*** thorst_ has joined #openstack-watcher | 00:00 | |
*** thorst_ has quit IRC | 00:04 | |
*** thorst_ has joined #openstack-watcher | 00:13 | |
*** thorst_ has quit IRC | 00:13 | |
*** acabot has quit IRC | 00:13 | |
*** thorst_ has joined #openstack-watcher | 01:44 | |
*** thorst_ has quit IRC | 01:49 | |
*** wootehfoot has quit IRC | 01:51 | |
*** sferna1_ has quit IRC | 02:03 | |
*** thorst_ has joined #openstack-watcher | 02:09 | |
*** thorst_ has quit IRC | 02:12 | |
*** thorst_ has joined #openstack-watcher | 02:57 | |
*** thorst_ has quit IRC | 02:57 | |
*** thorst_ has joined #openstack-watcher | 03:07 | |
*** thorst_ has quit IRC | 03:07 | |
*** sanfern has joined #openstack-watcher | 04:18 | |
*** hidekazu has joined #openstack-watcher | 04:42 | |
*** yuanying has joined #openstack-watcher | 04:48 | |
*** sanfern has quit IRC | 04:49 | |
*** sanfern has joined #openstack-watcher | 04:50 | |
*** atuly_ has quit IRC | 04:55 | |
*** atuly_ has joined #openstack-watcher | 05:07 | |
*** diga has joined #openstack-watcher | 05:17 | |
*** sanfern has quit IRC | 05:26 | |
*** sanfern has joined #openstack-watcher | 05:27 | |
*** sferna1_ has joined #openstack-watcher | 05:32 | |
*** sanfern has quit IRC | 05:34 | |
*** maestropandy has joined #openstack-watcher | 05:37 | |
maestropandy | Hi All, I am new to watcher, but have interest in learning watcher and do contribution, looking for mentor to guide and i will do best. Can i have someone to help me out | 05:38 |
---|---|---|
*** thorst_ has joined #openstack-watcher | 05:49 | |
*** thorst_ has quit IRC | 05:55 | |
*** adisky_ has joined #openstack-watcher | 05:59 | |
*** danpawlik has quit IRC | 07:38 | |
*** thorst_ has joined #openstack-watcher | 07:51 | |
*** thorst_ has quit IRC | 07:56 | |
*** sferna1__ has joined #openstack-watcher | 08:06 | |
*** sferna1_ has quit IRC | 08:08 | |
*** dtardivel has joined #openstack-watcher | 08:18 | |
*** vincentfrancoise has joined #openstack-watcher | 08:25 | |
*** alexchadin has joined #openstack-watcher | 08:30 | |
*** maestropandy has quit IRC | 08:33 | |
*** alexchadin has quit IRC | 08:38 | |
*** alexchadin has joined #openstack-watcher | 08:39 | |
alexchadin | vincentfrancoise: hi | 08:42 |
vincentfrancoise | alexchadin: morning PTL :p | 08:43 |
alexchadin | :) | 08:43 |
alexchadin | vincentfrancoise: I've started to implement versioned notifications of action plan but then noticed that you have already uploaded your implementation. | 08:44 |
vincentfrancoise | alexchadin: that was notification for "action" I talked to you wasn't it? | 08:45 |
alexchadin | vincentfrancoise: let me see | 08:46 |
vincentfrancoise | alexchadin: that's because I did the action plan part that I proposed you the action one to finish with this | 08:46 |
alexchadin | vincentfrancoise: oh I see | 08:47 |
vincentfrancoise | alexchadin: look at http://eavesdrop.openstack.org/irclogs/%23openstack-watcher/%23openstack-watcher.2017-01-27.log.html#t2017-01-27T09:52:04 I actually gave you the correct link :p | 08:48 |
alexchadin | vincentfrancoise: I had confused these two BPs:) | 08:48 |
vincentfrancoise | they are quite similar anyway | 08:48 |
vincentfrancoise | the spec was made very quickly and is not merged yet so you can change it if you want | 08:49 |
vincentfrancoise | I just wanted to start writing specs so people can more easily pick up some BPs | 08:49 |
alexchadin | vincentfrancoise: It is very clear BP so I'd prefer to implement it first and then to make some corrections to the spec | 08:51 |
vincentfrancoise | alexchadin: yeah the only thing that you shouldn't really be too concerned about is the content of the payloads | 09:17 |
vincentfrancoise | alexchadin: Overall it should be close to being all good but I may have made some mistakes here and there | 09:17 |
alexchadin | vincentfrancoise: Is it necessary to get goal and strategy payloads to every object type? It looks good for audit, but would it be good for action? | 09:21 |
vincentfrancoise | alexchadin: sure | 09:22 |
vincentfrancoise | alexchadin: overall what should be in each notification is the action fields + the directly related objects so action_plan in this case | 09:23 |
vincentfrancoise | alexchadin: then the action_plan field should not contain its related objects, only the UUIDs | 09:24 |
vincentfrancoise | alexchadin: IMHO that should be more than enough | 09:24 |
vincentfrancoise | alexchadin: I've copy/pasted the action plan BP so that's why I was telling you to not take into account the payloads too much | 09:25 |
alexchadin | vincentfrancoise: then there shouldn't be goal and strategy since they are not directly related to action | 09:26 |
vincentfrancoise | alexchadin: exactly | 09:26 |
vincentfrancoise | alexchadin: unless you think it's really needed | 09:27 |
alexchadin | vincentfrancoise: +1 | 09:27 |
*** efoley__ has joined #openstack-watcher | 09:33 | |
*** thorst_ has joined #openstack-watcher | 09:52 | |
*** maestropandy has joined #openstack-watcher | 09:56 | |
*** thorst_ has quit IRC | 09:57 | |
openstackgerrit | Alexander Chadin proposed openstack/watcher-specs: Add fixes to planner-storage-action-plan spec https://review.openstack.org/425261 | 09:59 |
*** maestropandy has quit IRC | 10:00 | |
*** yuanying_ has joined #openstack-watcher | 10:48 | |
*** yuanying has quit IRC | 10:51 | |
*** vincentfrancoise has quit IRC | 11:08 | |
*** yuanying_ has quit IRC | 11:25 | |
*** alexchadin has quit IRC | 11:32 | |
*** sferna1__ has quit IRC | 11:48 | |
*** thorst_ has joined #openstack-watcher | 11:53 | |
*** alexchadin has joined #openstack-watcher | 11:57 | |
*** thorst_ has quit IRC | 11:57 | |
*** diga has quit IRC | 11:59 | |
*** alexchadin has quit IRC | 12:01 | |
*** alexchadin has joined #openstack-watcher | 12:01 | |
*** alexchadin has quit IRC | 12:06 | |
*** thorst_ has joined #openstack-watcher | 12:41 | |
*** vincentfrancoise has joined #openstack-watcher | 13:08 | |
*** sanfern has joined #openstack-watcher | 13:18 | |
*** acabot has joined #openstack-watcher | 13:49 | |
*** alexchadin has joined #openstack-watcher | 13:59 | |
*** alexchadin has quit IRC | 14:00 | |
*** alexchadin has joined #openstack-watcher | 14:00 | |
*** alexchadin has quit IRC | 14:05 | |
chrisspencer | vincentfrancoise: coming back to my bug after getting pulled off for a couple of days and I have some further Qs for you | 15:31 |
vincentfrancoise | chrisspencer: hi | 15:32 |
vincentfrancoise | chrisspencer: go ahead | 15:32 |
chrisspencer | you had mentioned just running the tempest tests to verify the logging done in the nova notifications. I was able to run the latest fixes you had merged, but wasn't seeing any of the notification logging on console or at logs/watcher-decision-engine (or watcher-api, watcher-applier) | 15:33 |
chrisspencer | Did I need to try all of that out on a multinode setup? right now I just have an all in one with watcher installed | 15:34 |
openstackgerrit | Merged openstack/watcher-specs: Add fixes to planner-storage-action-plan spec https://review.openstack.org/425261 | 15:34 |
vincentfrancoise | chrisspencer: an all-in-one should suffice: if you create/update/delete an instance that should be enough | 15:35 |
chrisspencer | vincentfrancoise: hmm. Maybe I'm doing something else wrong. It should be getting logged in the watcher-decision-engine log right? | 15:37 |
vincentfrancoise | chrisspencer: yes, although you need to setup the log level down to DEBUG | 15:38 |
vincentfrancoise | chrisspencer: otherwise you won't be able to see it | 15:38 |
chrisspencer | ah, maybe that is what I'm missing | 15:38 |
vincentfrancoise | chrisspencer: the lovel level is INFO by default | 15:39 |
chrisspencer | vincentfrancoise: where is that conf file located? | 15:39 |
vincentfrancoise | chrisspencer: /etc/watcher/watcher.conf then in [DEFAULT] debug = True | 15:40 |
chrisspencer | vincentfrancoise: looks like already set to true for me | 15:41 |
vincentfrancoise | chrisspencer: that's the first param you can see http://docs.openstack.org/developer/watcher/deploy/conf-files.html | 15:41 |
chrisspencer | vincentfrancoise: does it need to be first? | 15:41 |
vincentfrancoise | chrisspencer: nope, just has to be in the right section | 15:41 |
vincentfrancoise | chrisspencer: do you see any other DEBUG level logs BTW? | 15:42 |
chrisspencer | vincentfrancoise: yes, I just didn't see any being written by watcher.decision_engine.model.notification or more specifically any of the strings we print in that nova.py file | 15:44 |
vincentfrancoise | chrisspencer: did you try with the master branch or your version when you say that? | 15:46 |
vincentfrancoise | chrisspencer: because if you look at the multinode logs for https://review.openstack.org/#/c/422295/ | 15:47 |
vincentfrancoise | chrisspencer: I can defnitely see the logs | 15:47 |
chrisspencer | let me just double check that. Is there a quicker way to restart just the watcher component without having to unstack/stack all over again? | 15:48 |
vincentfrancoise | chrisspencer: just quit the screen and rerun it | 15:49 |
vincentfrancoise | chrisspencer: to rerun the screen you can run: "screen -c ~/devstack/stack-screen" | 15:50 |
vincentfrancoise | chrisspencer: to quit a screen, that's "ctrl+a \" | 15:50 |
vincentfrancoise | chrisspencer: give me a shout if you still struggle ;) | 15:54 |
chrisspencer | vincentfrancoise: thanks. I only want to quit the watcher screen? | 15:55 |
chrisspencer | decision-engine * | 15:55 |
vincentfrancoise | chrisspencer: it's easier to quit it all, but to only restart the watcher services you have to jump to each watcher-* tab and "ctrl +c" to interrupt the process and run the last command | 15:57 |
chrisspencer | vincentfrancoise: ack. So once I've quit the screen, I just restart it the screen -c ~/devstack/stack-screen again? then it should be running my modified code? | 16:00 |
vincentfrancoise | chrisspencer: yes it should reload the code it has at hand in /opt/stack/watcher | 16:01 |
chrisspencer | vincentfrancoise: ahh, I think that is my problem. I probably missed reinstalling my changes to watcher. I'm making the changes in the git repo. I'll trying making the changes in opt/stack. Thanks | 16:03 |
vincentfrancoise | chrisspencer: that's why I explicited the path in case you didn't do that :p | 16:03 |
vincentfrancoise | chrisspencer: I personally use rsync to get my local code into my devstack VMs | 16:04 |
vincentfrancoise | chrisspencer: "rsync -a <WATCHER_DIR>/watcher stack@<IP_ADDRESS>:/opt/stack/watcher/watcher" | 16:05 |
chrisspencer | vincentfrancoise: I want to say I remember a way to install and use the local code. `python setup.py develop` or something like that | 16:05 |
vincentfrancoise | chrisspencer: what do you mean by local as opposed to the /opt/stack/watcher code? | 16:07 |
vincentfrancoise | chrisspencer: as in, where is it located? | 16:07 |
openstackgerrit | Antoine Cabot proposed openstack/watcher-specs: Remove voluptuous and Use JSONSchema as our only JSON validation tool https://review.openstack.org/425204 | 16:07 |
openstackgerrit | Antoine Cabot proposed openstack/watcher-specs: Remove voluptuous and Use JSONSchema as our only JSON validation tool https://review.openstack.org/425204 | 16:10 |
chrisspencer | vincentfrancoise: ah, actually it looks like my code is in /opt/stack/watcher so no need for any of the special workarounds | 16:10 |
openstackgerrit | David TARDIVEL proposed openstack/watcher: Added action_plan.create|update|delete notifs https://review.openstack.org/424556 | 16:14 |
openstackgerrit | David TARDIVEL proposed openstack/watcher: Added action_plan.execution.* actions https://review.openstack.org/424658 | 16:32 |
vincentfrancoise | chrisspencer: does it work now? | 16:33 |
chrisspencer | vincentfrancoise: unfortunately not. the only logs I notice are: http://paste.openstack.org/show/596915/ but nothing in the notification path | 16:47 |
vincentfrancoise | chrisspencer: did you try reverting back to the master branch code and see if there was something coming out? | 16:48 |
chrisspencer | vincentfrancoise: that last comment was a little misleading. There are other items in the log. but I was searching for matching string decision_engine.model | 16:48 |
chrisspencer | vincentfrancoise: will try that | 16:48 |
vincentfrancoise | chrisspencer: oh, one sec | 16:48 |
vincentfrancoise | chrisspencer: if you filter via the "Event" key, do you get something | 16:49 |
vincentfrancoise | ? | 16:49 |
chrisspencer | vincentfrancoise: I'm using vim and doing string search. I see a bunch of items: 2017-01-27 05:55:41.943 DEBUG watcher.common.service [^[[00;36m-] ^[[01;35moslo_reports.file_event_handler = None^[[00m ^[[00;33mfrom (pid=34506) log_opt_values /usr/local/lib/python2.7/dist-packages/oslo_config/cfg.py:2744^[[00m | 16:51 |
chrisspencer | it's catching the file_event_handler | 16:51 |
vincentfrancoise | chrisspencer: you should see logs like http://logs.openstack.org/56/424556/6/check/gate-watcher-dsvm-multinode-ubuntu-xenial-nv/4975d3d/logs/screen-watcher-decision-engine.txt.gz#_2017-01-30_16_42_43_465 | 16:52 |
vincentfrancoise | chrisspencer: this is what we currently have on master | 16:53 |
chrisspencer | vincentfrancoise: what is running in that gate to cause those logs? Right now I'm just doing: ./run_tempest.sh --config watcher-cloud/etc/tempest.conf -N -- watcher | 16:56 |
vincentfrancoise | chrisspencer: wow, this is a question to which I'd rather answer MAGIC | 16:58 |
chrisspencer | maybe I'll just push my changes to the review and let the gate magic tell me if it is working | 16:59 |
vincentfrancoise | chrisspencer: overall, it should be close to what you do though | 16:59 |
vincentfrancoise | chrisspencer: yeah push your code to see what happens | 16:59 |
vincentfrancoise | chrisspencer: also, can you send me a full copy/paste of the logs you have without any search | 17:00 |
vincentfrancoise | chrisspencer: so I can see what's here and what isn't | 17:00 |
chrisspencer | vincentfrancoise: will send you my logs | 17:01 |
vincentfrancoise | chrisspencer: I'll leave the office soon btw so I'll have time going through them today | 17:05 |
chrisspencer | vincentfrancoise: this file is large and is on a remote VM, do you know of an easy way to get it? | 17:08 |
vincentfrancoise | chrisspencer: no need to send me too much, just a subset that should have contained the logs | 17:10 |
vincentfrancoise | chrisspencer: it just has to be without any filtering | 17:10 |
chrisspencer | vincentfrancoise: http://paste.openstack.org/show/596918/ | 17:13 |
vincentfrancoise | chrisspencer: I need more logs to get past the initial log settings printing | 17:15 |
chrisspencer | sorry just noticed that too | 17:15 |
chrisspencer | http://paste.openstack.org/ | 17:15 |
chrisspencer | one sec | 17:16 |
chrisspencer | vincentfrancoise: http://paste.openstack.org/show/596919/ | 17:16 |
vincentfrancoise | chrisspencer: looks normal to me so push the code and we'll see what happens | 17:20 |
vincentfrancoise | chrisspencer: I have to go too so see you tomorrow ;) | 17:20 |
chrisspencer | vincentfrancoise: thanks for all your help | 17:21 |
vincentfrancoise | chrisspencer: you're welcome | 17:21 |
*** vincentfrancoise has quit IRC | 17:26 | |
*** harlowja has joined #openstack-watcher | 18:02 | |
*** alexchadin has joined #openstack-watcher | 18:03 | |
*** absubram has joined #openstack-watcher | 18:04 | |
*** efoley_ has joined #openstack-watcher | 18:04 | |
*** efoley__ has quit IRC | 18:08 | |
*** efoley_ has quit IRC | 18:12 | |
openstackgerrit | Chris Spencer proposed openstack/watcher: Adding additional details to notification logs https://review.openstack.org/422295 | 18:28 |
*** adisky_ has quit IRC | 18:29 | |
*** alexchadin has quit IRC | 18:43 | |
*** dtardivel has quit IRC | 21:14 | |
*** wootehfoot has joined #openstack-watcher | 21:19 | |
*** atuly_ has quit IRC | 21:41 | |
*** atuly_ has joined #openstack-watcher | 21:57 | |
*** thorst_ has quit IRC | 22:36 | |
*** thorst_ has joined #openstack-watcher | 23:06 | |
*** thorst_ has quit IRC | 23:10 | |
*** wootehfoot has quit IRC | 23:22 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!