| flwang1 | Eva-i: haha, ok, then I will give you some bugs as lunch or dinner | 00:10 |
|---|---|---|
| *** pradk has quit IRC | 00:14 | |
| flwang1 | kgriffs: any lucky you around? | 00:35 |
| kgriffs | whats up? | 00:36 |
| *** achanda has joined #openstack-zaqar | 00:41 | |
| flwang1 | kgriffs: oh, you're around. i'm trying to figure out if there is a good way to get the api request stats of zaqar | 00:42 |
| flwang1 | i know Rackspace is using EOM | 00:42 |
| flwang1 | but EOM doesn't work with keystone | 00:42 |
| flwang1 | so i'm wondering if if there is a way to do that easily | 00:42 |
| kgriffs | can you explain a little more about what you mean by "api request stats"? | 00:43 |
| *** achanda has quit IRC | 00:43 | |
| *** achanda has joined #openstack-zaqar | 00:43 | |
| *** achanda has quit IRC | 00:44 | |
| flwang1 | kgriffs: you know, SQS and Rackspace are charging the msg service by request number over 1 million | 00:44 |
| flwang1 | so we(catalyst Cloud) also need to figure out how to achieve that | 00:44 |
| *** davideagnello has quit IRC | 00:44 | |
| kgriffs | ah | 00:45 |
| *** achanda has joined #openstack-zaqar | 00:45 | |
| kgriffs | let's see, you need to know the user info, so you could add some WSGI middleware like: keystone ==> metering-thing ===> zaqar | 00:47 |
| kgriffs | is that what you were thinking? | 00:47 |
| *** achanda has quit IRC | 00:47 | |
| *** achanda has joined #openstack-zaqar | 00:50 | |
| flwang1 | kgriffs: yep, basically | 00:53 |
| kgriffs | using the middleware approach, you would probably want to update a counter on localhost, then have a daemon that periodically pushes the current counter value to a centralized service/DB | 00:53 |
| flwang1 | kgriffs: so basically we need to do the similar thing like keystonemiddleware, right? | 00:54 |
| kgriffs | if you go over a domain socket, it can be very fast so it doesn't slow down each request by much in order to increment the per-tenant counter | 00:54 |
| kgriffs | yeah, so just create a WSGI middleware app that checks the headers keystone injects, so that you can know which tenant's counter to increment | 00:54 |
| kgriffs | the keystone middleware will wrap your metering middleware which will then wrap the zaqar app | 00:55 |
| *** achanda has quit IRC | 00:55 | |
| flwang1 | kgriffs: which is the good sample to follow? keystonemiddleware https://github.com/openstack/keystonemiddleware or EOM? | 00:57 |
| *** achanda has joined #openstack-zaqar | 01:00 | |
| *** achanda has quit IRC | 01:01 | |
| kgriffs | hmm, to be honest I would just write a low-level WSGI middleware component and not try to use webob or anything. It will be faster, and for a simple thing like this, webob doesn't buy you much. | 01:02 |
| kgriffs | so you have a metering WSGI app | 01:03 |
| kgriffs | that would wrap the Zaqar app | 01:03 |
| kgriffs | finally, you could use AuthProtocol from keystone like: | 01:03 |
| kgriffs | app = AuthProtocol(metering_app, conf) | 01:03 |
| kgriffs | btw, for counter on localhost, it depends on how resilient you want to be to losing the counter; if your collector daemon runs fairly often, it may be OK to just keep all in RAM using memcached, redis, or tarantool | 01:04 |
| kgriffs | I believe tarantool is actually faster than redis | 01:04 |
| kgriffs | but anyway, the idea is to have the metering just update a counter as fast as possible then pass control to zaqar | 01:04 |
| *** achanda has joined #openstack-zaqar | 01:05 | |
| flwang1 | kgriffs: got, very helpful | 01:05 |
| kgriffs | and saving it locally then doing batched collections to a central location avoids having to play games with something like UDP | 01:05 |
| kgriffs | (ala statsd) | 01:05 |
| kgriffs | kk, gotta run. cheers! | 01:05 |
| flwang1 | kgriffs: cool, ttyl | 01:05 |
| *** achanda has quit IRC | 01:17 | |
| *** kgriffs is now known as kgriffs|afk | 01:19 | |
| *** achanda has joined #openstack-zaqar | 01:21 | |
| *** pradk has joined #openstack-zaqar | 01:23 | |
| *** achanda has quit IRC | 01:26 | |
| *** david-lyle has joined #openstack-zaqar | 02:22 | |
| openstackgerrit | Fei Long Wang proposed openstack/python-zaqarclient: Add subscriptions for queue object https://review.openstack.org/269398 | 02:33 |
| *** malini has joined #openstack-zaqar | 02:40 | |
| *** davideagnello has joined #openstack-zaqar | 02:45 | |
| *** mdnadeem has joined #openstack-zaqar | 02:48 | |
| *** davideagnello has quit IRC | 02:51 | |
| *** malini has quit IRC | 03:05 | |
| *** achanda has joined #openstack-zaqar | 03:21 | |
| *** davideagnello has joined #openstack-zaqar | 03:48 | |
| *** davideagnello has quit IRC | 03:52 | |
| *** flwang has quit IRC | 03:57 | |
| *** achanda has quit IRC | 04:00 | |
| *** achanda has joined #openstack-zaqar | 04:30 | |
| *** achanda has quit IRC | 04:38 | |
| *** achanda has joined #openstack-zaqar | 04:39 | |
| *** pradk has quit IRC | 04:58 | |
| *** achanda_ has joined #openstack-zaqar | 06:25 | |
| *** achanda has quit IRC | 06:28 | |
| *** davideagnello has joined #openstack-zaqar | 06:37 | |
| *** davideagnello has quit IRC | 06:38 | |
| *** flwang_home has joined #openstack-zaqar | 06:43 | |
| *** achanda_ has quit IRC | 06:43 | |
| *** flwang_home has quit IRC | 06:47 | |
| *** GB21 has joined #openstack-zaqar | 06:55 | |
| *** GB21 has quit IRC | 07:09 | |
| *** GB21 has joined #openstack-zaqar | 07:26 | |
| *** GB21 has quit IRC | 07:37 | |
| *** GB21 has joined #openstack-zaqar | 07:40 | |
| *** eliqiao_ is now known as eliqiao | 07:52 | |
| *** GB21 has quit IRC | 07:56 | |
| *** rcernin has joined #openstack-zaqar | 08:01 | |
| *** mdnadeem has quit IRC | 08:08 | |
| *** dynarro has joined #openstack-zaqar | 08:28 | |
| *** mdnadeem has joined #openstack-zaqar | 08:36 | |
| *** davideagnello has joined #openstack-zaqar | 08:38 | |
| *** davideagnello has quit IRC | 08:44 | |
| openstackgerrit | Nanke_Liu (lnk) proposed openstack/zaqar: replace deprecated iso8601_from_timestamp https://review.openstack.org/265406 | 09:35 |
| *** boris-42 has quit IRC | 09:43 | |
| *** davideagnello has joined #openstack-zaqar | 10:40 | |
| *** davideagnello has quit IRC | 10:46 | |
| openstackgerrit | Thomas Herve proposed openstack/zaqar: Use default pool for queue listing https://review.openstack.org/269165 | 10:51 |
| *** penchal has joined #openstack-zaqar | 10:55 | |
| *** malini has joined #openstack-zaqar | 11:13 | |
| *** dynarro has quit IRC | 11:21 | |
| *** malini has quit IRC | 12:30 | |
| *** dynarro has joined #openstack-zaqar | 12:35 | |
| *** davideagnello has joined #openstack-zaqar | 12:42 | |
| *** davideagnello has quit IRC | 12:47 | |
| *** mdnadeem has quit IRC | 12:51 | |
| *** mdnadeem has joined #openstack-zaqar | 13:10 | |
| *** shakamunyi has joined #openstack-zaqar | 13:24 | |
| *** AAzza has quit IRC | 13:50 | |
| *** AAzza has joined #openstack-zaqar | 13:53 | |
| *** venkat_ has joined #openstack-zaqar | 14:06 | |
| *** venkat_ has quit IRC | 14:08 | |
| *** sriram has joined #openstack-zaqar | 14:13 | |
| *** sriram has quit IRC | 14:13 | |
| *** sriram has joined #openstack-zaqar | 14:14 | |
| *** amitgandhinz has joined #openstack-zaqar | 14:19 | |
| *** kgriffs|afk is now known as kgriffs | 14:21 | |
| *** penchal has quit IRC | 14:29 | |
| *** mpanetta has joined #openstack-zaqar | 14:37 | |
| *** davideagnello has joined #openstack-zaqar | 14:43 | |
| *** davideagnello has quit IRC | 14:48 | |
| *** njohnston has joined #openstack-zaqar | 14:50 | |
| *** ametts has joined #openstack-zaqar | 14:54 | |
| openstackgerrit | Nanke_Liu (lnk) proposed openstack/zaqar: replace deprecated iso8601_from_timestamp https://review.openstack.org/265406 | 14:58 |
| *** sriram has quit IRC | 15:02 | |
| *** sriram has joined #openstack-zaqar | 15:04 | |
| *** malini has joined #openstack-zaqar | 15:07 | |
| *** mdnadeem has quit IRC | 15:19 | |
| *** pradk has joined #openstack-zaqar | 15:27 | |
| *** achanda has joined #openstack-zaqar | 15:29 | |
| *** achanda has quit IRC | 15:40 | |
| *** bradjones has joined #openstack-zaqar | 15:58 | |
| *** bradjones has joined #openstack-zaqar | 15:58 | |
| *** achanda has joined #openstack-zaqar | 15:59 | |
| *** boris-42 has joined #openstack-zaqar | 16:03 | |
| *** achanda has quit IRC | 16:03 | |
| *** sriram1 has joined #openstack-zaqar | 16:44 | |
| *** sriram has quit IRC | 16:44 | |
| *** sriram1 is now known as sriram | 16:45 | |
| *** sriram has quit IRC | 16:45 | |
| *** sriram has joined #openstack-zaqar | 16:45 | |
| *** rcernin has quit IRC | 16:53 | |
| therve | ryansb, With all the talk about creating signed URLs, did we even mention *using* them? :) | 16:56 |
| ryansb | hm? | 16:57 |
| ryansb | as in using them from the CLI? | 16:57 |
| therve | ryansb, More from the client library | 16:58 |
| ryansb | I haven't thought of a great way to surface that functionality personally, other than exposing flags to add them | 16:58 |
| *** david-lyle has quit IRC | 16:58 | |
| therve | Well presumably we should be able to create a client with the given data | 16:58 |
| therve | So that the http headers at least are managed by zaqarclient | 16:59 |
| ryansb | are you talking about the python client, rather than CLI? | 17:01 |
| therve | Yeah I'm talking from the python client | 17:02 |
| *** david-lyle has joined #openstack-zaqar | 17:06 | |
| therve | I think managing it with a new auth backend would be the best | 17:08 |
| *** flwang has joined #openstack-zaqar | 17:08 | |
| *** rcernin has joined #openstack-zaqar | 17:08 | |
| ryansb | so you'd instantiate a client with, say, signature, project ID, url, etc | 17:09 |
| ryansb | and it would do the auth backend for you? | 17:09 |
| *** dynarro has quit IRC | 17:10 | |
| therve | Basically | 17:12 |
| ryansb | I can take a look at that tomorrow, unless you've got time for it | 17:13 |
| *** ametts has quit IRC | 17:18 | |
| *** ametts has joined #openstack-zaqar | 17:31 | |
| *** davideagnello has joined #openstack-zaqar | 17:37 | |
| *** flwang has quit IRC | 17:53 | |
| *** pradk has quit IRC | 18:19 | |
| therve | ryansb, I'll take some time :) | 18:23 |
| ryansb | k | 18:24 |
| *** pradk has joined #openstack-zaqar | 18:26 | |
| *** pradk has quit IRC | 18:26 | |
| *** malini has quit IRC | 18:30 | |
| *** zhiyan has quit IRC | 18:33 | |
| *** zhiyan has joined #openstack-zaqar | 18:36 | |
| *** boris-42 has quit IRC | 18:53 | |
| *** malini has joined #openstack-zaqar | 19:08 | |
| *** eliqiao has quit IRC | 19:11 | |
| *** eliqiao has joined #openstack-zaqar | 19:13 | |
| *** mpanetta is now known as yourself | 19:53 | |
| *** yourself is now known as mpanetta | 19:53 | |
| *** malini has quit IRC | 20:21 | |
| *** flwang has joined #openstack-zaqar | 20:24 | |
| *** malini has joined #openstack-zaqar | 20:24 | |
| flwang1 | jasondotstar: ping | 20:27 |
| *** malini has quit IRC | 20:42 | |
| *** boris-42 has joined #openstack-zaqar | 20:42 | |
| jasondotstar | pong | 20:44 |
| flwang1 | jasondotstar: any progress about the puppet work? | 20:58 |
| jasondotstar | some. :-( | 20:58 |
| jasondotstar | give me a couple mins here | 20:59 |
| jasondotstar | I'm going to abandon the commits I have | 20:59 |
| jasondotstar | and re-submit what I've got. | 20:59 |
| flwang1 | jasondotstar: why you have to abandon? | 21:00 |
| jasondotstar | merge conflits | 21:05 |
| jasondotstar | *conflicts | 21:06 |
| jasondotstar | it'd rather submit a new review | 21:06 |
| jasondotstar | i'd rather submit a new review | 21:06 |
| flwang1 | jasondotstar: ok, no problem | 21:07 |
| flwang1 | jasondotstar: so how far away from the ready-status? :) | 21:07 |
| Eva-i | flwang1: hello, according to doc team's spec about api-ref migration to rst/swagger, the migration must be done by now. But seems like doc team can't keep up - in "api-site" repo still has wadl/docbook files and no swagger/rst files are being used at docs.openstack.org. Their fairy-slipper tool is working while throwing exceptions, but still able to generate working swagger/rst files from wadl/docbook files. Should I manually make rst/swagger | 21:09 |
| Eva-i | Even if I make these files, I don't know where to commit them. | 21:09 |
| flwang1 | Eva-i: what's the comments from docs team? | 21:10 |
| Eva-i | flwang1: silence | 21:10 |
| flwang1 | wow | 21:10 |
| jasondotstar | 2 days. | 21:10 |
| jasondotstar | flwang1: ^ | 21:10 |
| flwang1 | jasondotstar: so after 2 days, we can get a workable puppet zaqar? :) | 21:10 |
| flwang1 | i can't believe it :D | 21:11 |
| Eva-i | :o | 21:11 |
| flwang1 | Eva-i: ok, i will ping lana to confirm | 21:11 |
| Eva-i | flwang1: thanks | 21:11 |
| Eva-i | flwang1: Lana is one of doc team members? | 21:12 |
| flwang1 | Lana is the PTL | 21:12 |
| Eva-i | flwang1: oh | 21:12 |
| Eva-i | flwang1: remember you asked me add info about pipeline options to config-ref? I made a patch in openstack-manuals repo and checked you as reviewer. Have you received the email notification about it? | 21:14 |
| flwang1 | Eva-i: oh, yep | 21:15 |
| flwang1 | it's on my list, but recently i have some internal stuff, so bad review performance | 21:15 |
| flwang1 | sorry about that | 21:15 |
| Eva-i | flwang1: it's okay, take your time | 21:15 |
| flwang1 | Eva-i: thanks, now i'm working on the TTL issue of subscription | 21:17 |
| jasondotstar | flwang1: yep | 21:17 |
| flwang1 | and the zaqar UI | 21:17 |
| jasondotstar | flwang1: im sure there be a host of -1s as I cleanup the debian install | 21:17 |
| Eva-i | flwang1: cool, it's very important | 21:17 |
| jasondotstar | but the rpm based one should pass | 21:18 |
| jasondotstar | *there will be | 21:18 |
| flwang1 | jasondotstar: ok, i see | 21:18 |
| flwang1 | jasondotstar: does that mean we still can't make it work for ububtu/debian platform? | 21:18 |
| jasondotstar | flwang1: that's what I'm in the process of cleaning up | 21:19 |
| jasondotstar | in the end, we *should* be able to. | 21:19 |
| flwang1 | jasondotstar: ok, got it, thanks | 21:19 |
| jasondotstar | i've been testing against debian jessie. | 21:19 |
| Eva-i | jasondotstar: thank you, jasondotstar, for working on pupper zaqar | 21:19 |
| Eva-i | *puppet | 21:19 |
| jasondotstar | don't thank me just yet- | 21:19 |
| jasondotstar | little more work to do :-) | 21:20 |
| Eva-i | okay okay =) | 21:20 |
| flwang1 | jasondotstar: do you need a hand? should we ask EmilienM to get some help? | 21:21 |
| jasondotstar | flwang1: I'll ping him | 21:21 |
| flwang1 | jasondotstar: cool, again, many users are keen for it | 21:22 |
| jasondotstar | im sure. | 21:22 |
| jasondotstar | it's been busy, but I'm on it. | 21:22 |
| EmilienM | hello | 21:22 |
| flwang1 | EmilienM: hi | 21:23 |
| flwang1 | EmilienM: we're talking about the puppet zaqar work | 21:23 |
| EmilienM | no progress so far iirc | 21:23 |
| flwang1 | EmilienM: yep, so that's why i ask jasondotstar if he needs some help from you guys | 21:24 |
| *** mdnadeem has joined #openstack-zaqar | 21:29 | |
| jasondotstar | EmilienM: I will need help with the beaker testing stuff | 21:33 |
| jasondotstar | writing up the tests | 21:33 |
| EmilienM | sure | 21:33 |
| jasondotstar | right now, I've been concentrating on the manifest work | 21:33 |
| EmilienM | are you familiar with puppet? | 21:33 |
| jasondotstar | yes :-) | 21:33 |
| EmilienM | beaker is really nothing | 21:33 |
| EmilienM | just write the manifest you want | 21:33 |
| jasondotstar | isn't there q requirement | 21:33 |
| jasondotstar | s/q/a | 21:33 |
| jasondotstar | to have the beaker tests written? | 21:34 |
| jasondotstar | EmilienM: ^ | 21:34 |
| EmilienM | jasondotstar: well, it's always better to test what you code, isn't? | 21:35 |
| jasondotstar | "beaker is really nothing" <--- | 21:35 |
| EmilienM | I personally put in low priority in my review list the patches that has no tests | 21:35 |
| EmilienM | jasondotstar: I mean, nothing hard to make | 21:35 |
| EmilienM | it's just a manifest | 21:35 |
| jasondotstar | just trying to make sure everything is included in the patch | 21:35 |
| jasondotstar | right | 21:35 |
| jasondotstar | which is why i havent touched the beaker stuff at all. | 21:36 |
| EmilienM | there is ONE file to write: https://github.com/openstack/puppet-nova/blob/master/spec/acceptance/basic_nova_spec.rb | 21:36 |
| jasondotstar | so... if it's simple then, no worries. | 21:36 |
| EmilienM | and so easy to reproduce for zaqar | 21:36 |
| EmilienM | very simple | 21:36 |
| jasondotstar | cool. | 21:36 |
| EmilienM | but let me know if need help | 21:36 |
| jasondotstar | perhaps I don't, but I'll let you know. I just haven't look at it at all. as you said, low priority. | 21:37 |
| flwang1 | EmilienM: jasondotstar: goods to see the work is simple from your PoV :D | 21:38 |
| jasondotstar | ;-] | 21:39 |
| EmilienM | flwang1: the work is simple but I never found time to write puppet-zaqar itself | 21:40 |
| EmilienM | writting manifests & unit tests take time | 21:40 |
| EmilienM | but we have a lot of modules | 21:40 |
| EmilienM | so we can take examples from there | 21:41 |
| flwang1 | EmilienM: i totally understand | 21:41 |
| *** mdnadeem has quit IRC | 21:41 | |
| *** sriram has quit IRC | 21:42 | |
| *** eliqiao has quit IRC | 22:12 | |
| *** eliqiao has joined #openstack-zaqar | 22:13 | |
| *** rcernin has quit IRC | 22:21 | |
| openstackgerrit | Fei Long Wang proposed openstack/python-zaqarclient: Fix wrong api version type https://review.openstack.org/267906 | 22:33 |
| *** amitgandhinz has quit IRC | 22:45 | |
| openstackgerrit | Fei Long Wang proposed openstack/python-zaqarclient: Use API version checking directly https://review.openstack.org/269904 | 22:56 |
| *** mpanetta has quit IRC | 23:04 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!