| *** tongli has quit IRC | 02:12 | |
| *** craigbr has quit IRC | 03:29 | |
| *** garyh has quit IRC | 04:10 | |
| *** garyh has joined #openstack-monasca | 04:34 | |
| *** craigbr has joined #openstack-monasca | 04:50 | |
| *** craigbr has quit IRC | 04:54 | |
| *** craigbr has joined #openstack-monasca | 04:55 | |
| *** aviau has quit IRC | 10:40 | |
| *** aviau has joined #openstack-monasca | 10:52 | |
| *** mikedillion has joined #openstack-monasca | 11:18 | |
| *** mikedillion has quit IRC | 11:35 | |
| *** craigbr has quit IRC | 13:12 | |
| *** openstackgerrit has quit IRC | 13:21 | |
| *** craigbr has joined #openstack-monasca | 13:42 | |
| *** openstackgerrit has joined #openstack-monasca | 13:58 | |
| openstackgerrit | A change was merged to stackforge/monasca-agent: Remove hardcoded version https://review.openstack.org/115468 | 14:06 |
|---|---|---|
| *** tongli has joined #openstack-monasca | 14:11 | |
| sandywalsh | added some new screencasts to the playlist: Notigen event templates and EOD .exists processing | 14:20 |
| sandywalsh | https://www.youtube.com/playlist?list=PLmyM48VxCGaW5pPdyFNWCuwVT1bCBV5p3 | 14:20 |
| *** dschroeder has joined #openstack-monasca | 14:39 | |
| *** cindyo has joined #openstack-monasca | 15:00 | |
| *** mikedillion has joined #openstack-monasca | 15:05 | |
| sandywalsh | cindyo: not quite sure of your question? One set of rules per pipeline (triggering criteria), but that one ruleset could be responsible for thousands of active streams | 15:25 |
| sandywalsh | cindyo: https://github.com/StackTach/sandbox/blob/master/oahu_config.py#L22-L25 is one pipeline definition | 15:27 |
| sandywalsh | and https://github.com/StackTach/sandbox/blob/master/oahu_config.py#L31-L35 is the other (for example) | 15:27 |
| cindyo | sandywash: db_expiry_check in the inmemory driver vs mongodb_driver has a different number of parms. Which one should I be using just to see how it works? | 15:38 |
| cindyo | it seems like only mongodb one will work, since the pipeline is the one sending those parms | 15:39 |
| cindyo | sandywalsh: thanks for the links, that is helpful. the integration test was my guide and it was not matching the event request id key | 15:43 |
| sandywalsh | cindyo: sorry, I didn't get back to the inmemory driver to fix it up yet. | 15:46 |
| sandywalsh | only mongo right now | 15:46 |
| sandywalsh | I'll try and get to that shortly ... working on stackforge headaches :/ | 15:47 |
| cindyo | cindyo: no problem, I realize it’s in development, so it’s nice of you to help us to understand :) Good luck on the stackforge integration :) | 15:48 |
| sandywalsh | heh, thanks ... I'll need it. Let me know if you run into any more issues | 15:48 |
| sandywalsh | the new screencast might be useful too | 15:49 |
| cindyo | what is the link for the screencast? | 15:52 |
| *** mikedillion has quit IRC | 15:58 | |
| tongli | @sandywalsh, will someone send an agenda out for next week meetings? | 15:58 |
| *** mikedillion has joined #openstack-monasca | 16:00 | |
| cindyo | sandywalsh: I get a connection failure to mongo. I guess I have to figure out how to setup a mongdo db | 16:16 |
| cindyo | or can I just change my version of the inmemory db to have the same number of parms? | 16:18 |
| cindyo | i’ll look into that, because i think we want to go the inmemory way | 16:19 |
| sandywalsh | cindyo: the big limitation is the workers don't work with in-memory (since it would span N in-memory instances) | 16:30 |
| sandywalsh | cindyo: I'd need to change it to use shared memory or something | 16:31 |
| cindyo | was thinking of using the streams (in-memory?) to calculate the amount of time from beginning to end and generate a metric with the delta time. Not sure how you normally store this stream of data though | 16:36 |
| cindyo | we would have pipeline processing per tenant, so the same worker would always get that tenant data to add to its pipeline | 16:40 |
| openstackgerrit | Tim Kuhlman proposed a change to stackforge/monasca-thresh: Create a tar.gz file with the package target. https://review.openstack.org/116038 | 16:49 |
| sandywalsh | cindyo: that's a possibility, key off tenant_id and request_id. | 16:51 |
| cindyo | yeah, then the memory sharing is not a concern | 16:52 |
| kuhlmant | I'm getting an error trying to compile monasca-api. | 16:56 |
| kuhlmant | [ERROR] /Users/kuhlmant/src/monasca/api/src/test/java/com/hpcloud/mon/infrastructure/persistence/mysql/AlarmMySQLRepositoryImplTest.java:[50,7] error: class AlarmMySqlRepositoryImplTest is public, should be declared in a file named AlarmMySqlRepositoryImplTest.java | 16:56 |
| kuhlmant | Any ideas on what is wrong? | 16:56 |
| kuhlmant | kuhlmant@mac:~/src/monasca/api (master)$ java -version | 16:57 |
| kuhlmant | java version "1.7.0_25" | 16:57 |
| kuhlmant | Java(TM) SE Runtime Environment (build 1.7.0_25-b15) | 16:57 |
| kuhlmant | Java HotSpot(TM) 64-Bit Server VM (build 23.25-b01, mixed mode)y | 16:57 |
| sandywalsh | cindyo: well, the expired, ready and purge workers wouldn't work. Unless we explicitely call those operations in the yagi handler or add_event() | 16:57 |
| craigbr | kuhlmant: You have a capitalization problem | 16:57 |
| craigbr | Weird thing about a mac is filenames are case insenstive | 16:58 |
| cindyo | oh those are separate workers, then there is a memory sharing problem | 16:58 |
| kuhlmant | Hmm, so what do I do about it? | 16:59 |
| craigbr | I will help you after standup | 16:59 |
| sandywalsh | cindyo: but I think there is a hack workaround we can do | 16:59 |
| cindyo | i’m sure we’ll talk about some of this next week, but our goal was per tenant metrics from streams | 16:59 |
| kuhlmant | Oh wait Cindy isn't talking about my issues, capitilization I will try and fix. | 17:00 |
| cindyo | haha! | 17:00 |
| sandywalsh | cindyo: I'll try and get some time tomorrow to fix that stuff up. Nearly have stackforge done | 17:01 |
| cindyo | how is the communication between mongo db and the expired, ready, and purge workers implemented? | 17:03 |
| cindyo | mongo db streams | 17:04 |
| openstackgerrit | Tim Kuhlman proposed a change to stackforge/monasca-api: Added a tar ball to the package target https://review.openstack.org/116041 | 17:04 |
| openstackgerrit | David Schroeder proposed a change to stackforge/monasca-agent: Fix detection and operation of Kafka plugin https://review.openstack.org/110124 | 17:13 |
| sandywalsh | cindyo: they all talk to the same mongodb | 17:15 |
| sandywalsh | the streams are stored there | 17:15 |
| *** tong_ has joined #openstack-monasca | 17:23 | |
| *** tongli has quit IRC | 17:26 | |
| cindyo | sandywalsh: do the workers poll the db? | 17:34 |
| cindyo | checking for ready, etc | 17:34 |
| sandywalsh | cindyo: yes | 17:34 |
| *** tongli has joined #openstack-monasca | 17:48 | |
| *** tong_ has quit IRC | 17:52 | |
| *** mikedillion has quit IRC | 18:05 | |
| *** mikedillion has joined #openstack-monasca | 18:28 | |
| *** mikedill_ has joined #openstack-monasca | 18:32 | |
| *** mikedillion has quit IRC | 18:33 | |
| *** mikedill_ has quit IRC | 18:34 | |
| *** mikedillion has joined #openstack-monasca | 18:36 | |
| *** mikedillion has quit IRC | 19:05 | |
| *** mikedillion has joined #openstack-monasca | 19:08 | |
| *** tongli has quit IRC | 19:10 | |
| *** cindyo has quit IRC | 19:29 | |
| *** cindyo has joined #openstack-monasca | 19:35 | |
| cindyo | sandywalsh: what would I pass in for the “chunk” parameter to pipeline calls? | 19:53 |
| sandywalsh | anything is fine (like 100) https://github.com/StackTach/sandbox/blob/master/oahu_config.py | 20:12 |
| cindyo | sandywalsh: thanks, sandbox is a good source i see for examples :) will try using mongodb | 20:17 |
| cindyo | sandywalsh: I had 634 streams, it says it checked 100 and none were ready. Is the chunksize the number of streams it checks? I want to force them ready but don’t know how. I tried this: p.process_ready_streams(100,now) | 20:28 |
| cindyo | sandywalsh: I think I had the parameters reversed! | 20:32 |
| cindyo | hmmm maybe not, the pipeline code has this: def process_ready_streams(self, chunk, now=None), but sandbox tests have the reverse | 20:33 |
| *** cindyo has quit IRC | 22:27 | |
| *** cindyo has joined #openstack-monasca | 22:31 | |
| *** craigbr has quit IRC | 22:58 | |
| *** craigbr has joined #openstack-monasca | 22:59 | |
| *** craigbr has quit IRC | 23:03 | |
| *** cindyo has quit IRC | 23:16 | |
| *** dschroeder has quit IRC | 23:27 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!