| *** mattybrennan has quit IRC | 00:00 | |
| *** zhurong has joined #openstack-mistral | 01:28 | |
| *** gongysh has joined #openstack-mistral | 01:41 | |
| *** chlong has quit IRC | 01:50 | |
| *** zhurong has quit IRC | 02:30 | |
| *** thrash is now known as thrash|g0ne | 03:17 | |
| *** zhurong has joined #openstack-mistral | 04:00 | |
| *** zhurong has quit IRC | 04:30 | |
| *** zhurong has joined #openstack-mistral | 04:40 | |
| *** sharatss has joined #openstack-mistral | 04:42 | |
| openstackgerrit | Renat Akhmerov proposed openstack/mistral master: Revert "Remove 'sphinxcontrib.autohttp.flask' from sphinx config" https://review.openstack.org/471585 | 05:02 |
|---|---|---|
| openstackgerrit | Merged openstack/mistral master: Stop using abbreviation DSL in document https://review.openstack.org/470925 | 05:13 |
| gongysh | hi | 05:49 |
| gongysh | kong | 05:49 |
| *** jaosorior_away is now known as jaosorior | 06:11 | |
| openstackgerrit | Sharat Sharma proposed openstack/mistral master: Remove the deprecated configuration options https://review.openstack.org/471593 | 06:19 |
| rakhmerov | d0ugal, mgershen, kong: please review https://review.openstack.org/#/c/470928/ | 06:22 |
| rakhmerov | would appreciate if you do it today or tomorrow | 06:22 |
| sharatss | rakhmerov, you told you would take a look at https://review.openstack.org/#/c/455174/ | 06:52 |
| rakhmerov | sharatss: hi, yes, I looked at it but didn't understand yet why it fails | 07:15 |
| sharatss | rakhmerov, same with me.. that is why its stuck there | 07:16 |
| *** amoralej|off is now known as amoralej | 07:18 | |
| *** shardy has joined #openstack-mistral | 07:21 | |
| therve | fultonj, https://bugs.launchpad.net/mistral/+bug/1696128/comments/4 | 07:21 |
| openstack | Launchpad bug 1696128 in Mistral "unable to access a mistral env using mistral.environments_get" [Undecided,New] | 07:21 |
| flaper87 | fultonj: fwiw, I've not solved the ssh | 07:33 |
| flaper87 | issue | 07:33 |
| therve | flaper87, Ah that's interesting :) | 07:35 |
| therve | sharatss, I suspect you're hitting https://bugs.launchpad.net/python-mistralclient/+bug/1693266 | 07:35 |
| openstack | Launchpad bug 1693266 in python-mistralclient "Regression about errors with session usage" [Undecided,New] | 07:35 |
| flaper87 | yeah, I've copied the ssh keys into the mistral's user home | 07:35 |
| therve | sharatss, You need to fix the error handling | 07:35 |
| flaper87 | IIRC | 07:35 |
| flaper87 | I did something, I don't remember what | 07:35 |
| flaper87 | mmh | 07:35 |
| flaper87 | oh, I know what I did | 07:35 |
| flaper87 | just hacked an ansible.cfg and pointed it to the ssh keys | 07:35 |
| flaper87 | I know there's a patch that tries to solve this issue | 07:36 |
| * flaper87 has it in his review/test list | 07:36 | |
| therve | flaper87, What do you think about reusing the validation ssh keys? | 07:36 |
| flaper87 | that was the first thing I attempted to do but for the sake of completing the test I was doing I didn't explore that path further | 07:36 |
| flaper87 | I think we could re-use that, fwiw | 07:37 |
| flaper87 | It's part of the overcloud environment | 07:37 |
| flaper87 | I think | 07:37 |
| flaper87 | if not, we could make it part of the overcloud environment | 07:37 |
| therve | It's in a different env, called ssh_keys | 07:38 |
| flaper87 | ah, that, yeah | 07:39 |
| flaper87 | I think we could totally pass that env to the workflow when it's executed | 07:39 |
| therve | Well it's stored, so you don't really need to pass it. It could make things easier though | 07:40 |
| flaper87 | therve: mmh, I'd prefer "passing" the env to the workflow execution instead of having the workflow assume the env is there | 07:42 |
| flaper87 | but, this being an heat deployment thing, I think we can assume it's there | 07:42 |
| flaper87 | I'll play with this a bit and let you know my results | 07:42 |
| therve | flaper87, Yeah that's fair. I wonder about passing private keys around though | 07:42 |
| therve | sharatss, I have a change that may help on that patch, do you want me to submit it? | 07:51 |
| *** jpich has joined #openstack-mistral | 07:53 | |
| *** gongysh has quit IRC | 08:25 | |
| *** brunograz has quit IRC | 08:32 | |
| *** gongysh has joined #openstack-mistral | 08:33 | |
| *** brunograz has joined #openstack-mistral | 08:40 | |
| *** brunograz1 has joined #openstack-mistral | 08:42 | |
| *** brunograz has quit IRC | 08:42 | |
| *** brunograz1 has quit IRC | 08:47 | |
| *** zhurong has quit IRC | 08:47 | |
| gongysh | kong | 08:48 |
| gongysh | hi kong | 08:48 |
| *** jamielennox is now known as jamielennox|away | 08:50 | |
| gongysh | mistral service-list return error | 08:51 |
| gongysh | ERROR (app) Service API is not supported. | 08:51 |
| gongysh | who can help me out? | 08:54 |
| therve | gongysh, You're using devstack? | 08:57 |
| gongysh | yes | 08:57 |
| therve | I suspect it's because the API is running in Apache | 08:58 |
| therve | That said, I don't know the purpose of service-list | 08:58 |
| therve | Oh no | 08:59 |
| therve | gongysh, Have set up a coordination mechanism? | 08:59 |
| gongysh | therve, do you have a url for how to set up the coordination? | 09:00 |
| therve | Nope | 09:01 |
| *** zhurong has joined #openstack-mistral | 09:01 | |
| *** brunograz has joined #openstack-mistral | 09:03 | |
| gongysh | therve, gongysh@ubuntu64:~$ mistral execution-create my_workflow | 09:03 |
| gongysh | ERROR (app) MessagingTimeout: Timed out waiting for a reply to message ID a53bc693fcc34133860a73ab0ce6f5fa | 09:03 |
| gongysh | how to debug it? | 09:03 |
| gongysh | this is blocking tacker ci. | 09:04 |
| sharatss | therve, please | 09:04 |
| therve | gongysh, https://review.openstack.org/#/c/469806/ | 09:07 |
| gongysh | so you are meaning for us to update the oslo messaging package? | 09:08 |
| gongysh | how does the oslo change the code so careless? | 09:09 |
| *** jamielennox|away is now known as jamielennox | 09:10 | |
| therve | Well it's mostly mistral's fault to use the blocking executor | 09:18 |
| therve | gongysh, You can downgrade oslo.messaging to 5.24 | 09:19 |
| openstackgerrit | Thomas Herve proposed openstack/python-mistralclient master: Use keystoneauth plugins and session instead of keystoneclient https://review.openstack.org/455174 | 09:30 |
| gongysh | therve, it works with 5.24 | 09:34 |
| gongysh | thanks | 09:34 |
| *** zhurong has quit IRC | 09:51 | |
| *** sharatss has quit IRC | 09:55 | |
| *** sharat has joined #openstack-mistral | 09:55 | |
| *** sharat has quit IRC | 10:01 | |
| *** rook has quit IRC | 10:29 | |
| *** rook has joined #openstack-mistral | 10:30 | |
| *** rook is now known as Guest52590 | 10:30 | |
| kong | gongysh: for service api issue, you need to setup mistral deployment with zookeeper or some other tooz backend software | 10:43 |
| gongysh | kong, do we have installation guide for it? | 10:44 |
| kong | nope, you can look for zookeeper doc for installation and tooz doc for reference | 10:46 |
| kong | for mistral, you only need to config backend_url in coordination section | 10:46 |
| kong | mistral just use tooz for implementation | 10:47 |
| *** sam_mistral has joined #openstack-mistral | 10:48 | |
| *** goutham has joined #openstack-mistral | 10:59 | |
| gongysh | what feature do you use tooz? | 11:02 |
| gongysh | just for heath check of service? | 11:02 |
| gongysh | or distributed lock? | 11:02 |
| gongysh | kong | 11:02 |
| kong | gongysh: health check, but we could also leverage other features of tooz if needed | 11:04 |
| *** shardy is now known as shardy_lunch | 11:06 | |
| *** jkilpatr has joined #openstack-mistral | 11:10 | |
| *** Guest52590 is now known as rook | 11:40 | |
| *** shardy_lunch is now known as shardy | 11:43 | |
| *** thrash|g0ne is now known as thrash | 11:44 | |
| *** chlong has joined #openstack-mistral | 12:02 | |
| *** gongysh has quit IRC | 12:19 | |
| *** gongysh has joined #openstack-mistral | 12:19 | |
| fultonj | flaper87: therve looks like i missed the fun around 3 AM my time ^ | 12:19 |
| therve | fultonj, Not so much | 12:20 |
| fultonj | therve: as per your commment you think a new mistraclient ? | 12:21 |
| fultonj | 1. new mistral client | 12:21 |
| fultonj | 2. environment_get(ssh_key) | 12:21 |
| fultonj | 3. use rbrady's new patch to pass ssh-key | 12:21 |
| fultonj | i will try https://review.openstack.org/#/c/462917 | 12:22 |
| therve | fultonj, Yeah that'd be nice. I managed to reproduce the issue locally, and that fixed it | 12:22 |
| *** gongysh has quit IRC | 12:23 | |
| fultonj | thanks therve. i will try the same and update bug | 12:23 |
| *** goutham has quit IRC | 12:29 | |
| *** toure|biab is now known as toure | 13:13 | |
| *** bobh has joined #openstack-mistral | 13:14 | |
| toure | rakhmerov ping | 13:26 |
| toure | rakhmerov just wondering what you thought of the proposed refactored name | 13:27 |
| *** amoralej is now known as amoralej|lunch | 13:31 | |
| *** chlong has quit IRC | 13:43 | |
| fultonj | therve: how did you install your new mistral cient? | 13:46 |
| *** gkadam has joined #openstack-mistral | 13:50 | |
| fultonj | hey mistral people, i am just trying a new mistral client on an undercloud deployed by tripleo-quickstart but didn't seem to get it set up right... | 13:53 |
| fultonj | MistralAction.environments.get failed: <type 'exceptions.AttributeError'>: 'NoneType' object has no attribute 'Client'" | 13:53 |
| fultonj | i had gotten it this way... | 13:53 |
| fultonj | git clone https://github.com/openstack/python-mistralclient ; cd python-mistralclient/ ; sudo pip install -e . | 13:53 |
| *** zhurong has joined #openstack-mistral | 13:54 | |
| mgershen | try the other option | 13:54 |
| fultonj | pip install -r requirements.txt $ python setup.py install | 13:55 |
| * fultonj tries ^ | 13:55 | |
| mgershen | run them as 2 separate commands | 13:56 |
| mgershen | sudo pip install -r requirements.txt | 13:56 |
| fultonj | yeah | 13:56 |
| mgershen | sudo python setup.py install | 13:56 |
| therve | fultonj, I didn't, I changed the code locally | 13:56 |
| fultonj | therve: i had tried that | 13:57 |
| fultonj | adding the 'v' on line 37 of keystone.py | 13:57 |
| therve | fultonj, And? | 13:57 |
| fultonj | therve: was that sufficient for you? nothing else (e.g. pythoncompileall) | 13:57 |
| therve | fultonj, Well restart the mistral services | 13:58 |
| fultonj | therve: ran it twice no comlaint, ran it a 3rd time complaint | 13:58 |
| fultonj | tried that too | 13:58 |
| therve | fultonj, It fails as before? | 13:58 |
| fultonj | therve: yes | 14:00 |
| fultonj | therve: i will re-test | 14:00 |
| fultonj | sudo yum reinstall python-mistralclient | 14:00 |
| d0ugal | I've always had issues if I ever run the python packaging install | 14:01 |
| d0ugal | Manually updating files seems to be safer, but gets messy | 14:01 |
| therve | d0ugal, Hi! Do you know what we can wrt the oslo.messaging issue? | 14:03 |
| d0ugal | therve: I only seen some some messages pass by - whats the issue? | 14:04 |
| therve | d0ugal, The latest release broke the "blocking" executor. The functional gates are kaput | 14:05 |
| therve | Which only block mistralclient of course | 14:05 |
| therve | At any rate, the blocking executor is deprecated, so it'd be nice to use something else | 14:05 |
| d0ugal | therve: latest release? do you mean master? | 14:05 |
| *** amoralej|lunch is now known as amoralej | 14:06 | |
| d0ugal | oh | 14:06 |
| therve | d0ugal, No, both 5.25 and 5.26 | 14:06 |
| d0ugal | the latest release of oslo.messaging | 14:06 |
| therve | Right, sorry | 14:06 |
| d0ugal | I didn't know something was broken :) | 14:06 |
| d0ugal | I shall take a look | 14:06 |
| d0ugal | rakhmerov: are you around? | 14:07 |
| therve | It should be fixed by the next release | 14:07 |
| therve | But mistral needs to move away from that executor anyway, so we may try that | 14:07 |
| d0ugal | Right | 14:07 |
| d0ugal | I have no idea what is involved in doing that tbh | 14:08 |
| therve | When I tried to mention it I was not received super well :) | 14:09 |
| d0ugal | therve: was that back when ddeja was working on related issues? | 14:10 |
| therve | Yeah | 14:10 |
| d0ugal | right | 14:10 |
| d0ugal | I never fully got my head around it then | 14:11 |
| therve | AFAIU we relied on the blocking executor to avoid locking issues with join tasks | 14:12 |
| therve | Which always seems to me like a giant bandaid | 14:12 |
| d0ugal | yeah, that doesn't sound great. | 14:13 |
| *** gkadam has quit IRC | 14:13 | |
| therve | https://review.openstack.org/#/c/356343/12/mistral/engine/rpc_backend/oslo/oslo_server.py | 14:14 |
| thrash | therve: I don't think that's current... | 14:16 |
| therve | thrash, Well, it's still using the blocking executor | 14:17 |
| fultonj | therve: i seem to have broken my env | 14:17 |
| thrash | therve: I don't think so... | 14:17 |
| *** zhurong has quit IRC | 14:17 | |
| therve | thrash, I have the code right there :) | 14:17 |
| thrash | https://github.com/openstack/mistral/blob/master/mistral/executors/executor_server.py#L52 | 14:17 |
| therve | fultonj, That happens | 14:17 |
| thrash | therve: ^^^ | 14:17 |
| thrash | therve: Am I missing something? | 14:18 |
| fultonj | i will rebuild but when it's back i want to clarify how you got the new client | 14:18 |
| therve | thrash, https://github.com/openstack/mistral/blob/master/mistral/engine/engine_server.py#L65 | 14:18 |
| therve | It was moved around | 14:18 |
| therve | Because doing so in the executor was horribly broken | 14:18 |
| therve | Whereas it's just simply broken in the engine | 14:19 |
| thrash | therve: ok... So I misunderstood... We were talking about executor so that's where I was looking. | 14:19 |
| therve | thrash, Ah yeah I'm talking about oslo.messaging executor | 14:19 |
| rbrady | therve: is there a difference between horribly broken and simply broken? :) | 14:19 |
| fultonj | vim /usr/lib/python2.7/site-packages/mistralclient/auth/keystone.py +37 | 14:20 |
| fultonj | add a 'v' as per https://review.openstack.org/#/c/462917/1/mistralclient/auth/keystone.py | 14:20 |
| fultonj | that's it? | 14:20 |
| fultonj | no extra steps? | 14:20 |
| therve | rbrady, Surprising, right? | 14:20 |
| therve | fultonj, And restart the services | 14:20 |
| fultonj | therve: ok | 14:21 |
| fultonj | openstack-mistral-api.service openstack-mistral-engine.service openstack-mistral-executor.service | 14:21 |
| fultonj | ^ restart | 14:21 |
| fultonj | therve: thanks | 14:21 |
| fultonj | i'll be back... | 14:21 |
| *** jaosorior is now known as jaosorior_away | 14:31 | |
| *** sam_mistral has quit IRC | 14:35 | |
| *** chlong has joined #openstack-mistral | 15:27 | |
| therve | rbrady, d0ugal I filled https://bugs.launchpad.net/mistral/+bug/1696469 | 15:32 |
| openstack | Launchpad bug 1696469 in Mistral "Move away from oslo.messaging blocking executor" [Undecided,New] | 15:32 |
| therve | At that point I think waiting for oslo.messaging to release is the quickest fix for the master issue | 15:34 |
| d0ugal | thanks | 15:34 |
| d0ugal | that was my conclusion, because I have no idea how to fix it :-D | 15:34 |
| *** gkadam has joined #openstack-mistral | 16:04 | |
| fultonj | /away | 16:36 |
| *** jpich has quit IRC | 16:36 | |
| *** jkilpatr has quit IRC | 16:36 | |
| prometheanfire | has the mistral team decided on the cycle-with-intermediary issue? | 16:39 |
| *** jkilpatr has joined #openstack-mistral | 16:40 | |
| fultonj | therve: the new client worked for me too. i updated 1696128 | 16:56 |
| *** shardy has quit IRC | 17:18 | |
| fultonj | flaper87: http://paste.openstack.org/show/611755/ | 17:46 |
| *** dtantsur is now known as dtantsur|afk | 17:50 | |
| *** gkadam has quit IRC | 17:59 | |
| *** amoralej is now known as amoralej|off | 18:54 | |
| *** fultonj has quit IRC | 19:28 | |
| *** fultonj has joined #openstack-mistral | 19:35 | |
| *** thrash is now known as thrash|bbl | 20:51 | |
| fultonj | therve: Not sure if you're around, but my workflow gets the key using mistral.environments_get but I get a keystone exception when triggering it via Heat. I made my own little issue with details: https://github.com/fultonj/tripleo-ceph-ansible/issues/11 | 21:01 |
| fultonj | MistralAction.environments.get failed: <class 'keystoneauth1.exceptions.auth.AuthorizationFailure'>: Authorization failed: You are not authorized to perform the requested action: Using trust-scoped token to create another token. Create a new trust-scoped token instead. (HTTP 403) | 21:02 |
| rbrady | fultonj: this is why we have a tripleo base action to create clients | 21:04 |
| rbrady | fultonj: this is also why I've suggested using the action that gets(or generates) the public key by using the action that is already in tripleo-common | 21:05 |
| fultonj | rbrady: the issue is that i need not the public key, but the private key | 21:06 |
| fultonj | and there's no action for the private key | 21:06 |
| fultonj | however.... | 21:06 |
| fultonj | perhaps i should could add an action to get the private key | 21:06 |
| rbrady | fultonj: is the private key in the env vars? | 21:06 |
| fultonj | rbrady: yes | 21:06 |
| rbrady | then copy the action and change it accordingly :) | 21:07 |
| fultonj | rbrady: ok, cool | 21:07 |
| fultonj | i will do that thanks | 21:07 |
| fultonj | FWIW: this had worked http://paste.openstack.org/show/611755/ | 21:07 |
| fultonj | but i understand it's a authorization issue | 21:08 |
| *** fultonj has quit IRC | 21:22 | |
| *** cargonza_ has joined #openstack-mistral | 21:24 | |
| *** cargonza has quit IRC | 21:29 | |
| *** lkannan has quit IRC | 21:29 | |
| *** Kevin_Zheng has quit IRC | 21:29 | |
| *** cargonza_ is now known as cargonza | 21:29 | |
| *** fultonj has joined #openstack-mistral | 21:34 | |
| therve | fultonj, That sounds like something we should be able to fix though | 21:44 |
| fultonj | therve: my isp dropped me temporarily, the last thing i think you said is... | 21:45 |
| fultonj | 17:44 therve: fultonj, That sounds like something we should be able to fix though | 21:45 |
| therve | Cool that's the only thing I said :D | 21:45 |
| fultonj | i assume you mean that i could fix it by creating a new trust-scoped | 21:45 |
| therve | Well, it should be fixed in mistral presumably | 21:46 |
| *** mattybrennan has joined #openstack-mistral | 21:46 | |
| fultonj | i'm trying rbrady's suggestion | 21:46 |
| fultonj | adding a new action like https://github.com/openstack/tripleo-common/blob/master/tripleo_common/actions/validations.py#L26-L44 | 21:47 |
| fultonj | but for GetPrikeyAction | 21:47 |
| *** catintheroof has joined #openstack-mistral | 21:47 | |
| *** bobh has quit IRC | 22:02 | |
| therve | fultonj, https://bugs.launchpad.net/mistral/+bug/1690787 is why I think your issue may get fixed | 22:23 |
| openstack | Launchpad bug 1690787 in Mistral "cron trigger uses trust-scoped token to create another token" [High,New] | 22:23 |
| therve | It's not an isolated problem | 22:23 |
| fultonj | therve: thanks for pointing that out | 22:25 |
| *** chlong has quit IRC | 22:32 | |
| *** jkilpatr has quit IRC | 22:34 | |
| *** catintheroof has quit IRC | 22:46 | |
| fultonj | http://paste.openstack.org/raw/611775/ | 22:46 |
| fultonj | rbrady: ^ any idea what I overlooked while trying to get that new action in? | 22:47 |
| fultonj | i think i did the right things to add a GetPrikeyAction but it doesn't show up when i run 'mistral action-list | grep tripleo.validations.get_' | 22:48 |
| fultonj | - updated validations.py | 22:48 |
| fultonj | - added it to setup.cfg | 22:48 |
| mattybrennan | fultonj: do you need to supply a config file to mistral-db-manage? | 22:53 |
| fultonj | http://paste.openstack.org/raw/611776/ | 22:56 |
| fultonj | mattybrennan: thanks for the suggestion i tried it ^ | 22:56 |
| fultonj | previously i have added extra actions and not needed to do this; i just ran the populate in the same dir | 22:57 |
| mattybrennan | cool, maybe it's not necessary. but I didn't mean that config...I've previously had to supply mistral.conf. | 22:58 |
| fultonj | ah | 22:58 |
| mattybrennan | but I'm not in tripleo or any other openstack stuff, so maybe it "just works" then | 22:59 |
| fultonj | sudo mistral-db-manage --config-file /etc/mistral/mistral.conf populate | 23:00 |
| fultonj | ^ no diff | 23:00 |
| fultonj | but thanks for the suggestion | 23:00 |
| fultonj | got it | 23:05 |
| fultonj | https://github.com/openstack/tripleo-common#action-development | 23:05 |
| fultonj | sudo rm -Rf /usr/lib/python2.7/site-packages/tripleo_common* | 23:05 |
| fultonj | i hadn't dont ^ | 23:05 |
| fultonj | sudo diff -u /usr/lib/python2.7/site-packages/tripleo_common/actions/validations.py /home/stack/tripleo-common/tripleo_common/actions/validations.py | 23:07 |
| fultonj | even though ^ returned nothing after I did the install | 23:07 |
| fultonj | i guess it needed a bigger change to know to load it | 23:07 |
| *** bobh has joined #openstack-mistral | 23:10 | |
| *** openstack has joined #openstack-mistral | 23:14 | |
| *** lkannan has joined #openstack-mistral | 23:20 | |
| *** Kevin_Zheng has joined #openstack-mistral | 23:21 | |
| *** thrash|bbl is now known as thrash | 23:27 | |
| fultonj | thanks rbrady I have updated my patch to use your suggestion. | 23:49 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!