| *** bcoca has quit IRC | 01:33 | |
| *** harlowja has quit IRC | 01:46 | |
| *** jparrill has quit IRC | 05:23 | |
| *** jparrill has joined #ara | 05:24 | |
| *** openstackgerrit has quit IRC | 06:04 | |
| *** gvincent_ has joined #ara | 07:04 | |
| *** gvincent has quit IRC | 07:07 | |
| *** jcl has joined #ara | 07:23 | |
| *** gvincent_ has quit IRC | 07:30 | |
| *** jparrill has quit IRC | 07:40 | |
| *** jparrill has joined #ara | 07:40 | |
| *** jparrill has quit IRC | 08:42 | |
| *** jparrill has joined #ara | 08:42 | |
| *** gvincent has joined #ara | 08:47 | |
| *** gvincent has quit IRC | 09:10 | |
| *** gvincent has joined #ara | 09:11 | |
| *** gvincent has quit IRC | 09:19 | |
| *** gvincent has joined #ara | 09:36 | |
| *** sshnaidm|afk is now known as sshnaidm | 10:30 | |
| *** tbielawa has joined #ara | 12:52 | |
| *** Bakey has joined #ara | 13:24 | |
| *** evrardjp has joined #ara | 13:27 | |
| *** evrardjp_ has quit IRC | 13:30 | |
| *** themurph has joined #ara | 13:40 | |
| *** bcoca has joined #ara | 14:10 | |
| *** bcoca has joined #ara | 14:10 | |
| *** tbielawa is now known as tbielawa|mtg | 14:28 | |
| *** jcl has quit IRC | 14:41 | |
| *** tbielawa|mtg is now known as tbielawa | 15:47 | |
| *** harlowja has joined #ara | 16:12 | |
| ara-slack | <dbpiv> OK, I am back…. Still have this insert into files failure. I thought there was a fix put in 0.15.0 and we shouldn’t have this issue anymore? I am running 0.15.0 with Ansible 2.5.5. and Python 2.7. I know I can walk through and delete each file from the sqlite db, but I was hoping to avoid that. Any ideas? | 16:36 |
|---|---|---|
| *** harlowja has quit IRC | 17:10 | |
| *** tbielawa is now known as tbielawa|lunch | 17:15 | |
| *** tbielawa|lunch is now known as tbielawa | 18:19 | |
| *** cliles has quit IRC | 18:27 | |
| *** sshnaidm has quit IRC | 19:09 | |
| *** tbielawa has quit IRC | 19:18 | |
| *** tbielawa has joined #ara | 19:18 | |
| *** sshnaidm has joined #ara | 19:50 | |
| *** openstackgerrit has joined #ara | 20:26 | |
| openstackgerrit | David Moreau Simard proposed openstack/ara-server master: Remove trailing slashes from URL https://review.openstack.org/577603 | 20:26 |
| ara-slack | <dmsimard> @dbpiv hey, saw your message just now. Are you able to reliably reproduce the issue or something ? | 20:27 |
| ara-slack | <dbpiv> @dmsimard Yes…. It is a daily issue for me at this point. | 20:28 |
| ara-slack | <dmsimard> If this is what I'm thinking about, the workaround I merged was https://github.com/openstack/ara/commit/b85df540af6df608e833d78a7d6e0e16f8deb71b | 20:29 |
| ara-slack | <dbpiv> yeah I pulled the models.py and compared them…. They where the same. | 20:30 |
| ara-slack | <dmsimard> Translated, that boils down to "if you find a playbook without a file, create a fake file and attach it, then return that" | 20:30 |
| ara-slack | <dmsimard> It's kind of dirty but should work, if it doesn't, we're missing something | 20:30 |
| ara-slack | <dmsimard> It should even be self-healing, in the sense that the fix is then persisted in the database | 20:31 |
| ara-slack | <dbpiv> my error would indicate that the file entry is already inserted into the db thus it fails to load. | 20:31 |
| ara-slack | <dmsimard> do you have a stack trace ? | 20:31 |
| ara-slack | <dbpiv> UNIQUE constraint failed: files.playbook_id, files.path [SQL: u’INSERT INTO files (id, playbook_id, path, content_id, is_playbook) VALUES (?, ?, ?, ?, ?)’] [parameters: (‘c6c8727d-9239-4e18-840c-48239d4fe8b5’, u'40dbaa51-8a2e-4d76-a981-22f4173c06f0', u’/Users/dbpiv/Documents/PSFT/Ansible/CacheClear_TEST.yml’, None, 1)] | 20:32 |
| ara-slack | <dbpiv> stack trace? | 20:32 |
| ara-slack | <dbpiv> or tell me how to get a stack trace | 20:32 |
| ara-slack | <dmsimard> yeah, like a python traceback that tells you where the error came from | 20:32 |
| ara-slack | <dmsimard> do you get that when your playbook runs ? after ? from the web interface ? | 20:33 |
| ara-slack | <dbpiv> only from the web interface. | 20:33 |
| ara-slack | <dmsimard> do you run the standalone server or something else ? | 20:33 |
| ara-slack | <dbpiv> with the --debug option on I get a webpage full of Traceback…. | 20:33 |
| ara-slack | <dbpiv> is that what you are looking for? | 20:34 |
| dmsimard | maybe | 20:34 |
| dmsimard | if there's no sensitive info there, can you screenshot it ? | 20:34 |
| ara-slack | <dbpiv> File uploaded https://ara-community.slack.com/files/U6U9U3BV1/FBFE0C3BQ/integrityerror_--_werkzeug_debugger.pdf / https://slack-files.com/T6VAB05L7-FBFE0C3BQ-e753d58268 | 20:35 |
| ara-slack | <dbpiv> nothing sensitive…. just my userid and some paths. | 20:35 |
| dmsimard | ok I think I understand what's going on | 20:37 |
| dmsimard | hang on | 20:37 |
| openstackgerrit | David Moreau Simard proposed openstack/ara master: Improve self-healing of missing playbook file references https://review.openstack.org/578202 | 20:52 |
| openstackgerrit | David Moreau Simard proposed openstack/ara master: Improve self-healing of missing playbook file references https://review.openstack.org/578202 | 20:53 |
| ara-slack | <dmsimard> @dbpiv can you try this patch ? ^ in patch format: https://git.openstack.org/cgit/openstack/ara/patch/?id=043538fa6fedd3fcf1f11ac66609b42bcff434fd | 20:54 |
| ara-slack | <dmsimard> disclaimer: I don't currently have the means to easily reproduce your issue and test the fix | 20:54 |
| ara-slack | <dbpiv> No worries… I can test this. Give me a couple of minutes. | 20:55 |
| ara-slack | <dmsimard> in a nutshell | 20:55 |
| ara-slack | <dmsimard> the fix that had merged prior was to address a race condition where a playbook was created but the playbook's file had not yet been created | 20:56 |
| ara-slack | <dmsimard> it's two operations and if there's something interrupted at a very specific point, you'd end up with an orphaned playbook with no files | 20:56 |
| ara-slack | <dmsimard> now, in the trace you sent me -- the file had been created BUT it had not been flagged as a playbook file *yet*, so this new patch also attempts to recover from that. | 20:57 |
| ara-slack | <dbpiv> and it worked like a charm | 20:57 |
| ara-slack | <dmsimard> so everything auto-healed then ? | 20:57 |
| ara-slack | <dmsimard> thankfully this kind of stuff shouldn't happen in the new 1.0 code paths, operations are more... atomic | 20:59 |
| ara-slack | <dbpiv> yes it loaded… however interestingly there are two run listed at 2018-06-25 08:28'ish from something I was testing…. They are blank with no task and no records. I assume those were the issues. | 20:59 |
| ara-slack | <dbpiv> but yes everything loaded without issue. | 20:59 |
| ara-slack | <dmsimard> do you run your playbooks locally on your machine ? | 21:00 |
| *** tbielawa has quit IRC | 21:00 | |
| ara-slack | <dbpiv> All of my development is on my machine… so yes. but the production runs are on a different server. | 21:00 |
| ara-slack | <dbpiv> the diff server lets me change ansible versions with worrying so I can test. | 21:01 |
| dmsimard | ok, this specific issue is essentially because of interrupted runs... like ctrl+c early enough in the process | 21:01 |
| dmsimard | it's not handled gracefully | 21:01 |
| dmsimard | bcoca: I wonder if Ansible could send a callback hook on interrupt ? I guess it'd be best effort | 21:02 |
| *** themurph has quit IRC | 21:02 | |
| bcoca | we do have an interrupt handler | 21:02 |
| bcoca | but callback might not be available | 21:02 |
| ara-slack | <dbpiv> I might can buy that…. Our standard playbook make extensive use of vars_prompt so it is highly likely that runs are killed at the vars_prompt stage before any contact with the server. | 21:03 |
| ara-slack | <dmsimard> ok so I'll merge that and it'll be available in 0.15.1 (along with some other fixes) | 21:04 |
| ara-slack | <dmsimard> thanks for letting me know about it | 21:04 |
| ara-slack | <dbpiv> hey thanks for the help…. ARA is awesome for developing playbooks. | 21:04 |
| openstackgerrit | David Moreau Simard proposed openstack/ara-server master: Remove trailing slashes from URL https://review.openstack.org/577603 | 21:07 |
| openstackgerrit | David Moreau Simard proposed openstack/ara master: Improve self-healing of missing playbook file references https://review.openstack.org/578202 | 21:19 |
| openstackgerrit | David Moreau Simard proposed openstack/ara master: Improve self-healing of missing playbook file references https://review.openstack.org/578202 | 21:19 |
| *** myoung is now known as myoung|off | 21:56 | |
| openstackgerrit | Merged openstack/ara-server master: Remove trailing slashes from URL https://review.openstack.org/577603 | 22:07 |
| *** Bakey has quit IRC | 23:24 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!