| *** toddjohn has joined #openstack-mistral | 00:35 | |
| *** dprince has quit IRC | 00:44 | |
| *** toddjohn has quit IRC | 01:24 | |
| *** bobh has joined #openstack-mistral | 01:44 | |
| *** bobh has quit IRC | 02:09 | |
| *** rrecio_ has quit IRC | 02:19 | |
| openstackgerrit | Merged openstack/mistral: Added releasenote for https support https://review.openstack.org/349474 | 02:37 |
|---|---|---|
| openstackgerrit | Dawid Deja proposed openstack/mistral: Add state info for synchronous actions run from CLI https://review.openstack.org/349993 | 04:35 |
| openstackgerrit | Merged openstack/mistral: Add target parameters to REST API https://review.openstack.org/339349 | 05:22 |
| rakhmerov | ddeja: hi, I left a comment in https://review.openstack.org/#/c/349993 | 06:17 |
| rakhmerov | be sure to think about what I wrote carefully, data type in case of action result is a tricky thing | 06:18 |
| *** janki has joined #openstack-mistral | 06:23 | |
| hparekh | rakhmerov, Hi | 06:25 |
| rakhmerov | hi | 06:25 |
| hparekh | rakhmerov, Retry policy will only work if task is in error state right ? | 06:25 |
| rakhmerov | yes | 06:26 |
| rakhmerov | right | 06:26 |
| hparekh | strange, One person is claiming that it gets executed even if task action is in success | 06:27 |
| openstackgerrit | Renat Akhmerov proposed openstack/mistral: Towards non-locking model: removing env update from WF controller https://review.openstack.org/349466 | 06:28 |
| openstackgerrit | Renat Akhmerov proposed openstack/mistral: Towards non-locking model: fix obvious workflow controller issues https://review.openstack.org/349457 | 06:28 |
| *** SonalOjha has joined #openstack-mistral | 06:28 | |
| openstackgerrit | Renat Akhmerov proposed openstack/mistral: Towards non-locking model: Add 'unique_key' for delayed calls https://review.openstack.org/349445 | 06:28 |
| rakhmerov | hparekh: hm.. interesting | 06:29 |
| rakhmerov | who is this person and what is he/she trying to do? | 06:29 |
| rakhmerov | we need more details | 06:29 |
| SonalOjha | I am trying to upgrade to mitaka version of mistral from liberty .. I see some schema changes and was trying to use mistral-db-manage cli but that doesnt migrate the data | 06:30 |
| hparekh | rakhmerov, Ah. If task is in success and continue-on tag is evaluated to true then Mistral will retry | 06:30 |
| SonalOjha | what should be done for data migration? | 06:30 |
| openstackgerrit | Merged openstack/mistral: Add _get_fake_client to ironic-inspector actions https://review.openstack.org/349968 | 06:31 |
| rakhmerov | SonalOjha: hi, what data changes you see? | 06:31 |
| rakhmerov | mistral-db-manage upgrade head should help | 06:31 |
| rakhmerov | if if doesn't happen then there might be a bug in migration scripts, please give more details, we'll check it | 06:32 |
| rakhmerov | ideally we need a filed bug | 06:32 |
| SonalOjha | that doesnt help I tried running it with --sql command and what all it shows is create queries | 06:32 |
| rakhmerov | ok, what changes in DB schema are you talking about? | 06:32 |
| SonalOjha | I see executions_v2 table having workflow_id introduced | 06:34 |
| rakhmerov | yes | 06:34 |
| SonalOjha | I tried "mistral-db-manage --config-file /path/to/mistral.conf upgrade 009:011 --sql" | 06:35 |
| SonalOjha | and it shows me just the schema upgrade queries | 06:35 |
| rakhmerov | try without revision numbers | 06:35 |
| rakhmerov | just upgrade head | 06:35 |
| rakhmerov | I'm now looking at our scripts | 06:35 |
| rakhmerov | we do have a script for this change | 06:36 |
| rakhmerov | it should work | 06:36 |
| SonalOjha | I tried that .. it fails as it tries to create table which already exists | 06:36 |
| SonalOjha | do I need to generate the sql and modify it accordingly ? | 06:37 |
| rakhmerov | yes, this is a workaround | 06:37 |
| rakhmerov | I think here's what happened in your case | 06:37 |
| rakhmerov | ok, in your mistral database do you see table alembic_version? | 06:38 |
| SonalOjha | yes I do | 06:38 |
| rakhmerov | do select * from it | 06:38 |
| rakhmerov | what do you see? | 06:38 |
| SonalOjha | nothing in there .. empty | 06:39 |
| SonalOjha | MariaDB [mistral]> select * from alembic_version; Empty set (0.00 sec) | 06:39 |
| rakhmerov | yeah, and that is the reason why it's not working | 06:39 |
| rakhmerov | so | 06:39 |
| rakhmerov | as a workaround you can just set a version in this table manually | 06:39 |
| SonalOjha | ohhkk | 06:39 |
| rakhmerov | if you exactly what version are you at | 06:39 |
| *** ddeja|away is now known as ddeja | 06:39 | |
| rakhmerov | so I guess here's what happened: | 06:40 |
| rakhmerov | initially mistral DB was not created by alembic initially (e.g. you didn't run mistral-db-manage) but DB was created by just the first run if mistral-server | 06:41 |
| rakhmerov | automatically | 06:41 |
| SonalOjha | yeah | 06:41 |
| rakhmerov | and for that reason Alembic (migration tool) thought that we were on revision 0 and that's why it tried to create existing tables | 06:41 |
| rakhmerov | makes sense? | 06:42 |
| SonalOjha | yes it does | 06:42 |
| rakhmerov | so just set a version manually and it should work | 06:42 |
| rakhmerov | hope this helps you :) | 06:42 |
| SonalOjha | how do I find out the version then , I tried looking here https://github.com/openstack/mistral/tree/stable/liberty/mistral/db/sqlalchemy/migration/alembic_migrations/versions | 06:42 |
| SonalOjha | based on the migration script I guessed I am on 009 version | 06:42 |
| SonalOjha | is that correct | 06:42 |
| SonalOjha | ? | 06:42 |
| rakhmerov | simply look our migration scripts and compare with your actual DB state | 06:43 |
| SonalOjha | ok | 06:43 |
| rakhmerov | yes, this should work | 06:43 |
| SonalOjha | great ! would try that out | 06:43 |
| rakhmerov | hparekh: hm.. this might be a bug | 06:43 |
| rakhmerov | with continue-on | 06:43 |
| SonalOjha | so I remember using sync_db script | 06:43 |
| rakhmerov | and btw I remember that nmakhotkin already fixed long ago something like this | 06:44 |
| SonalOjha | shouldnt that do the needful? | 06:44 |
| hparekh | rakhmerov, continue-on is suppose to work with task.state= success right ? | 06:44 |
| rakhmerov | SonalOjha: no, sync_db is used by mistral-db-manage implicitly, no need to use it yourself | 06:44 |
| rakhmerov | we'll remove it later | 06:44 |
| hparekh | rakhmerov, If so we need to improve our documentation. it is little bit confusing | 06:44 |
| SonalOjha | ohhkk | 06:44 |
| rakhmerov | hparekh: let me recall... | 06:45 |
| hparekh | rakhmerov, https://review.openstack.org/#/c/188385/ | 06:46 |
| rakhmerov | hparekh: hm.. in my understanding it's just the opposite of "break-on" | 06:49 |
| rakhmerov | but it's not supposed to be taken into account if task is in SUCCESS | 06:49 |
| rakhmerov | so we have two options: | 06:49 |
| rakhmerov | 1) "break-on" - if need to break retry loop on some condition regardless of iteration count | 06:50 |
| rakhmerov | hm... yeah. this is confusing | 06:51 |
| rakhmerov | seems like I forgot its semantics | 06:51 |
| hparekh | rakhmerov, I guess we should treat continue-on as separate policy. and file a bug for that | 06:52 |
| rakhmerov | no-no, there was a 100% semantics | 06:52 |
| rakhmerov | let me talk to someone, I'll get back to you | 06:52 |
| hparekh | rakhmerov, No issue | 06:53 |
| rakhmerov | 100% clear.. | 06:53 |
| hparekh | ok | 06:53 |
| rakhmerov | hparekh: here's the thing | 07:00 |
| rakhmerov | I recalled :) | 07:00 |
| rakhmerov | this is actually the same as "break-on" but! | 07:01 |
| rakhmerov | it switches the task to SUCCESS instead of ERROR | 07:01 |
| rakhmerov | makes sense? | 07:01 |
| rakhmerov | so it's for a use case when during retry iteration loop we realized that we don't need to continue but we want to mark a task as successful | 07:02 |
| openstackgerrit | Dawid Deja proposed openstack/mistral: Add state info for synchronous actions run from CLI https://review.openstack.org/349993 | 07:06 |
| openstackgerrit | Dawid Deja proposed openstack/mistral: Add state info for synchronous actions run from CLI https://review.openstack.org/349993 | 07:13 |
| hparekh | rakhmerov, So you are saying that we want to continue workflow execution on some condition (continue-on) | 07:19 |
| rakhmerov | yes | 07:19 |
| hparekh | hmmm | 07:19 |
| rakhmerov | and leave this task in SUCCESS state | 07:19 |
| rakhmerov | yeah, it's confusing, we definitely need to improve our docs | 07:20 |
| hparekh | rakhmerov, Then there is bug. See comment in source code https://github.com/openstack/mistral/blob/c7aa89e03d46609a4960e26a880f14cbd8d84b3e/mistral/engine/policies.py#L279 | 07:24 |
| hparekh | rakhmerov, Also it will stop execution if task is in success and continue_on is not specified. https://github.com/openstack/mistral/blob/c7aa89e03d46609a4960e26a880f14cbd8d84b3e/mistral/engine/policies.py#L330 | 07:25 |
| rakhmerov | the last is correct | 07:26 |
| rakhmerov | the first one... thinking | 07:26 |
| openstackgerrit | Dawid Deja proposed openstack/mistral: Add state info for synchronous actions run from CLI https://review.openstack.org/349993 | 07:26 |
| ddeja | hi rakhmerov, I've made changes in my patch | 07:27 |
| rakhmerov | ddeja: ok, thanks :) | 07:27 |
| ddeja | I know what was purpose of the _prepare_output | 07:27 |
| rakhmerov | ddeja: I addressed your comments too in mine ) | 07:27 |
| ddeja | that is why I've added the method to_dict to Result class | 07:27 |
| rakhmerov | hparekh: I am confused again... I guess what I told about semantics of "continue-on" is not quite right.. | 07:28 |
| hparekh | :) | 07:28 |
| ddeja | but I delted it usage in some places while merging by mistake... :| | 07:28 |
| rakhmerov | hparekh: I still need to talk to someone :) It's the person this requirement was coming from.. | 07:28 |
| ddeja | it is fixed now | 07:28 |
| hparekh | Ok take your time :) | 07:28 |
| rakhmerov | ddeja: yeah, it looked like a mistake :) | 07:28 |
| rakhmerov | hparekh: ok ) | 07:28 |
| * ddeja should first take off from bed before git review | 07:29 | |
| d0ugal | What are the risks with increasing the execution_field_size_limit_kb? | 07:29 |
| d0ugal | Was 1024 chosen for a specific reason as the default? | 07:30 |
| rakhmerov | hparekh: yes, what I said is right | 07:32 |
| rakhmerov | if you think there's a bug pls file it | 07:32 |
| d0ugal | ddeja: Did you look at the failures on https://review.openstack.org/#/c/337696/? The change causes some of the other tests to fail - it might just be that they need updated. I'm not sure. | 07:33 |
| rakhmerov | d0ugal: no risk I think, it seemed to enough for investigating all typical errors | 07:33 |
| rakhmerov | d0ugal: I didn't | 07:33 |
| d0ugal | rakhmerov: okay, cool - we hit the limit in TripleO :) | 07:34 |
| rakhmerov | what do you mean? It started getting exceptions? | 07:34 |
| d0ugal | rakhmerov: That question was for ddeja, he pushed an update to the review to fix the test I added | 07:35 |
| rakhmerov | it should be shrinked if the actual size exceeds this limit | 07:35 |
| *** jpich has joined #openstack-mistral | 07:35 | |
| openstackgerrit | Michal Gershenzon proposed openstack/mistral: DB migration to three execution tables and increase some columns https://review.openstack.org/350183 | 07:58 |
| *** shardy has joined #openstack-mistral | 07:59 | |
| *** jistr is now known as jistr|training | 07:59 | |
| hparekh | rakhmerov, Sorry but I am again confused. Take a look https://blueprints.launchpad.net/mistral/+spec/mistral-retry-continue-on | 08:02 |
| rakhmerov | hparekh: sorry, will do later | 08:03 |
| hparekh | rakhmerov, Ok | 08:03 |
| ddeja | d0ugal: oh, I was thinking you will continue the work | 08:07 |
| ddeja | I can take a look on it in an hour | 08:07 |
| *** Ravikiran_K has joined #openstack-mistral | 08:18 | |
| d0ugal | ddeja: No, it's fine. I will look. | 08:18 |
| d0ugal | ddeja: I just didn't want to duplicate effort if you had looked into the new errors | 08:18 |
| d0ugal | Please don't feel like I am asking you to finish my work :) | 08:19 |
| ddeja | d0ugal: OK :) | 08:30 |
| *** gmann has joined #openstack-mistral | 08:48 | |
| gmann | hparekh: ping | 08:49 |
| *** dkushwaha has joined #openstack-mistral | 08:49 | |
| *** jistr|training is now known as jistr | 08:49 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/mistral: Updated from global requirements https://review.openstack.org/350466 | 09:02 |
| *** shardy is now known as shardy_mtg | 09:02 | |
| *** mflobo has joined #openstack-mistral | 09:21 | |
| openstackgerrit | Michal Gershenzon proposed openstack/mistral: DB migration to three execution tables and increase some columns https://review.openstack.org/350183 | 09:34 |
| hparekh | gmann, hi | 09:39 |
| *** evgenyl has quit IRC | 09:42 | |
| *** shardy_mtg is now known as shardy | 09:43 | |
| *** evgenyl has joined #openstack-mistral | 09:44 | |
| ddeja | rakhmerov, all: Just for the record: Thursday and Friday I'm on Openstack Meetup in Wroclaw (Poland). It is around 400km from my city, so I will have limited internet access | 10:16 |
| rakhmerov | ddeja: ok, sure | 10:17 |
| rakhmerov | thanks for letting us know | 10:17 |
| hparekh | ddeja, Ok Thanks | 10:20 |
| hparekh | rakhmerov, Bug has been filled https://bugs.launchpad.net/mistral/+bug/1609335 Take a look when you get some time | 10:28 |
| openstack | Launchpad bug 1609335 in Mistral "Retry get executed even if action is success" [Undecided,New] | 10:28 |
| rakhmerov | sure | 10:29 |
| *** Ravikiran_K has quit IRC | 10:56 | |
| *** mflobo has quit IRC | 11:18 | |
| *** mflobo has joined #openstack-mistral | 11:28 | |
| *** mflobo has quit IRC | 11:36 | |
| openstackgerrit | Merged openstack/mistral: Updated from global requirements https://review.openstack.org/350466 | 11:40 |
| *** dprince has joined #openstack-mistral | 11:45 | |
| jtomasek | rakhmerov: a quick question, every Mistral api response returns an object with properties and all of the values of those properties are json strings. wouldn't it be better if it was objects? | 12:44 |
| jtomasek | rakhmerov: example: http://paste.openstack.org/show/545045/ | 12:44 |
| jtomasek | rakhmerov: I am not sure what the standard is, but every time I want to work with such API response, I need to parse JSON first. e.g. JSON.parse(response.output), JSON.parse(response.input)... | 12:46 |
| rakhmerov | jtomasek: yes, it's basically a side-effect that they are strings | 12:46 |
| rakhmerov | yeah, sorry for this | 12:46 |
| jtomasek | rakhmerov: ok, so there is nothing we can do to improve it, right? | 12:46 |
| rakhmerov | btw, it's rather an intermediate state, we didn't finish something in the implementation | 12:46 |
| rakhmerov | we can but it's not finished yet | 12:46 |
| jtomasek | rakhmerov: sounds good! | 12:47 |
| jtomasek | rakhmerov: thanks | 12:47 |
| rakhmerov | np | 12:47 |
| jtomasek | rakhmerov: is there a blueprint to watch? | 12:47 |
| rakhmerov | no I guess | 12:48 |
| rakhmerov | I just know it ) | 12:48 |
| rakhmerov | there's too many things like this and I don't file BPs for some of them | 12:48 |
| rakhmerov | could you please file a bug for this? | 12:48 |
| jtomasek | rakhmerov: ok, sure | 12:48 |
| rakhmerov | thank you | 12:49 |
| jtomasek | rakhmerov: https://bugs.launchpad.net/mistral/+bug/1609402 | 12:56 |
| openstack | Launchpad bug 1609402 in Mistral "api responses nested properties are JSON strings" [Undecided,New] | 12:56 |
| jpich | jtomasek, rakhmerov: I agree it'd be more convenient though changing it will break existing consumers now? | 13:03 |
| * jpich subscribes to bug as well | 13:03 | |
| jtomasek | jpich: it will, It will have to be taken care of in versioning somehow. or better go through the deprecation process if possible | 13:04 |
| jpich | jtomasek: Gotcha! | 13:04 |
| rakhmerov | jtomasek: what version of Mistral are you using? | 13:07 |
| jtomasek | rakhmerov: 2.0.1 | 13:08 |
| rakhmerov | ooh, ok | 13:08 |
| rakhmerov | in fact, we already broke it a little bit in newton | 13:08 |
| jtomasek | rakhmerov: in which case? is that version old? we're getting Mistral from RDO packages, which are (afaik) udpated on patch basis | 13:09 |
| rakhmerov | no, it's not too old, it's Mitaka | 13:13 |
| rakhmerov | but we've made some huge changes (no worries, they are to the better and are mostly about the implementation) in Newton | 13:13 |
| rakhmerov | the format of how we display errors may be slightly different now | 13:14 |
| rakhmerov | although it is still represented as a string | 13:14 |
| jtomasek | rakhmerov: cool, I think the package we use is fairly recent, it probably just uses old version tag because newton has not been done yet | 13:14 |
| rakhmerov | yeah | 13:16 |
| rakhmerov | jtomasek: just for your information, we're now working on tremendous performance improvements | 13:16 |
| rakhmerov | we've already made some | 13:16 |
| rakhmerov | so Newton version will be orders or magnitude faster than Mitaka | 13:16 |
| rakhmerov | and more resistant to concurrent problems | 13:17 |
| jtomasek | rakhmerov: great! Is is about running large amount of workflows at once as discussed on the Summit? | 13:17 |
| jtomasek | rakhmerov: ah, that sounds very good | 13:18 |
| rakhmerov | and this too | 13:18 |
| rakhmerov | right now we have serious problems with highly concurrent workflows | 13:18 |
| rakhmerov | even if it's just one workflow | 13:18 |
| rakhmerov | for example, if in a workflow we have >50 tasks running in parallel it will be very slow | 13:19 |
| jtomasek | I see | 13:19 |
| *** tonytan4ever has joined #openstack-mistral | 13:32 | |
| *** Ravikiran_K has joined #openstack-mistral | 13:33 | |
| *** toddjohn has joined #openstack-mistral | 13:49 | |
| *** krot_sickleave is now known as krotscheck | 14:00 | |
| *** janki has quit IRC | 14:11 | |
| *** janki has joined #openstack-mistral | 14:34 | |
| *** bobh has joined #openstack-mistral | 14:43 | |
| *** bobh has quit IRC | 14:49 | |
| *** dprince has quit IRC | 15:10 | |
| *** rrecio has joined #openstack-mistral | 15:11 | |
| *** SonalOjha has quit IRC | 15:16 | |
| *** brian_price has joined #openstack-mistral | 15:34 | |
| *** mflobo has joined #openstack-mistral | 15:36 | |
| *** jpeeler has quit IRC | 15:43 | |
| *** brunograz has quit IRC | 15:50 | |
| *** jistr is now known as jistr|biab | 16:06 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/mistral: Updated from global requirements https://review.openstack.org/350670 | 16:10 |
| *** janki has quit IRC | 16:12 | |
| *** brian_price has quit IRC | 16:18 | |
| *** brian_price has joined #openstack-mistral | 16:19 | |
| *** brian_price has left #openstack-mistral | 16:21 | |
| *** jistr|biab is now known as jistr | 16:22 | |
| *** brian_price has joined #openstack-mistral | 16:28 | |
| *** dprince has joined #openstack-mistral | 16:33 | |
| *** bobh has joined #openstack-mistral | 16:46 | |
| *** bobh has quit IRC | 16:51 | |
| *** krotscheck is now known as kro_focused | 16:53 | |
| *** mgershen has quit IRC | 16:55 | |
| *** openstackgerrit_ has joined #openstack-mistral | 17:06 | |
| *** openstackgerrit_ has quit IRC | 17:08 | |
| *** shardy has quit IRC | 17:44 | |
| *** bobh has joined #openstack-mistral | 17:47 | |
| *** bobh has quit IRC | 17:52 | |
| *** harlowja has quit IRC | 17:58 | |
| *** harlowja has joined #openstack-mistral | 18:00 | |
| *** tonytan4ever has quit IRC | 18:08 | |
| *** bobh has joined #openstack-mistral | 18:11 | |
| *** toddjohn has quit IRC | 18:11 | |
| *** toddjohn has joined #openstack-mistral | 18:13 | |
| *** toddjohn_ has joined #openstack-mistral | 18:14 | |
| *** toddjohn has quit IRC | 18:17 | |
| *** jpich has quit IRC | 18:22 | |
| *** mflobo has quit IRC | 18:27 | |
| *** toddjohn_ has quit IRC | 18:52 | |
| *** toddjohn has joined #openstack-mistral | 18:54 | |
| *** tonytan4ever has joined #openstack-mistral | 19:08 | |
| *** tonytan4ever has quit IRC | 19:13 | |
| *** clenimar has quit IRC | 19:34 | |
| *** clenimar has joined #openstack-mistral | 19:35 | |
| *** tonytan4ever has joined #openstack-mistral | 19:49 | |
| *** Ravikiran_K has quit IRC | 19:57 | |
| *** dprince has quit IRC | 19:58 | |
| *** toddjohn has quit IRC | 20:22 | |
| *** toddjohn has joined #openstack-mistral | 20:22 | |
| *** toddjohn_ has joined #openstack-mistral | 20:24 | |
| *** toddjohn has quit IRC | 20:27 | |
| *** toddjohn_ has quit IRC | 20:29 | |
| openstackgerrit | Dougal Matthews proposed openstack/mistral: Fix the use of both adhoc actions and "with-items" in workflows https://review.openstack.org/337696 | 20:43 |
| *** toddjohn has joined #openstack-mistral | 20:59 | |
| *** bobh has quit IRC | 22:47 | |
| *** toddjohn has quit IRC | 23:09 | |
| *** brian_price has quit IRC | 23:18 | |
| *** tonytan4ever has quit IRC | 23:33 | |
| *** bobh has joined #openstack-mistral | 23:52 | |
| *** toddjohn has joined #openstack-mistral | 23:54 | |
| *** rrecio has quit IRC | 23:54 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!