harlowja | dmsimard is there anything i can do to try to move that output capturing stuff to existence | 00:22 |
---|---|---|
harlowja | even if its in the form of what zuul did? | 00:22 |
harlowja | another one that u might know about https://gist.github.com/harlowja/f6439c083e349356c422c6c964b5f893 | 00:24 |
harlowja | that's been causing `Internal Server Error | 00:25 |
harlowja | ah, maybe its that jinja`>=2.9` | 00:27 |
*** cuongnv has joined #ara | 00:52 | |
dmsimard | harlowja: yeah macros error is jinja>=2.9 | 02:36 |
dmsimard | harlowja: also, no progress on that output capturing stuff | 02:36 |
dmsimard | harlowja: there's simply no way for a pure callback to collect that data -- I've even checked with Tower devs, tower basically just runs ansible-playbook with pexpect so that's how it ends up capturing data | 02:37 |
dmsimard | so the only way to capture data would be to wrap around ansible playbook | 02:38 |
dmsimard | say instead of running ansible-playbook, you run ara-playbook or something | 02:38 |
dmsimard | I don't mean to say it's not doable, but I would much, much rather avoid "breaking" the transparency/seamlessness of ara by requiring users to run something other than ansible-playbook | 02:39 |
dmsimard | There's probably a way but I haven't dug enough yet | 02:40 |
*** cliles has quit IRC | 03:15 | |
*** ChanServ changes topic to "ARA - Ansible Run Analysis: Making playbook runs easier to visualize, understand and troubleshoot | Git: https://github.com/openstack/ara - Docs: https://ara.readthedocs.io/en/latest/ - Latest version: 0.13.0: https://github.com/openstack/ara/releases/tag/0.13.0 | This channel is logged: http://eavesdrop.openstack.org/irclogs/%23ara/" | 04:04 | |
-openstackstatus- NOTICE: Sufficient free space has been reclaimed that jobs are passing again; any POST_FAILURE results can now be rechecked. | 04:04 | |
*** jparrill has joined #ara | 06:15 | |
*** karimb has joined #ara | 07:35 | |
*** cuongnv has quit IRC | 10:11 | |
*** jparrill has quit IRC | 13:43 | |
*** jparrill has joined #ara | 13:44 | |
*** tbielawa has joined #ara | 13:46 | |
*** tbielawa has quit IRC | 13:47 | |
*** tbielawa has joined #ara | 13:48 | |
*** jparrill has quit IRC | 13:51 | |
*** jparrill has joined #ara | 13:51 | |
*** tbielawa has quit IRC | 13:58 | |
*** tbielawa_ has joined #ara | 13:58 | |
*** tbielawa_ is now known as tbielawa|mtg | 14:04 | |
*** jparrill has quit IRC | 14:12 | |
*** jparrill has joined #ara | 14:13 | |
*** viniciusfs has joined #ara | 14:14 | |
*** jparrill_ has joined #ara | 14:16 | |
*** jparrill has quit IRC | 14:17 | |
*** jparrill has joined #ara | 14:58 | |
*** jparrill_ has quit IRC | 14:59 | |
*** tbielawa|mtg is now known as tbielawa | 15:01 | |
*** tbielawa has quit IRC | 15:01 | |
*** viniciusfs has quit IRC | 15:16 | |
*** viniciusfs has joined #ara | 15:31 | |
*** jparrill has quit IRC | 15:59 | |
harlowja | dmsimard ya, the other way is there any way i can upload the logs to ara? | 16:57 |
harlowja | i can capture them, and then just associate them after the fact? | 16:57 |
dmsimard | harlowja: would need a new feature in ara_record to be able to attach data to a specific playbook that you could then run as adhoc command | 17:00 |
dmsimard | Because right now ara_record expects to run from the playbook run context so it knows in which playbook to attach the data. | 17:00 |
dmsimard | It could work. | 17:01 |
harlowja | dmsimard or even outside of ara_record somehow? | 17:23 |
dmsimard | ara_record's the only way to add "dynamic" stuff to ara outside of a regular task from a playbook | 17:23 |
dmsimard | it wouldn't be too hard to add support for ara_record/ara_read as adhoc tasks | 17:25 |
dmsimard | I can take a stab at it | 17:25 |
harlowja | ya, i was even thinking some kind of CLI entrypoint dmsimard | 17:30 |
harlowja | `ara-associate-logs` | 17:30 |
harlowja | that i'd call after running ansible to 'associate the logs' | 17:30 |
harlowja | associate/upload/whatever | 17:31 |
dmsimard | We can make a "helper" command that'd really wrap around ara_record but I don't really want to create something from scratch considering ara_record already exists so you can put whatever you want in it :p | 17:31 |
harlowja | ya, i suppose this would work , ithink | 17:32 |
harlowja | either way | 17:32 |
harlowja | stupid logs, why so hard to do thigns with | 17:33 |
harlowja | someone needs to talk to the ansible devs and tell them 'wtf' | 17:33 |
harlowja | lol | 17:33 |
*** logan- has quit IRC | 17:34 | |
dmsimard | harlowja: it'd look like this: http://paste.openstack.org/raw/610586/ | 17:34 |
dmsimard | (in ad-hoc obv different but you get the point) | 17:34 |
harlowja | ya | 17:34 |
dmsimard | sure, we can do that | 17:35 |
*** logan- has joined #ara | 17:35 | |
harlowja | let me know how asking the ansible devs 'wtf mate why are the runtime logs so hard to get' | 17:38 |
harlowja | lol | 17:38 |
*** openstackgerrit has joined #ara | 17:58 | |
openstackgerrit | David Moreau Simard proposed openstack/ara master: WIP: Allow ara_record and ara_read to run adhoc with specified playbook https://review.openstack.org/467696 | 17:58 |
dmsimard | harlowja: no integration or unit tests yet but testing ^ quickly wfm | 17:58 |
harlowja | cools | 17:59 |
dmsimard | i.e, ansible localhost -m ara_record -a "playbook=42707d21-38b8-41f0-8f8c-c9ab6a8d7c80 key=logs value=lulz" | 17:59 |
harlowja | cools | 18:02 |
dmsimard | harlowja: hm, I'm not sure there is a good way to put a bunch of stuff into "value" from adhoc | 18:24 |
dmsimard | like, hundreds of lines of text | 18:25 |
dmsimard | maybe ara_record needs a 'log' type that picks up a file | 18:25 |
harlowja | i'd be fine with that :) | 18:26 |
* dmsimard sighs | 18:26 | |
dmsimard | Now that we have permalinks for results we can do stuff like this https://gist.github.com/dmsimard/0b6e1720b7fa1f7d0f696fc00b6418ab | 19:21 |
*** openstackgerrit has quit IRC | 19:48 | |
*** karimb has quit IRC | 22:12 | |
harlowja | cool | 22:50 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!