have a `_queue_ctrl` attribute in the _pool_catalog that has the instance of the queue_controller ? | 11:15 |
| flaper87 | that way we don't need to query the pool's table to then query the queue's one | 11:16 |
| flaper87 | in the Catalog init you coud do: self._queue_ctrl = control.queue_controller | 11:16 |
| flwang1 | here https://github.com/openstack/zaqar/blob/master/zaqar/storage/pooling.py#L418 ? | 11:18 |
| flwang1 | i think it works | 11:18 |
| flaper87 | flwang1: yup | 11:19 |
| flwang1 | ok, sounds a good idea. will post a new patch set | 11:19 |
| flaper87 | flwang1: thanks | 11:20 |
| flwang1 | flaper87: in your non-candidacy mail, you mentioned some items we should focus, cross-project integration, api, more use case | 11:22 |
| flwang1 | would you like to discuss it a bit ? | 11:22 |
| flwang1 | if you're not really hungry to run for lunch | 11:23 |
| flaper87 | flwang1: I already had lunch, lets talk | 11:31 |
| flwang1 | flaper87: cool | 11:31 |
| flwang1 | for cross project, i think we need to help get the swift middleware in and push websocket of zaqar for horizon, guest agent comm via zaqar for sahara/trove, anything else? | 11:32 |
| flaper87 | Those are the main topics we've discussed in the past | 11:39 |
| flaper87 | flwang1: However, If I were to choose, I'd focus on helping Sahara and Horizon | 11:40 |
| flaper87 | but normally, it comes down to other projects also willing to do the same | 11:40 |
| flwang1 | flaper87: what's the current sahara's attitude for the idea? | 11:40 |
| *** dynarro has joined #openstack-zaqar | 11:40 | |
| flwang1 | i know sahara and trove have the critical security hole, so it would be nice if we can work it out by zaqar | 11:41 |
| flaper87 | flwang1: Sahara is evaluating to move towards a guest-agent approach and if they do, they are willing to use Zaqar | 11:42 |
| flaper87 | Trove, on the other hand, is thinking to move away from guest-agent and use ssh (which is what sahara is using now)... | 11:43 |
| flaper87 | but I'm not sure where that discussion is going | 11:43 |
| flwang1 | flaper87: is there any people from sahara we can contact? | 11:43 |
| flaper87 | In vancouver we talked w/ folks from trove and they weren't very happy w/ the idea of using Zaqar | 11:44 |
| flaper87 | flwang1: I'm working close to the Sahara team on this, I'm happy to coordinate this work | 11:44 |
| flaper87 | What we don't have any holder for is Horizon | 11:44 |
| flwang1 | flaper87: it would be nice if you can continually follow the sahara integration | 11:45 |
| flaper87 | I will, that's a use-case I'm very excited about | 11:45 |
| flwang1 | as for horizon, IIRc, vkmc said she will follow up that, right? | 11:45 |
| flwang1 | me too | 11:45 |
| flwang1 | ok, let's move on to next topic, API | 11:46 |
| flaper87 | flwang1: yup, she said that. You should ping her again to check what her time schedule looks like on that front | 11:46 |
| flwang1 | flaper87: ok, i will check with her | 11:46 |
| flwang1 | we have released 3 versions, and i would say v1.1 is the stable one, but we have many exciting features in v2 | 11:47 |
| flwang1 | so what's our recommended api version if there is a potential deployer? | 11:48 |
| flaper87 | flwang1: fwiw, v1.1 is "supported" and v2 is "current" | 11:49 |
| flaper87 | so, use v2 | 11:50 |
| flwang1 | flaper87: cool | 11:52 |
| flwang1 | flaper87: do we have an official place to announce this? | 11:52 |
| flwang1 | given we're not on the api-ref site | 11:53 |
| flaper87 | flwang1: documentation would be the best place | 11:53 |
| flaper87 | Change the README, change devs docs | 11:53 |
| flaper87 | Put it in the main page | 11:53 |
| flaper87 | etc | 11:53 |
| flaper87 | and, obviously, have support for it in the client | 11:53 |
| flaper87 | which we don't have | 11:53 |
| flwang1 | flaper87: hmm... it's shame | 11:54 |
| flwang1 | i will pay more effort on client side in M | 11:54 |
| flaper87 | flwang1: yup, as I mentioned in the meeting yday. I will as well | 11:54 |
| flwang1 | flaper87: awesome | 11:55 |
| flaper87 | I'd like to work on extending the functional tests and support for v2 | 11:55 |
| flaper87 | dynarro: did a lot of work there | 11:55 |
| flaper87 | and her patches need to be reviewed | 11:55 |
| flaper87 | flwang1: dynarro will be in Tokyo | 11:55 |
| flwang1 | flaper87: awesome | 11:55 |
| flwang1 | i just booked my flights yesterday | 11:55 |
| flwang1 | flaper87: should we take microversion into consideration in M? | 11:56 |
| flaper87 | flwang1: I'd like us to start discussing it, yes. | 11:57 |
| flaper87 | I don't think I understand them well enough | 11:57 |
| flaper87 | but it'd be great to be prepared so that whenever we'll need them, we'll be ready to use them | 11:57 |
| flwang1 | flaper87: good point | 11:58 |
| flaper87 | I don't think we need a new API | 11:58 |
| flwang1 | i just read some doc of nova's microversion | 11:58 |
| flwang1 | me too | 11:58 |
| flaper87 | or at least, I don't think there are required, backwards incompatible, changes to do | 11:58 |
| flaper87 | Additions can still be under v2 | 11:59 |
| flaper87 | brb | 11:59 |
| flwang1 | just ping me when you back | 12:02 |
| openstackgerrit | Thomas Herve proposed openstack/zaqar: Fix usage of mongo test URL https://review.openstack.org/223553 | 12:03 |
| therve | zigo, OK! Sorry about that. The above patch should fix things again, hopefully | 12:05 |
| flwang1 | therve: morning | 12:06 |
| therve | hi flwang1 | 12:07 |
| openstackgerrit | Thomas Herve proposed openstack/zaqar: Fix usage of mongo test URL https://review.openstack.org/223553 | 12:08 |
| flwang1 | hah, i'm going to ask why use 27019 ;) | 12:08 |
| zigo | therve: Thanks, trying it now. | 12:08 |
| flwang1 | and i saw there is a new patchset | 12:08 |
| therve | zigo, Awesome thanks | 12:09 |
| flaper87 | flwang1: back | 12:10 |
| flwang1 | flaper87: as for the more user case you mentioned in your mail, what does that mean? | 12:11 |
| flwang1 | therve: thanks for the patch, i'm thinking if we should do the same thing for redis | 12:12 |
| flaper87 | flwang1: i'll have to re-read that email | 12:13 |
| flaper87 | ahhahaha | 12:13 |
| therve | flwang1, Possibly. I never run them | 12:14 |
| flwang1 | therve: ok, i will take a look | 12:14 |
| flaper87 | therve: I was hoping that config var would be more like: ZAQAR_STORAGE_URL | 12:15 |
| flaper87 | ZAQAR_TEST_STORAGE_URL | 12:15 |
| zigo | therve: http://paste.openstack.org/show/462886/ <=== Not working ... :( | 12:15 |
| flaper87 | BUT | 12:15 |
| flaper87 | it can't be because redis doesn't have support for the management plane | 12:15 |
| therve | flaper87, Good I'm here to crush your hopes | 12:15 |
| zigo | therve: Is this because I need trunk? | 12:15 |
| flaper87 | therve: :P | 12:15 |
| therve | zigo, Yeah | 12:16 |
| zigo | therve: Ok, will try to do that then. | 12:16 |
| therve | flaper87, I don't know, I'd like to be able to run both mongo and redis tests in one run | 12:16 |
| therve | If I have a generic var I can't | 12:17 |
| flwang1 | flaper87: what's our plan for redis? i mean, is it really necessary to support redis for management plane | 12:17 |
| flaper87 | therve: yup and we can't use redis for management, therefore we need a specific var | 12:17 |
| flaper87 | flwang1: no no | 12:17 |
| flaper87 | flwang1: the lack of proper durability in redis makes it not suitable for the management plane | 12:18 |
| flwang1 | flaper87: cool cool | 12:18 |
| flwang1 | flaper87: btw, i have a question as for the client | 12:18 |
| flwang1 | after install the client, with openstack client, i can only see 'queue' as the sub command | 12:19 |
| flaper87 | flwang1: shoot | 12:19 |
| flwang1 | am i missing anything? | 12:19 |
| flaper87 | flwang1: no, zaqarclient CLI is missing something, code. | 12:20 |
| flaper87 | :D | 12:20 |
| flaper87 | That's another thing we should change: we're currently using `queue` as subcommand but I think we shouldn't | 12:20 |
| flaper87 | we should either use messaging or something else | 12:20 |
| flwang1 | flaper87: yes | 12:20 |
| zigo | Much better from trunk indeed... :P | 12:20 |
| zigo | It seems to be still failing though. | 12:22 |
| flaper87 | zigo: does it work? | 12:22 |
| *** malini has joined #openstack-zaqar | 12:22 | |
| flaper87 | please please, tell me it works | 12:22 |
| zigo | It's going to take like forever to have it to finish to run all tests and give me some results. | 12:22 |
| flaper87 | :D | 12:22 |
| flaper87 | :( | 12:22 |
| flwang1 | flaper87: is that on dynarro's to-do list? | 12:22 |
| flaper87 | flwang1: no, dynarro is working on the v2 code for the library | 12:23 |
| flaper87 | flwang1: exploreshaifali is working on the CLI | 12:23 |
| flaper87 | flwang1: and diga | 12:23 |
| flaper87 | but he's been away lately | 12:23 |
| flwang1 | flaper87: ok, cool, i will sync with them | 12:24 |
| flwang1 | i know the puppet status, so i'm good now | 12:24 |
| flaper87 | man, this patch is driving me crazy: https://review.openstack.org/#/c/178955/ | 12:24 |
| flwang1 | flaper87: thanks for your time, man | 12:24 |
| flaper87 | :P | 12:24 |
| flaper87 | flwang1: w000h000 | 12:24 |
| flaper87 | flwang1: anytime | 12:25 |
| flwang1 | flaper87: more patches on client side will be coming | 12:25 |
| openstackgerrit | Ryan Brown proposed openstack/zaqar: Fix wrong query for pools when creating queue https://review.openstack.org/220849 | 12:35 |
| * flaper87 fixed devstack's patch | 12:41 | |
| *** malini has quit IRC | 12:46 | |
| *** Qiming has joined #openstack-zaqar | 12:50 | |
| *** Qiming has quit IRC | 13:07 | |
| *** Qiming has joined #openstack-zaqar | 13:07 | |
| *** Venkat has joined #openstack-zaqar | 13:18 | |
| openstackgerrit | Flavio Percoco proposed openstack/zaqar: Allow running zaqarclient for gate tests https://review.openstack.org/221068 | 13:20 |
| openstackgerrit | Flavio Percoco proposed openstack/zaqar: Allow running zaqarclient for gate tests https://review.openstack.org/221068 | 13:20 |
| flaper87 | ah wait, I know what's happening | 13:20 |
| *** sriram has joined #openstack-zaqar | 13:21 | |
| openstackgerrit | Flavio Percoco proposed openstack/zaqar: Allow running zaqarclient for gate tests https://review.openstack.org/221068 | 13:22 |
| therve | flwang1, So redis tests seem to be broken | 13:25 |
| therve | And some of them because subscriptions are broken I believe | 13:25 |
| openstackgerrit | Thomas Herve proposed openstack/zaqar: Fix usage of mongo test URL https://review.openstack.org/223553 | 13:27 |
| Venkat | Hello.. I am trying to configure Redis in Zaqar and test it with basic APIs like create, list, delete | 13:28 |
| Venkat | is there anyone can help me - how to configure Redis in zaqar.conf ? | 13:28 |
| flaper87 | Venkat: hey | 13:30 |
| flaper87 | Venkat: you need to use redis in for the message_store and sqlalchemy for the management store | 13:30 |
| flaper87 | unlike the mongodb storage, redis doesn't have support for the management API | 13:30 |
| flaper87 | therve: what tests are broken? | 13:30 |
| therve | flaper87, the subscription ones at least. A bunch of others seem easily fixable | 13:31 |
| therve | There are not that many anyway | 13:31 |
| *** Venkat has quit IRC | 13:32 | |
| *** Venkat has joined #openstack-zaqar | 13:36 | |
| *** akanksha has quit IRC | 13:38 | |
| Venkat | currently I configured MongoDB and verified | 13:38 |
| Venkat | How I can configure redis or sqlalchemy as backend in zaqar conf without mongodb | 13:39 |
| Venkat | any inputs ? | 13:45 |
| therve | Venkat, https://github.com/openstack/zaqar/blob/master/devstack/plugin.sh#L156 shows how we configure it in devstack | 13:54 |
| therve | It ought to work | 13:54 |
| therve | flaper87, I don't know much about redis, but I wonder why we're using sorted sets when the score is always 1 | 13:54 |
| ryansb | therve: we are? | 13:55 |
| therve | ryansb, That's how I understand it | 13:55 |
| ryansb | for what, messages? | 13:56 |
| therve | messages, queues, subscriptions... | 13:57 |
| ryansb | because the messages sorted set actually changes the score | 13:57 |
| ryansb | it's how we guarantee fifo, IIRC | 13:57 |
| therve | Add claims actually | 13:57 |
| therve | ryansb, Can you show me where it's done? | 13:58 |
| * therve tries to understand | 13:58 | |
| therve | Ah | 14:00 |
| therve | ryansb, Is that the lua script? | 14:01 |
| ryansb | uh, lemme check | 14:01 |
| therve | That would make sense | 14:01 |
| ryansb | yeah, it's in zaqar/storage/redis/scripts/index_messages.lua | 14:02 |
| ryansb | it gets the counter and sets that as the rank | 14:02 |
| therve | OK that one makes sense | 14:02 |
| therve | Not for the others though | 14:02 |
| Venkat | therve, if redis is enabled as zaqar-backend then why drivers:management_store:sqlalchemy is configured with sqlite:// ? | 14:05 |
| ryansb | hm, not sure then. That's a flavio question. | 14:05 |
| ryansb | Venkat: the reason is because redis is only a datastore backend | 14:06 |
| therve | What Ryan said | 14:06 |
| ryansb | sqlalchemy is still used to handle othe bookkeeping (flavor names, etc etc) | 14:06 |
| Venkat | Oh ok.. so redis is used as message_store and sqlalchemy for mgmt_store | 14:07 |
| therve | yep | 14:07 |
| Venkat | thanks therve and ryansb for your inputs here.. | 14:07 |
| Venkat | One more question - which one is best to use with good performance ? MongoDB or Redis-with-SQLAlchemy ? | 14:10 |
| ryansb | I don't know the answer to that one, sorry | 14:14 |
| *** mpanetta_ has joined #openstack-zaqar | 14:15 | |
| Venkat | any zaqar benchmark results available ? just want to know the performance results etc | 14:16 |
| flaper87 | Venkat: https://wiki.openstack.org/wiki/Zaqar/Performance | 14:16 |
| flaper87 | but those are a bit old | 14:16 |
| *** malini has joined #openstack-zaqar | 14:17 | |
| *** exploreshaifali has joined #openstack-zaqar | 14:17 | |
| *** david-lyle has quit IRC | 14:20 | |
| *** Venkat has quit IRC | 14:20 | |
| *** Nesze has joined #openstack-zaqar | 14:22 | |
| *** Venkat has joined #openstack-zaqar | 14:23 | |
| Venkat | Thanks flaper | 14:24 |
| *** Venkat has quit IRC | 14:27 | |
| *** achanda has joined #openstack-zaqar | 14:30 | |
| *** achanda has quit IRC | 14:31 | |
| Nesze | Hello everyone, I’d have a question. From the API doc: “The size of this body, in characters and including whitespace must be <= 64 KiB (configurable)”. Does configurable means, that there is no upper limit for message size? My use case would be to deal with occasional messages above a few mega bytes (max 16 MB). Can zaqar deal with that? | 14:32 |
| Nesze | Ehh, the quote is from the “Set Queue Metadata” section. My question is still valid though : ) . | 14:33 |
| vkmc | hi Nesze! | 14:34 |
| vkmc | that value is configurable | 14:34 |
| vkmc | metadata_size_uplimit is for metadata, which as you said, the default maximum is 64kib | 14:34 |
| vkmc | message_size_uplimit is for messages, the default is 256kib | 14:35 |
| vkmc | that is for v1 though, let me check if that still holds for v1.1/v2.0 | 14:36 |
| Nesze | thanks vkmc! doh, I wasn’t attentive enough, it’s right in the docs… | 14:41 |
| vkmc | no, it's ok! | 14:42 |
| vkmc | though seems to have changed... so you did right in asking | 14:42 |
| vkmc | Nesze, the configuration for message size is max_messages_post_size | 14:47 |
| vkmc | :) I knew it has changed | 14:47 |
| vkmc | and the default is 256kib as we did in v1 | 14:48 |
| vkmc | let us know if we can give you a hand with something | 14:48 |
| exploreshaifali | flaper87, vkmc is there any other openstack component which use "python-openstackclient"? | 14:58 |
| ryansb | heat is working on it | 14:58 |
| exploreshaifali | so that I can look at example of how to implement update functionality | 14:58 |
| ryansb | https://review.openstack.org/#/c/195867/ | 14:58 |
| vkmc | ehm, IIRC currently most of the openstack components are using the openstackclient | 14:58 |
| *** achanda has joined #openstack-zaqar | 14:58 | |
| ryansb | keystone uses openstackclient almost exclusively, I think they're deprecating their CLI soonish | 14:59 |
| *** Qiming has quit IRC | 14:59 | |
| exploreshaifali | ryansb, yes keystone use it but not "cliff" | 15:00 |
| ryansb | Hm? | 15:00 |
| vkmc | ah well | 15:00 |
| exploreshaifali | basically I want example to implement update command using cliff | 15:00 |
| vkmc | yes, we use that library | 15:00 |
| exploreshaifali | for update flavor I believe I will need to implement two commands: | 15:00 |
| exploreshaifali | openstack flavor update --name <old_name> <new_name> | 15:01 |
| vkmc | http://docs.openstack.org/developer/cliff/ | 15:01 |
| exploreshaifali | and one for pool | 15:01 |
| exploreshaifali | yes I looked at it | 15:02 |
| exploreshaifali | and I think command.Command is something will need for update also | 15:02 |
| Nesze | thanks, much appreciated, so it’s configurable, and there is no upper limit. that’s what I was mainly curious about | 15:02 |
| flaper87 | exploreshaifali: keystone, nova | 15:03 |
| flaper87 | exploreshaifali: there are others | 15:03 |
| exploreshaifali | keystone does not use cliff | 15:04 |
| flaper87 | exploreshaifali: the use python-openstackclient, therefore they use cliff | 15:05 |
| exploreshaifali | okay | 15:05 |
| flaper87 | thing is the code in their codebase is the old CLI, whereas the new one lives in the python-openstackclient repo | 15:05 |
| * exploreshaifali searching more | 15:06 | |
| *** achanda has quit IRC | 15:06 | |
| exploreshaifali | Thanks guys :) | 15:06 |
| *** csoukup has joined #openstack-zaqar | 15:07 | |
| flaper87 | exploreshaifali: anytime :D | 15:08 |
| *** david-lyle has joined #openstack-zaqar | 15:14 | |
| *** Nesze has quit IRC | 15:16 | |
| openstackgerrit | Doraly Navarro proposed openstack/python-zaqarclient: Added v1 folder https://review.openstack.org/216159 | 15:28 |
| *** exploreshaifali has quit IRC | 15:32 | |
| openstackgerrit | Doraly Navarro proposed openstack/python-zaqarclient: Added support to claims https://review.openstack.org/213165 | 15:32 |
| openstackgerrit | Doraly Navarro proposed openstack/python-zaqarclient: Added support to notifications https://review.openstack.org/209238 | 15:32 |
| openstackgerrit | Doraly Navarro proposed openstack/python-zaqarclient: Added v1 folder https://review.openstack.org/216159 | 15:32 |
| *** boris-42 has joined #openstack-zaqar | 15:35 | |
| openstackgerrit | Thomas Herve proposed openstack/zaqar: Fix redis tests https://review.openstack.org/223656 | 15:41 |
| therve | flaper87, ^^^ this seems to fix tests for me | 15:41 |
| therve | Another builder to add... | 15:41 |
| openstackgerrit | Thomas Herve proposed openstack/zaqar: Fix redis tests https://review.openstack.org/223656 | 15:53 |
| openstackgerrit | Doraly Navarro proposed openstack/python-zaqarclient: Added support to claims https://review.openstack.org/213165 | 16:17 |
| openstackgerrit | Doraly Navarro proposed openstack/python-zaqarclient: Added support to v2 https://review.openstack.org/209238 | 16:32 |
| *** amitgandhinz has joined #openstack-zaqar | 16:45 | |
| *** fifieldt_ has quit IRC | 16:49 | |
| *** dynarro has quit IRC | 16:52 | |
| *** malini has quit IRC | 16:56 | |
| *** achanda has joined #openstack-zaqar | 17:28 | |
| *** malini has joined #openstack-zaqar | 17:32 | |
| *** amitgandhinz has quit IRC | 17:37 | |
| *** amitgandhinz has joined #openstack-zaqar | 17:37 | |
| *** malini has quit IRC | 17:51 | |
| *** malini has joined #openstack-zaqar | 18:02 | |
| *** exploreshaifali has joined #openstack-zaqar | 18:20 | |
| *** exploreshaifali has quit IRC | 18:45 | |
| *** malini has quit IRC | 19:25 | |
| *** AAzza has quit IRC | 19:30 | |
| *** AAzza has joined #openstack-zaqar | 19:31 | |
| *** AAzza has quit IRC | 19:35 | |
| *** AAzza has joined #openstack-zaqar | 19:36 | |
| *** AAzza has quit IRC | 19:40 | |
| *** kgriffs is now known as kgriffs|afk | 19:44 | |
| *** AAzza has joined #openstack-zaqar | 19:46 | |
| *** flwang1 has quit IRC | 19:48 | |
| *** kgriffs|afk is now known as kgriffs | 19:49 | |
| *** kgriffs is now known as kgriffs|afk | 19:58 | |
| *** achanda has quit IRC | 20:01 | |
| *** kgriffs|afk is now known as kgriffs | 20:08 | |
| *** malini has joined #openstack-zaqar | 20:33 | |
| *** malini has quit IRC | 20:33 | |
| *** achanda has joined #openstack-zaqar | 20:35 | |
| *** achanda has quit IRC | 20:35 | |
| *** achanda has joined #openstack-zaqar | 20:35 | |
| *** kgriffs is now known as kgriffs|afk | 20:44 | |
| *** kgriffs|afk is now known as kgriffs | 20:45 | |
| *** malini has joined #openstack-zaqar | 20:57 | |
| *** malini has quit IRC | 20:58 | |
| *** flwang1 has joined #openstack-zaqar | 21:02 | |
| *** malini has joined #openstack-zaqar | 21:04 | |
| *** csoukup has quit IRC | 21:07 | |
| *** sriram has quit IRC | 21:19 | |
| *** mpanetta_ has quit IRC | 21:25 | |
| *** flwang1 has left #openstack-zaqar | 21:36 | |
| *** malini has quit IRC | 21:40 | |
| *** njohnston is now known as nate_gone | 21:47 | |
| *** amitgandhinz has quit IRC | 21:47 | |
| *** malini has joined #openstack-zaqar | 21:51 | |
| *** malini has quit IRC | 21:53 | |
| *** csoukup has joined #openstack-zaqar | 21:56 | |
| *** csoukup has quit IRC | 21:57 | |
| flwang | anybody has the ATC code but can't go Tokyo? | 21:59 |
| vkmc | flwang, pm | 22:01 |
| *** Qiming has joined #openstack-zaqar | 22:57 | |
| *** kgriffs is now known as kgriffs|afk | 23:02 | |
| *** openstackgerrit has quit IRC | 23:16 | |
| *** openstackgerrit has joined #openstack-zaqar | 23:16 | |
| *** david-lyle has quit IRC | 23:17 | |
| *** Qiming has quit IRC | 23:52 | |
| *** achanda_ has joined #openstack-zaqar | 23:55 | |
| *** achanda_ has quit IRC | 23:57 | |
| *** achanda has quit IRC | 23:58 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!