| *** nosnos has joined #openstack-marconi | 00:58 | |
| *** nosnos_ has joined #openstack-marconi | 01:53 | |
| *** nosnos has quit IRC | 01:54 | |
| *** balajiiyer has joined #openstack-marconi | 01:59 | |
| *** oz_akan_ has quit IRC | 03:08 | |
| openstackgerrit | liusheng proposed a change to openstack/marconi: add copyright notices to some source files https://review.openstack.org/69265 | 03:09 |
|---|---|---|
| *** oz_akan_ has joined #openstack-marconi | 03:18 | |
| *** oz_akan_ has quit IRC | 03:24 | |
| *** haomaiwang has joined #openstack-marconi | 03:27 | |
| *** haomaiwang has quit IRC | 03:29 | |
| *** haomaiwa_ has joined #openstack-marconi | 03:30 | |
| *** chandankumar_ has joined #openstack-marconi | 04:10 | |
| *** chandankumar_ has quit IRC | 04:24 | |
| *** chandankumar_ has joined #openstack-marconi | 04:29 | |
| *** chandankumar_ has quit IRC | 04:53 | |
| *** chandankumar_ has joined #openstack-marconi | 05:23 | |
| *** balajiiyer has quit IRC | 05:44 | |
| *** renlt has joined #openstack-marconi | 05:44 | |
| *** chandankumar_ has quit IRC | 05:52 | |
| *** chandankumar_ has joined #openstack-marconi | 06:38 | |
| *** dustin_ has joined #openstack-marconi | 06:41 | |
| *** chandankumar_ has quit IRC | 06:46 | |
| *** renlt has quit IRC | 07:00 | |
| *** flaper87|afk is now known as flaper87 | 07:37 | |
| *** al-maisan has quit IRC | 07:40 | |
| *** renlt has joined #openstack-marconi | 07:45 | |
| *** ykaplan has joined #openstack-marconi | 08:12 | |
| flwang | flaper87: I'm asking your return favor on this https://review.openstack.org/#/c/66787/ ? :D | 08:13 |
| flaper87 | flwang: LOOOL | 08:14 |
| *** renlt has quit IRC | 08:16 | |
| flwang | :P | 08:16 |
| *** dustin_ has quit IRC | 08:17 | |
| flaper87 | flwang: done | 08:22 |
| flwang | flaper87: thanks | 08:25 |
| flwang | flaper87: as for "What about adding a separate example for this? `management.py` (or something along those lines) ?" did you mean add a new file to place the verification code of admin? | 08:25 |
| flwang | flaper87: i mean /health and /shards | 08:26 |
| flaper87 | flwang: yeah, I'd like the simple.py example to be... simple | 08:26 |
| flaper87 | nothing more there | 08:26 |
| flaper87 | just get the client and post messages | 08:26 |
| flaper87 | management examples could live in their own module | 08:26 |
| flwang | flaper87: or should I just remove my code from example.py? | 08:27 |
| flwang | i mean simple.py | 08:27 |
| flaper87 | flwang: yeah, remove it from simple and add a new example for health | 08:27 |
| flwang | okay | 08:27 |
| flaper87 | flwang: maybe something that actually uses the response | 08:28 |
| flaper87 | flwang: like, how do I know everything is fine from that response ? | 08:28 |
| flaper87 | brb | 08:28 |
| *** tmu has quit IRC | 08:29 | |
| *** tmu has joined #openstack-marconi | 08:29 | |
| flaper87 | back | 08:44 |
| *** werebutt has joined #openstack-marconi | 08:47 | |
| *** werebutt has left #openstack-marconi | 08:47 | |
| flwang | flaper87: https://review.openstack.org/#/c/66787/3/marconiclient/queues/v1/api.py | 09:14 |
| flaper87 | flwang: 'sup ? | 09:14 |
| flwang | flaper87: did you mean add a comment above 'health': {? | 09:15 |
| flaper87 | flwang: nope, add a new key to the health dict | 09:16 |
| flaper87 | 'admin': True | 09:16 |
| flwang | flaper87: ok, got it | 09:43 |
| *** flaper87 is now known as flaper87|afk | 09:55 | |
| *** chandankumar_ has joined #openstack-marconi | 10:13 | |
| openstackgerrit | Fei Long Wang proposed a change to openstack/python-marconiclient: Add /health support to the client https://review.openstack.org/66787 | 10:14 |
| *** haomaiwa_ has quit IRC | 10:29 | |
| *** haomaiwang has joined #openstack-marconi | 10:30 | |
| *** chandankumar_ has quit IRC | 10:37 | |
| *** chandankumar_ has joined #openstack-marconi | 11:02 | |
| *** chandankumar_ has quit IRC | 11:08 | |
| *** chandankumar_ has joined #openstack-marconi | 11:15 | |
| *** flaper87|afk is now known as flaper87 | 11:18 | |
| *** chandankumar_ has quit IRC | 11:39 | |
| *** malini_afk is now known as malini | 11:47 | |
| *** malini has left #openstack-marconi | 12:01 | |
| *** chandankumar_ has joined #openstack-marconi | 12:05 | |
| *** chandankumar_ has quit IRC | 12:08 | |
| *** tedross has joined #openstack-marconi | 12:54 | |
| *** chandankumar_ has joined #openstack-marconi | 12:54 | |
| *** oz_akan_ has joined #openstack-marconi | 13:16 | |
| *** oz_akan_ has quit IRC | 13:17 | |
| *** oz_akan_ has joined #openstack-marconi | 13:17 | |
| *** chandankumar_ has quit IRC | 13:22 | |
| *** chandankumar_ has joined #openstack-marconi | 13:23 | |
| flwang | flaper87: can you revisit my patch https://review.openstack.org/#/c/66787/ ? | 13:26 |
| flaper87 | flwang: will do | 13:26 |
| flwang | flaper87: thanks | 13:26 |
| *** chandankumar_ has quit IRC | 13:44 | |
| *** rustlebee is now known as russellb | 14:12 | |
| *** jcru has joined #openstack-marconi | 14:21 | |
| *** chandankumar_ has joined #openstack-marconi | 14:23 | |
| *** amitgandhi has joined #openstack-marconi | 14:25 | |
| *** chandankumar_ has quit IRC | 14:27 | |
| *** chandankumar_ has joined #openstack-marconi | 14:29 | |
| *** amitgandhi has quit IRC | 14:30 | |
| *** amitgandhi has joined #openstack-marconi | 14:30 | |
| *** chandankumar_ has quit IRC | 14:33 | |
| *** chandankumar_ has joined #openstack-marconi | 14:34 | |
| *** balajiiyer has joined #openstack-marconi | 14:43 | |
| *** chandankumar_ has quit IRC | 14:44 | |
| *** balajiiyer has quit IRC | 14:47 | |
| *** mpanetta has joined #openstack-marconi | 14:50 | |
| *** chandankumar_ has joined #openstack-marconi | 14:56 | |
| *** alcabrera has joined #openstack-marconi | 15:01 | |
| alcabrera | Good morning! | 15:02 |
| *** alcabrera has quit IRC | 15:03 | |
| *** alcabrera has joined #openstack-marconi | 15:03 | |
| *** chandankumar_ has quit IRC | 15:06 | |
| *** malini has joined #openstack-marconi | 15:09 | |
| *** rwsu has joined #openstack-marconi | 15:16 | |
| flaper87 | alcabrera: goood morning | 15:21 |
| alcabrera | flaper87: hey! How are you? :) | 15:22 |
| flaper87 | alcabrera: I'm doing well, how are you doing? | 15:23 |
| *** dustin_ has joined #openstack-marconi | 15:27 | |
| alcabrera | I'm alright. It's a slow Monday for me. :P | 15:27 |
| malini | o/ | 15:28 |
| malini | I am planning to stalk folks (no..not really) on openstack-infra, to get some eyes on my devstack patches | 15:29 |
| alcabrera | malini: the battle for the patches! :D | 15:30 |
| malini | flaper87: do you know any core infra nicks, hanging ready on your finger tips? | 15:30 |
| flaper87 | malini: mordred ? | 15:31 |
| flaper87 | malini: sdage | 15:31 |
| flaper87 | malini: fungi | 15:31 |
| flaper87 | malini: jblair | 15:31 |
| malini | awesome..here I go :) | 15:31 |
| malini | thanks flaper87!! | 15:31 |
| flwang | alcabrera: ping | 15:32 |
| alcabrera | flwang: hey! :D | 15:33 |
| flwang | alcabrera: I'm working on this https://bugs.launchpad.net/marconi/+bug/1243899 | 15:33 |
| *** cpallares has joined #openstack-marconi | 15:34 | |
| flwang | alcabrera: two questions I'd like to ask, are you free now? | 15:34 |
| alcabrera | flwang: got it. how's it coming along, and yes, I've some free time at the moment. | 15:34 |
| flwang | alcabrera: firstly, when i trying to recreate, I always get 201 instead of 204, am I missing anything? | 15:35 |
| alcabrera | hmmm | 15:36 |
| alcabrera | lemme double check the current state of the code | 15:37 |
| flwang | alcabrera: ok, thanks, and the 2nd question, when I request GET http://127.0.0.1:8888/v1/shards I got 405 error | 15:38 |
| alcabrera | hmmm, it *does* always return 201. | 15:38 |
| alcabrera | flwang: could you verify whether this is already implemented? (https://bugs.launchpad.net/marconi/+bug/1243899) | 15:39 |
| alcabrera | what I'm seeing is that it already does the replacing behavior. Someone might have already taken care of this. I might've even done it at some point and forgotten to update the status. | 15:40 |
| alcabrera | Now, about the 405 | 15:40 |
| flwang | alcabrera: based on my test with the latest code, it's always return 201, so I'm going to close the bug | 15:40 |
| alcabrera | flwang: works for me | 15:41 |
| flwang | alcabrera: can you recreate the 405 error? | 15:42 |
| alcabrera | I'm going to try it now. | 15:43 |
| alcabrera | GET http://127.0.0.1:8888/v1/shards gives me 204 when empty, flwang. Now testing one shard case. | 15:47 |
| flwang | alcabrera: I can create sharding with PUT and delete it with DELETE, but the GET doesn't work, weird | 15:48 |
| alcabrera | http get localhost:8000/v1/shards worked with one shard, too | 15:48 |
| alcabrera | hmmm | 15:48 |
| alcabrera | the sharding interface should only be implemented when runnning on a mongodb storage | 15:49 |
| alcabrera | (so far, anyway) | 15:49 |
| alcabrera | lemme retry these tests with sqlite | 15:49 |
| alcabrera | flwang: reproduced! | 15:50 |
| alcabrera | If I set my storage to sqlite... | 15:50 |
| alcabrera | GET http://localhost:8000/v1/shards => HTTP 405 | 15:50 |
| flwang | alcabrera: I noticed that by debug, but the question is I'm using mongodb | 15:51 |
| flwang | alcabrera: it's set in marconi-queues.conf # Storage driver module (e.g., mongodb, sqlite) | 15:53 |
| flwang | storage = mongodb | 15:53 |
| alcabrera | weird. :x | 15:53 |
| alcabrera | I was using storage = mongodb while getting the 201/204s. | 15:53 |
| alcabrera | then swapped over to storage = sqlite and started getting 405s... | 15:53 |
| alcabrera | hmmm... | 15:53 |
| flwang | I think 405 is make sense for sqlite, since it's not implemented, is it? | 15:54 |
| alcabrera | flwang: that's correct. storage/sqlite/[shards, catalogue] still need to be filled on. | 15:56 |
| alcabrera | s/on/in | 15:56 |
| flwang | alcabrera: the only conf may impact this is marconi-queues.conf, right? | 15:57 |
| flwang | alcabrera: if so, would you mind sharing yours with http://dpaste.com/ ? | 15:58 |
| alcabrera | flwang: I think we migrated away from having marconi-queues.conf. It should be just marconi.conf now. I'll paste mine in a moment. :) | 15:59 |
| flwang | oh, really? maybe that's the root cause :) | 16:00 |
| flaper87 | we migrate away from it | 16:00 |
| alcabrera | flwang: http://paste.openstack.org/show/61947/ | 16:00 |
| *** ykaplan has quit IRC | 16:00 | |
| *** whenry has joined #openstack-marconi | 16:03 | |
| flwang | alcabrera: cool, works now | 16:04 |
| flwang | alcabrera: that's the root cause | 16:04 |
| alcabrera | heh - nothing quite like configuration file changes to wreak havoc. :) | 16:05 |
| *** ykaplan has joined #openstack-marconi | 16:13 | |
| *** nosnos_ has quit IRC | 16:14 | |
| *** jergerber has joined #openstack-marconi | 16:22 | |
| alcabrera | be back in a bit. | 16:50 |
| *** alcabrera has quit IRC | 16:50 | |
| *** ykaplan has quit IRC | 16:52 | |
| *** kgriffs_afk is now known as kgriffs | 16:52 | |
| kgriffs | malini: have you seen this failure? https://review.openstack.org/#/c/68701/ | 16:56 |
| kgriffs | specifically: http://logs.openstack.org/01/68701/2/gate/gate-marconi-pypy/c9c64af/console.html.gz | 16:56 |
| kgriffs | flaper87: ^^^ | 16:56 |
| kgriffs | so, it doesn't fail every time | 16:56 |
| kgriffs | and it only fails on pypy when it does | 16:56 |
| malini | checking now | 16:56 |
| flaper87 | kgriffs: yup | 16:56 |
| flaper87 | that's the failure I asked help with the other day | 16:57 |
| kgriffs | I looked at the code and couldn't see how /messages/messages could be possible | 16:57 |
| flaper87 | exactly | 16:57 |
| flaper87 | :P | 16:57 |
| flaper87 | that was fixed once | 16:57 |
| flaper87 | flwang: did the fix for that | 16:57 |
| flaper87 | not sure *why* that happens | 16:57 |
| kgriffs | gtk I'm not (totally) crazy | 16:57 |
| malini | kgriffs: YESY!! | 16:57 |
| malini | YES! | 16:57 |
| flaper87 | but I'm sure it happens just w/ pypy | 16:57 |
| malini | I have seen it | 16:57 |
| flaper87 | so, there's something evil behind pypy, but don't tell Alex_Gaynor | 16:57 |
| flaper87 | opppppppppppppssssssssssssss | 16:57 |
| kgriffs | shhhhh | 16:57 |
| kgriffs | malini: any idea of how to repro? | 16:58 |
| malini | for a while I have seen them move around py27 & py33 as well | 16:58 |
| kgriffs | smells like a race condition | 16:58 |
| flaper87 | malini: yeah, at that point there was actually a bug | 16:59 |
| malini | no..I have never seen tht thing, when I was writing those tests against mongo db | 16:59 |
| flaper87 | malini: but it got fixed | 16:59 |
| flaper87 | and moved into pyyp | 16:59 |
| flaper87 | pypy | 16:59 |
| flaper87 | :P | 16:59 |
| kgriffs | heh | 16:59 |
| flaper87 | we jitted that bug | 16:59 |
| kgriffs | so, that test is running against mongo or sqlite? | 16:59 |
| flaper87 | sqlite | 16:59 |
| flaper87 | the gate ain't do mongo | 16:59 |
| malini | hmm…I saw tht happening after flwang fixed the 'bug' | 16:59 |
| flaper87 | then flwang is liying to us | 17:00 |
| flaper87 | flwang: dude, not fair, not fair | 17:00 |
| kgriffs | hmmm | 17:00 |
| malini | coz, I added tht test after the bug fix was merged | 17:00 |
| kgriffs | ok, so maybe we just run that test in a loop and see if it eventually fails? | 17:00 |
| malini | https://bugs.launchpad.net/marconi/+bug/1240897 | 17:01 |
| flaper87 | kgriffs: I tried that once but it didn't work | 17:01 |
| flaper87 | not saying you shouldn't try, though | 17:01 |
| malini | we can skip the test till we figure this out, if its becoming a blocker for gate | 17:04 |
| *** reed has joined #openstack-marconi | 17:05 | |
| *** mpanetta has quit IRC | 17:08 | |
| *** mpanetta has joined #openstack-marconi | 17:08 | |
| kgriffs | malini: ok, we may need to. Can you put it on your todo list to get a reliable repro? | 17:17 |
| kgriffs | actually, malini: can you also file a bug? | 17:18 |
| kgriffs | I don't want this to be forgotten about | 17:18 |
| malini | yeap..I will reopen the original one | 17:18 |
| kgriffs | chances are there really is a sneaky bug in there somewhere | 17:18 |
| kgriffs | kk | 17:18 |
| kgriffs | pls. link to the above error log | 17:18 |
| malini | is there anything different b/w sqlite & mongo drivers in how the href is constructed? | 17:19 |
| openstackgerrit | Cindy Pallares proposed a change to openstack/marconi: Removed copyright from files without code https://review.openstack.org/69417 | 17:22 |
| malini | kgriffs: https://bugs.launchpad.net/marconi/+bug/1273335 | 17:25 |
| *** alcabrera has joined #openstack-marconi | 17:27 | |
| kgriffs | triaged. I marked it High until/unless we find out it is not a real bug | 17:28 |
| *** vkmc has joined #openstack-marconi | 17:29 | |
| malini | thanks! | 17:31 |
| openstackgerrit | Kurt Griffiths proposed a change to openstack/marconi: Skip message_get_many schema test until we can repro reliably https://review.openstack.org/69426 | 18:00 |
| kgriffs | flaper87, alcabrera, malini: ^^^ | 18:01 |
| alcabrera | kgriffs: +2'd, because a bug was filed, and because (Maybe) false positives are annoying. | 18:02 |
| *** rossk has joined #openstack-marconi | 18:12 | |
| alcabrera | flaper87: ping | 18:24 |
| alcabrera | I noticed something curious about the proper release of oslo.cache - there doesn't appear to be an unset function. :x | 18:24 |
| alcabrera | there's unset_many | 18:24 |
| kgriffs | hmmm | 18:25 |
| kgriffs | curious | 18:25 |
| kgriffs | can you just pass a single item to unset_many? | 18:25 |
| alcabrera | kgriffs: it can be done, as long as it is wrapped in an iterable, say | 18:26 |
| alcabrera | unset_many(['key']) | 18:26 |
| kgriffs | makes me think you may be missing an opportunity for optimizations | 18:26 |
| kgriffs | i mean, if the backend is faster when you tell it to unset a single thing vs many | 18:26 |
| kgriffs | may not be significant, just a thought | 18:26 |
| kgriffs | but anway, why not just have it take *args | 18:27 |
| kgriffs | then you don't have to wrap in an iterable | 18:27 |
| alcabrera | that's an interesting point | 18:28 |
| alcabrera | *args unpacking never ceases to please. :) | 18:28 |
| kgriffs | i mean, unless we expect someone to have an array already | 18:28 |
| kgriffs | but then, they could just pass *things | 18:28 |
| kgriffs | anyway | 18:29 |
| kgriffs | food for thought | 18:29 |
| openstackgerrit | Alejandro Cabrera proposed a change to openstack/marconi: fix(shards): guard against registering bad shard URIs https://review.openstack.org/68267 | 18:33 |
| *** alcabrera is now known as alcabrera|afk | 18:33 | |
| *** alcabrera|afk is now known as alcabrera | 18:43 | |
| flaper87 | back | 18:44 |
| alcabrera | flaper87: welcome back! | 18:47 |
| flaper87 | alcabrera: :D | 18:48 |
| *** reed has quit IRC | 18:48 | |
| openstackgerrit | Kurt Griffiths proposed a change to openstack/marconi: Skip message_get_many schema test until we can repro reliably https://review.openstack.org/69426 | 18:49 |
| kgriffs | stupid pep8 | 18:49 |
| kgriffs | :p | 18:49 |
| kgriffs | ok, I love pep8. usually | 18:49 |
| alcabrera | line too long error? ;) | 18:50 |
| kgriffs | yeah | 18:52 |
| *** reed has joined #openstack-marconi | 18:53 | |
| openstackgerrit | Alejandro Cabrera proposed a change to openstack/marconi: fix(shards): guard against registering bad shard URIs https://review.openstack.org/68267 | 19:02 |
| alcabrera | closed one critical bug, one test bug, and the bug originally planned for in that. ^^ | 19:03 |
| alcabrera | kgriffs, flaper87: ^ | 19:03 |
| alcabrera | so - two new bugs discovered while working on this patch. | 19:03 |
| alcabrera | it'd be awesome to get tempest support in. mongodb-related functionality gets little breaks from time to time since we aren't able to gate on those. | 19:04 |
| alcabrera | :( | 19:04 |
| alcabrera | thanks for working on getting tempest support for marconi, malini! | 19:04 |
| flaper87 | malini: +! | 19:04 |
| flaper87 | malini: +1 | 19:04 |
| kgriffs | yes, we needs teh tempest | 19:04 |
| malini | its fun, till the review part :( | 19:04 |
| *** balajiiyer has joined #openstack-marconi | 19:12 | |
| *** vkmc has quit IRC | 19:37 | |
| kgriffs | flaper87: hey, thoughts on this patch? https://review.openstack.org/#/c/69426/ | 19:41 |
| flaper87 | back | 19:59 |
| flaper87 | sorry | 19:59 |
| flaper87 | alcabrera: btw, thanks for the comment on my post ;) | 19:59 |
| flaper87 | kgriffs: unfortunately, I agree! | 20:00 |
| flaper87 | lets skip it | 20:00 |
| openstackgerrit | A change was merged to openstack/marconi: Skip message_get_many schema test until we can repro reliably https://review.openstack.org/69426 | 20:03 |
| alcabrera | flaper87: on real-time systems? sure, no problem. It made for good reading. :) | 20:03 |
| flaper87 | alcabrera: :) I'm glad you like it. I had the feeling I was rushing the post | 20:04 |
| flaper87 | as in not getting into enough details | 20:04 |
| flaper87 | but hey, I've the same feeling every time I write something | 20:04 |
| flaper87 | so... | 20:04 |
| alcabrera | it's a start. waiting to get all the details in on the first go would flood the reader. | 20:04 |
| flaper87 | :P | 20:04 |
| *** rossk_ has joined #openstack-marconi | 20:04 | |
| openstackgerrit | Alejandro Cabrera proposed a change to openstack/marconi: fix(shards): guard against registering bad shard URIs https://review.openstack.org/68267 | 20:05 |
| mpanetta | flaper87: Link? :) | 20:05 |
| alcabrera | there we go - jenkins will be happy now that kgriffs put in the patch to skip the nondeterministic test. | 20:05 |
| flaper87 | mpanetta: http://blog.flaper87.com/post/52e43f93d987d2cc17700b58/ | 20:05 |
| *** openstackgerrit has quit IRC | 20:06 | |
| mpanetta | flaper87: Fun stuff! :) | 20:06 |
| *** openstackgerrit has joined #openstack-marconi | 20:06 | |
| *** jcru_ has joined #openstack-marconi | 20:06 | |
| kgriffs | kewl | 20:06 |
| kgriffs | alcabrera: do you remember what python hash library we almost used in marconi? | 20:07 |
| kgriffs | https://pypi.python.org/pypi/mmh3/2.3 | 20:07 |
| kgriffs | was it that one? | 20:07 |
| kgriffs | I want to say it was something else. | 20:07 |
| alcabrera | hmmm | 20:07 |
| flaper87 | I think I know what kgriffs is going to say | 20:07 |
| kgriffs | but, maybe that was it after all | 20:08 |
| alcabrera | it was either that or pyhash | 20:08 |
| *** rossk has quit IRC | 20:08 | |
| flaper87 | it was that one, AFAIK | 20:08 |
| kgriffs | flaper87: which? | 20:08 |
| *** jcru has quit IRC | 20:08 | |
| alcabrera | https://pypi.python.org/pypi/pyhashxx or https://pypi.python.org/pypi/mmh3/2.3 | 20:08 |
| flaper87 | mmh3 | 20:08 |
| flaper87 | (git log) | 20:09 |
| alcabrera | w00t | 20:09 |
| alcabrera | git log knows much better than I | 20:09 |
| flaper87 | my laptop is dying | 20:09 |
| flaper87 | can't find it | 20:11 |
| flaper87 | :/ | 20:11 |
| flaper87 | I think we were using pip-requires at that time | 20:11 |
| openstackgerrit | Kurt Griffiths proposed a change to openstack/marconi: chore: Fix pylint errors https://review.openstack.org/68701 | 20:13 |
| *** vkmc has joined #openstack-marconi | 20:13 | |
| mpanetta | flaper87: Yeah people seem to confuse real-time with real-fast. I guess mostly ones that don't understand the definition of real-time. | 20:14 |
| flaper87 | mpanetta: indeed. TBH, the term `real-time` bothers me even more. Not just because the misconception or lack of precision. It bothers me because there's no such thing as `real-time` to begin with | 20:17 |
| kgriffs | malini: before I forget... | 20:17 |
| jraim_i___ | kgriffs: ping | 20:17 |
| flaper87 | real-time in the real world doesn't mean anything | 20:17 |
| kgriffs | we need to figure out a way to be testing with the env variable MARCONI_TEST_SLOW=1 | 20:17 |
| kgriffs | jraim_i___: yo yo | 20:17 |
| flaper87 | There's no principle, statement nor theory describing what "real time" is | 20:18 |
| flaper87 | (AFAIK) | 20:18 |
| jraim_i___ | kgriffs: do you guys have your devstack stuff working? I copied a lot of the work you all did for barbican and we are running into problems | 20:18 |
| malini | kgriffs: hmmm..I'll chk what other projects are doing..is that variable for slow tests? | 20:18 |
| kgriffs | malini: ^^ | 20:18 |
| flaper87 | and since computers are ruled by the laws of physics, the should stick to that | 20:18 |
| mpanetta | I think the idea came from interactive systems, they had to control things in 'real time' | 20:18 |
| kgriffs | malini: yes - thing is, we don't want to run it normally as part of the gate | 20:18 |
| mpanetta | Except real-time does not really exist... | 20:18 |
| kgriffs | malini: but, we should run at least once before tagging each milestone | 20:18 |
| flaper87 | but anyway, this is definitely not something I wanted to bring up in that post | 20:18 |
| malini | jraim_i___: depends on what you call devstack stuff | 20:18 |
| malini | kgriffs: I'll add it to my list of stuff..do we need a bp? | 20:19 |
| kgriffs | jraim_i___: malini has been working in that neck of the woods lately | 20:19 |
| jraim_i___ | malini: When we enabled the marconi service in devstack, stack.sh blows up | 20:19 |
| flaper87 | mpanetta: yeah and then nodejs came out caliming it is "real-time" | 20:19 |
| mpanetta | Heh, but it isn't deterministic... | 20:19 |
| flaper87 | jraim_i___: :S that doesn't sound good | 20:19 |
| kgriffs | malini: mmm. Not sure. Maybe a bug. Unless you think it is going to be a big work item | 20:19 |
| malini | jraim_i___: I was able to get it running locally in devstack | 20:19 |
| flaper87 | the devstack thing should work | 20:20 |
| mpanetta | Real-time to me, mean deterministic if nothing else... | 20:20 |
| kgriffs | flaper87: WAH?! I thought Socket.IO === Realtime | 20:20 |
| jraim_i___ | hmm..okay. Maybe it is our test environment or something | 20:20 |
| kgriffs | and Rails routes == REST | 20:20 |
| flaper87 | mpanetta: not necessarily. There are real-time systems that are not deterministic | 20:20 |
| kgriffs | ;) | 20:20 |
| flaper87 | however, it's a really low percentage | 20:20 |
| flaper87 | kgriffs: LOOOOL | 20:20 |
| malini | jraim_i___: do you have logs or something we can look at ? | 20:20 |
| mpanetta | flaper87: true | 20:20 |
| jraim_i___ | malini: sure. I'll have our guys through up a gist or something | 20:20 |
| flaper87 | real-time just means: You gotta do this stuff in N units of time | 20:21 |
| mpanetta | As a pretty basic definition, yes. | 20:21 |
| flaper87 | then you can add whatever you want to it | 20:21 |
| mpanetta | I used to have a little black book of 'real-time' that I picked up at embedded systems conference years ago... | 20:21 |
| flaper87 | you want it to be deterministic? Then you gotta make it fault-tolerant and blah blah blah | 20:21 |
| mpanetta | I wonder if I still have it somewhere | 20:21 |
| flaper87 | I really recommend reading the bugs I linked at the bottom of that post | 20:22 |
| *** jraim has quit IRC | 20:23 | |
| flaper87 | books* | 20:23 |
| flaper87 | WTF? | 20:23 |
| flaper87 | why did I write bugs? | 20:23 |
| flaper87 | damn, computers are killing me | 20:24 |
| flaper87 | or "bugging me" | 20:24 |
| flaper87 | ok, bad joke | 20:24 |
| * flaper87 gets back to bug triage | 20:24 | |
| mpanetta | Oh, specific to distributed systems though. | 20:24 |
| *** jraim_i___ is now known as jraim | 20:24 | |
| *** jmckind has joined #openstack-marconi | 20:25 | |
| flaper87 | well, the last one "Fault-Tolerant Real-Time Systems: The Problem of Replica Determinism" is really nice | 20:25 |
| flaper87 | it's very expensive | 20:25 |
| flaper87 | but IMHO, it's worth it | 20:25 |
| flaper87 | it explains replica determinism based on a car system | 20:26 |
| flaper87 | engines control, events, fault-tolerance etc | 20:26 |
| mpanetta | That one I may have to look in to. | 20:26 |
| mpanetta | HOLY... Wow it really is expensive! | 20:27 |
| mpanetta | And it was released on my bday... | 20:28 |
| mpanetta | SOmeone should buy it for me :P | 20:28 |
| flaper87 | LOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOL | 20:28 |
| flaper87 | alcabrera: ^^^^ mpanetta is talking to you | 20:28 |
| mpanetta | Why are all the books I like so expensive? :P | 20:29 |
| alcabrera | lol | 20:29 |
| alcabrera | :P | 20:30 |
| alcabrera | for realz, though - books as gifts are amazing (to me) | 20:30 |
| flaper87 | mpanetta: because you'll be rich someday | 20:30 |
| flaper87 | alcabrera: same here! | 20:30 |
| alcabrera | I've been fortunate over the past months that the books that I want are in the $20 - $50 range. | 20:31 |
| alcabrera | So I can get one or two a month without making a dent in the budget. :) | 20:31 |
| mpanetta | flaper87: hah, I donno about that. :P | 20:32 |
| *** malini is now known as malini_afk | 20:38 | |
| *** ayoung has quit IRC | 20:40 | |
| *** ayoung has joined #openstack-marconi | 20:44 | |
| alcabrera | flaper87, kgriffs: Can I set some eyes on this patch between today and tomorrow? https://review.openstack.org/#/c/68267/ . At the moment, our sharding implementation is broken as a result of oslo.cache changes. (unset -> unset_many) | 20:58 |
| flaper87 | alcabrera: erm, I'm sorry? :( | 20:59 |
| *** rossk_ has quit IRC | 21:00 | |
| *** rossk has joined #openstack-marconi | 21:00 | |
| alcabrera | no worries. :) | 21:00 |
| *** jcru_ has quit IRC | 21:01 | |
| alcabrera | I don't think anyone is deploying off of master at the moment, so I'm sure it's okay. | 21:01 |
| *** jcru has joined #openstack-marconi | 21:01 | |
| alcabrera | Have a good night, everyone. :) | 21:07 |
| *** alcabrera has quit IRC | 21:08 | |
| *** cpallares has quit IRC | 21:08 | |
| *** amitgandhi has quit IRC | 21:42 | |
| *** amitgandhi has joined #openstack-marconi | 21:59 | |
| *** amitgandhi has quit IRC | 22:00 | |
| *** amitgandhi has joined #openstack-marconi | 22:01 | |
| flaper87 | kgriffs: ping | 22:16 |
| flaper87 | kgriffs: btw, when talking about the things comming in mongodb 2.6, did I mention they basically re-wrote the query planner? | 22:16 |
| flaper87 | and that there'll be basic index intersection ? | 22:17 |
| kgriffs | flaper87: pong | 22:17 |
| flaper87 | :D :D :D :D :D :D :D | 22:17 |
| kgriffs | ooooooh | 22:17 |
| flaper87 | :) | 22:17 |
| * kgriffs drools all over keyboard | 22:17 | |
| flaper87 | I'm tempted to propose that the minor mongodb version required by Marconi should be 2.6 | 22:18 |
| flaper87 | :P | 22:18 |
| kgriffs | heh | 22:19 |
| kgriffs | I wonder how much ObjectRocket lags | 22:19 |
| kgriffs | (the latest version) | 22:19 |
| *** mpanetta has quit IRC | 22:48 | |
| *** flaper87 is now known as flaper87|afk | 22:50 | |
| amitgandhi | kgriffs: im playing around with pecan, and i see ceilometer uses wsme also | 22:51 |
| amitgandhi | what are your thoughts around wsme in a pecan world | 22:51 |
| *** dustin_ has quit IRC | 22:59 | |
| *** tedross has quit IRC | 23:02 | |
| kgriffs | amitgandhi: the oslo/requirements team recommends using them together | 23:09 |
| kgriffs | but | 23:09 |
| kgriffs | there has been some push-back on WSME for serialization | 23:09 |
| kgriffs | mostly because it doesn't afford API extensions, and so people like Nova can't use it | 23:09 |
| amitgandhi | oic | 23:09 |
| kgriffs | personally, I feel like WSME suffers architecturally from trying to do both SOAP and REST in a single library | 23:10 |
| amitgandhi | yeh as im playing with it it makes me want to take a shower | 23:10 |
| amitgandhi | feels really xmlish | 23:10 |
| kgriffs | it brings along the baggage of strongly-typed, static resources where everything is magically serialized back and forth and pay-no-attention-to-the-man-behind-the-current | 23:10 |
| amitgandhi | prob from the SOAP | 23:10 |
| kgriffs | s/current/curtain | 23:10 |
| kgriffs | amitgandhi: glad to hear I'm not alone! | 23:11 |
| amitgandhi | i guess devs from a java world can relate to it better | 23:11 |
| kgriffs | so, the crux of the issue wrt extensions is the models are defined at compile-time | 23:11 |
| amitgandhi | but i feel like i just stepped back in time 10 years | 23:11 |
| kgriffs | so, you can't just add fields on the fly or anything | 23:12 |
| amitgandhi | yeh that takes away some of the advantages of json | 23:12 |
| kgriffs | even if we were doing media type versioning instead of extensions and major API versioning, WSME would blow up | 23:12 |
| amitgandhi | makes it to strongly typed | 23:12 |
| kgriffs | amitgandhi: LOL re stepped back in time | 23:12 |
| kgriffs | +1 | 23:12 |
| * amitgandhi i have to admin i prefer falcon over pecan too | 23:13 | |
| amitgandhi | *admit | 23:13 |
| kgriffs | heh | 23:13 |
| kgriffs | well, I keep telling people "try both. I don't want to bias you" | 23:13 |
| amitgandhi | falcon seems more intuitive | 23:13 |
| kgriffs | but... | 23:13 |
| kgriffs | I have yet to have someone come back and say "you know what, I like pecan better." | 23:13 |
| amitgandhi | lol | 23:13 |
| kgriffs | but, I'm sure they are out there | 23:13 |
| amitgandhi | ok i have to go now, but thanks for the insight | 23:14 |
| kgriffs | no problem | 23:14 |
| kgriffs | as always, use your own best judgment! | 23:14 |
| * kgriffs is often wrong | 23:14 | |
| amitgandhi | will do | 23:14 |
| amitgandhi | currently im just experimenting | 23:15 |
| kgriffs | cheers | 23:15 |
| amitgandhi | playing around with what feels best | 23:15 |
| kgriffs | +1 | 23:15 |
| amitgandhi | getting the basic frameworks up etc | 23:15 |
| amitgandhi | ok later | 23:15 |
| kgriffs | ciao! | 23:15 |
| *** amitgandhi has quit IRC | 23:15 | |
| *** jergerber has quit IRC | 23:16 | |
| *** jmckind has quit IRC | 23:18 | |
| *** flwang has quit IRC | 23:22 | |
| *** flwang has joined #openstack-marconi | 23:23 | |
| *** kgriffs is now known as kgriffs_afk | 23:30 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!