| *** snadge has joined #openstack-ansible | 00:24 | |
| snadge | if i want to run a python script that uses python-novaclient.. should I install it on the controller host, or within the utility container? | 00:25 |
|---|---|---|
| snadge | does it matter | 00:25 |
| *** tosky has quit IRC | 01:02 | |
| *** prometheanfire has quit IRC | 01:28 | |
| *** prometheanfire has joined #openstack-ansible | 01:32 | |
| *** macz_ has joined #openstack-ansible | 02:24 | |
| *** macz_ has quit IRC | 02:29 | |
| *** evrardjp_ has joined #openstack-ansible | 02:50 | |
| *** evrardjp has quit IRC | 02:50 | |
| * prometheanfire migrated a down instance and now, again I cannot confirm... | 03:10 | |
| prometheanfire | I can't even revert, wtf | 03:19 |
| *** evrardjp_ has quit IRC | 03:33 | |
| *** evrardjp has joined #openstack-ansible | 03:33 | |
| *** shyamb has joined #openstack-ansible | 04:09 | |
| *** macz_ has joined #openstack-ansible | 04:46 | |
| *** macz_ has quit IRC | 04:51 | |
| *** shyamb has quit IRC | 05:09 | |
| *** jamesdenton has quit IRC | 05:38 | |
| *** jamesdenton has joined #openstack-ansible | 05:39 | |
| *** macz_ has joined #openstack-ansible | 05:43 | |
| *** macz_ has quit IRC | 05:48 | |
| *** yasemind has joined #openstack-ansible | 05:54 | |
| *** noonedeadpunk has joined #openstack-ansible | 07:12 | |
| *** miloa has joined #openstack-ansible | 07:17 | |
| *** miloa has quit IRC | 07:28 | |
| jrosser | morning | 07:56 |
| LowKey | Hi jrosser, morning | 08:00 |
| jrosser | good morning | 08:00 |
| LowKey | how are you ? | 08:00 |
| jrosser | good, thankyou :) | 08:02 |
| LowKey | great :) | 08:02 |
| noonedeadpunk | mornings! | 08:05 |
| * noonedeadpunk back | 08:06 | |
| noonedeadpunk | what's happening forum? | 08:07 |
| *** jamesdenton has quit IRC | 08:13 | |
| *** jamesdenton has joined #openstack-ansible | 08:14 | |
| *** andrewbonney has joined #openstack-ansible | 08:16 | |
| *** rpittau|afk is now known as rpittau | 08:22 | |
| *** priteau has joined #openstack-ansible | 08:25 | |
| ebbex | mornin' | 08:43 |
| *** tosky has joined #openstack-ansible | 08:53 | |
| openstackgerrit | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_neutron master: Remove third-party OVS NSH support in favor of built-in support https://review.opendev.org/c/openstack/openstack-ansible-os_neutron/+/780117 | 08:58 |
| *** SiavashSardari has joined #openstack-ansible | 09:01 | |
| admin0 | morning | 09:02 |
| admin0 | anyone know of this issue or seen it before? any cli command i type , even openstack token issue gives => Could not find project: b78683d67f8d4253b624db55b23d1393 | 09:03 |
| noonedeadpunk | is project domain is valid in openrc? | 09:05 |
| openstackgerrit | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_neutron master: Prevent neutron-l3-agent restart killing metadata service https://review.opendev.org/c/openstack/openstack-ansible-os_neutron/+/775388 | 09:06 |
| openstackgerrit | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_neutron stable/victoria: Add integrated tests various Neutron deployment scenarios https://review.opendev.org/c/openstack/openstack-ansible-os_neutron/+/780457 | 09:07 |
| ebbex | noonedeadpunk: https://review.opendev.org/c/openstack/openstack-ansible-os_neutron/+/780117/1..2 is the extra zypper/opensuse cleanup intentional here? | 09:08 |
| noonedeadpunk | it's rebase on master. so suse cleanup is laready merged | 09:09 |
| noonedeadpunk | https://opendev.org/openstack/openstack-ansible-os_neutron/commit/d8c643eeb2c02fb8c450be71fff5f9fddcb75481 | 09:09 |
| noonedeadpunk | I was just resolving merge conflict there | 09:13 |
| ebbex | ah, i think i was just expecting to see the 'rebase' tag :) The total patch looks good :) | 09:13 |
| noonedeadpunk | it's there only when you rebase with gerrit... and you can't do this while having conflict so yeah:( | 09:14 |
| ebbex | that explains it ;) | 09:16 |
| openstackgerrit | Dmitriy Rabotyagov proposed openstack/openstack-ansible master: Use proper distro path for senlin https://review.opendev.org/c/openstack/openstack-ansible/+/777384 | 09:20 |
| openstackgerrit | Merged openstack/openstack-ansible-ops master: It's always systemd. Remove references to upstart https://review.opendev.org/c/openstack/openstack-ansible-ops/+/779066 | 09:33 |
| openstackgerrit | Merged openstack/ansible-hardening master: Use ansible_facts[] instead of fact variables https://review.opendev.org/c/openstack/ansible-hardening/+/778483 | 09:36 |
| openstackgerrit | Dmitriy Rabotyagov proposed openstack/openstack-ansible master: Fix serialized playbook runs https://review.opendev.org/c/openstack/openstack-ansible/+/752040 | 10:13 |
| openstackgerrit | Merged openstack/openstack-ansible-os_neutron master: powervm: drop custom code https://review.opendev.org/c/openstack/openstack-ansible-os_neutron/+/780116 | 10:41 |
| *** openstackgerrit has quit IRC | 10:43 | |
| *** gshippey has joined #openstack-ansible | 12:23 | |
| jrosser | feels like time to have an etherpad for wallaby to-do list https://etherpad.opendev.org/p/osa-wallaby | 12:36 |
| *** SiavashSardari has quit IRC | 12:36 | |
| mgariepy | noonedeadpunk, jrosser the issue with ovs test on os_neutron+victoria. can it be related to : https://github.com/openstack/openstack-ansible-os_neutron/blob/stable/victoria/tests/ansible-role-requirements.yml | 12:36 |
| mgariepy | the version of tempest installed in the ovs build vs the one in aio build isn't the same. | 12:37 |
| *** yasemind has quit IRC | 12:37 | |
| *** SiavashSardari has joined #openstack-ansible | 12:37 | |
| jrosser | mgariepy: i wonder if it's worth backporting the re-work of all that test stuff jamesdenton overhauled to V | 12:38 |
| mgariepy | tempest-26.1.1.dev24 in the ovs build: and tempest-25.0.1 | 12:38 |
| mgariepy | in the aio one. | 12:38 |
| jrosser | right, on stable branches the AIO takes tempest from pip packages | 12:38 |
| jrosser | and quite possibly from tip-of-master git for the wierd ovs job on os_neutron | 12:39 |
| jrosser | it's kind of questionable putting effort in to fix up the functional tests vs. use this https://github.com/openstack/openstack-ansible-os_neutron/commit/69845d63a023f58beef172a03e48400ed3671742 | 12:40 |
| jrosser | unless it's a trivial fix in openstack-ansible-tests | 12:40 |
| jrosser | it's probably this making it use a git source https://github.com/openstack/openstack-ansible-tests/blob/stable/victoria/test-vars.yml#L360 | 12:42 |
| *** vakuznet has joined #openstack-ansible | 12:50 | |
| jamesdenton | good morning | 12:52 |
| noonedeadpunk | I think we should backport and patch already there https://review.opendev.org/c/openstack/openstack-ansible-os_neutron/+/780457 | 12:54 |
| jamesdenton | you may also need to backport this? https://review.opendev.org/c/openstack/openstack-ansible/+/778008/ | 12:54 |
| noonedeadpunk | yeah, totally, good catch | 12:55 |
| noonedeadpunk | ok, gerritbot is dead | 13:08 |
| mgariepy | noonedeadpunk, jamesdenton we might want to have bionic tests on victoria | 13:13 |
| noonedeadpunk | uh.... | 13:14 |
| noonedeadpunk | that is tough... | 13:14 |
| *** spatel_ has joined #openstack-ansible | 13:14 | |
| *** jamesdenton has quit IRC | 13:14 | |
| noonedeadpunk | I don;'t think we should test every scenario for both bionic and focal... | 13:14 |
| noonedeadpunk | from perspective of CI resource allocation... | 13:15 |
| *** jamesdenton has joined #openstack-ansible | 13:15 | |
| mgariepy | well . good point. | 13:15 |
| *** spatel_ has quit IRC | 13:15 | |
| *** spatel_ has joined #openstack-ansible | 13:16 | |
| noonedeadpunk | maybe we can add at least OVS... | 13:16 |
| mgariepy | ovs lxc/metal or just metal ? | 13:17 |
| mgariepy | perhaps only one would be good enoug. | 13:18 |
| noonedeadpunk | yeah, let's set metal | 13:18 |
| mgariepy | cool Thanks a lot. | 13:19 |
| *** yasemind has joined #openstack-ansible | 13:38 | |
| vakuznet | i have an issue of disappeared erlang package. https://review.opendev.org/c/openstack/openstack-ansible-rabbitmq_server/+/780562 but i can't find 22.1.5 either | 13:41 |
| *** spatel_ has quit IRC | 13:48 | |
| jrosser | vakuznet: it's on the third page of this? https://packagecloud.io/app/rabbitmq/erlang/search?dist=&filter=all&page=3&q= | 13:48 |
| jrosser | https://packagecloud.io/rabbitmq/erlang/packages/el/8/erlang-23.1.5-1.el8.x86_64.rpm | 13:49 |
| jrosser | oh whoops | 13:49 |
| vakuznet | thei is el8 | 13:49 |
| jrosser | *23*! | 13:49 |
| *** spatel_ has joined #openstack-ansible | 13:49 | |
| jrosser | https://packagecloud.io/app/rabbitmq/erlang/search?q=&filter=all&filter=all&dist=el%2F7 | 13:50 |
| jrosser | noonedeadpunk: ^ which erlang did you intend? | 13:50 |
| vakuznet | erlang versions: http://paste.openstack.org/show/803563/ | 13:52 |
| vakuznet | 22.1.* are dropped? | 13:52 |
| noonedeadpunk | vakuznet: yep, on centos they are | 14:00 |
| noonedeadpunk | I just posted https://review.opendev.org/c/openstack/openstack-ansible-rabbitmq_server/+/780562 | 14:00 |
| noonedeadpunk | jrosser: you mean https://review.opendev.org/c/openstack/openstack-ansible-rabbitmq_server/+/780565 ? | 14:01 |
| noonedeadpunk | I wanted to try out latest one | 14:01 |
| jrosser | no this one https://review.opendev.org/c/openstack/openstack-ansible-rabbitmq_server/+/780562/1/vars/redhat.yml | 14:02 |
| noonedeadpunk | uh, yes, I looked at 23.... | 14:05 |
| jrosser | we have to keep the rabbit/erlang versions compatible iirc | 14:06 |
| noonedeadpunk | yep, we do.... but I think we can use 22.2.8... | 14:06 |
| noonedeadpunk | need to double check | 14:06 |
| jrosser | https://www.rabbitmq.com/which-erlang.html | 14:07 |
| noonedeadpunk | 22.x is recommended | 14:07 |
| noonedeadpunk | then let's bump to latest 22?:) | 14:07 |
| jrosser | sounds like a plan | 14:08 |
| *** openstackgerrit has joined #openstack-ansible | 14:10 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_aodh master: Use ansible_facts[] instead of fact variables https://review.opendev.org/c/openstack/openstack-ansible-os_aodh/+/780594 | 14:10 |
| openstackgerrit | Dmitriy Rabotyagov proposed openstack/openstack-ansible-rabbitmq_server stable/train: Bump erlang version for CentOS 7 https://review.opendev.org/c/openstack/openstack-ansible-rabbitmq_server/+/780562 | 14:11 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_barbican master: Use ansible_facts[] instead of fact variables https://review.opendev.org/c/openstack/openstack-ansible-os_barbican/+/780595 | 14:18 |
| noonedeadpunk | jrosser: what a mess we made to the https://github.com/ansible/ansible/issues/73654 | 14:19 |
| jrosser | i know :) | 14:19 |
| jrosser | i had no idea that was a thing | 14:20 |
| noonedeadpunk | yeash, me neither | 14:20 |
| jrosser | i have sooooo many more patches to push | 14:20 |
| noonedeadpunk | I think they should feel the pain haha | 14:20 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_blazar master: Use ansible_facts[] instead of fact variables https://review.opendev.org/c/openstack/openstack-ansible-os_blazar/+/780596 | 14:21 |
| seanp | When I deploy a new instance via the CLI, it shows up in "Error" state. In Horizon the only error is something about exceeding the max retries, doesn't seem to be pertinent. In the syslog I don't get any (recent) hits for Error or Fail. Is there something else I could be looking for? Nothing that comes up in the syslog immediately after deploying and going into the error state looks interesting | 14:22 |
| seanp | I have deployed via Horizon with the same effect | 14:22 |
| noonedeadpunk | jrosser: so our plan is to set `INJECT_FACTS_AS_VARS = False` at the end of the day ? | 14:23 |
| jrosser | it would be nice to do that | 14:23 |
| jrosser | but on the other hand minimising the facts collection seems like a seperate thing | 14:23 |
| *** macz_ has joined #openstack-ansible | 14:24 | |
| jrosser | because it goes from many 100's kb down to 5kb with the minimal fact gathering | 14:24 |
| jrosser | so being clear that theres two things going on is important, and i think they're a bit mixed up at the moment | 14:24 |
| jrosser | i've separated it into two patches, this one https://review.opendev.org/c/openstack/openstack-ansible/+/778396 | 14:26 |
| jrosser | and the one underneath | 14:26 |
| noonedeadpunk | uh, external dependencies.... | 14:27 |
| jrosser | yeah, indeed | 14:27 |
| jrosser | i've got forks for the most important set on github | 14:27 |
| jrosser | not got round to making PR yes | 14:27 |
| jrosser | *yet | 14:27 |
| noonedeadpunk | feels like ceph-ansible would be huge blocker here | 14:28 |
| *** macz_ has quit IRC | 14:28 | |
| jrosser | yeah, so only for the INJECT_FACTS_AS_VARS bit | 14:28 |
| noonedeadpunk | yep | 14:28 |
| jrosser | we should be able to handle pretty much everything for minimal facts gathering | 14:28 |
| jrosser | i wonder if i should make two completely seperate stacks of patches for that | 14:29 |
| spatel_ | noonedeadpunk welcome back!!! where are we with this patch - https://review.opendev.org/c/openstack/openstack-ansible/+/777384 | 14:29 |
| noonedeadpunk | jrosser: considfering amount of patches I'd say you can merge them? | 14:31 |
| openstackgerrit | Dmitriy Rabotyagov proposed openstack/openstack-ansible master: Use proper distro path for senlin https://review.opendev.org/c/openstack/openstack-ansible/+/777384 | 14:35 |
| noonedeadpunk | spatel_: thanks) let's see how it goes that way | 14:37 |
| spatel_ | Thank you | 14:38 |
| noonedeadpunk | jrosser: regarding https://review.opendev.org/c/openstack/openstack-ansible/+/734883 | 14:38 |
| jrosser | yes i saw the comment | 14:39 |
| jrosser | -ENOTIME :) | 14:39 |
| noonedeadpunk | it seems it's working nicely now. but, um.... I see that there's no neat place for `systemd-python` but that's hardly a requirement.... | 14:39 |
| *** spatel_ has quit IRC | 14:40 | |
| jrosser | that requirements file controls what gets installed in /opt/ansible-runtime, nowhere else? | 14:40 |
| noonedeadpunk | mm, yes | 14:40 |
| *** SiavashSardari has quit IRC | 14:41 | |
| noonedeadpunk | you mean that whatever? | 14:42 |
| noonedeadpunk | since it's only ansible runtime - it won't hurt? | 14:42 |
| *** spatel_ has joined #openstack-ansible | 14:56 | |
| *** spatel_ has quit IRC | 14:59 | |
| jrosser | i guess, we use the requirements file not really as a traditional python requirements file | 15:03 |
| jrosser | but if we can install it in the test scripts somewhere, thats fine too | 15:03 |
| *** macz_ has joined #openstack-ansible | 15:09 | |
| *** spatel has joined #openstack-ansible | 15:14 | |
| jrosser | noonedeadpunk: sorry been in meetings | 16:01 |
| jrosser | we could have another file of packages to install only for in zuul | 16:01 |
| jrosser | or maybe just having it called requirements.txt is confusing becasue as you say it kind of isnt a requirements file for normal python use | 16:01 |
| noonedeadpunk | which should be test-requirements, but we use them for linters only... | 16:02 |
| jrosser | right | 16:02 |
| jrosser | so it's all a bit messy | 16:02 |
| noonedeadpunk | yeah, requirements.txt are frustrating, and I'm not sue if we want frustrate ppl even more by adding there ci-only things... | 16:02 |
| noonedeadpunk | should we add gate-requirements.txt?:) | 16:03 |
| jrosser | well hmm https://opendev.org/openstack/openstack-ansible/src/branch/master/scripts/scripts-library.sh#L86 | 16:03 |
| jrosser | i wonder if that actually uses it for real | 16:03 |
| noonedeadpunk | well https://opendev.org/openstack/openstack-ansible/src/branch/master/setup.cfg is kind of placeholder... | 16:05 |
| noonedeadpunk | oh, wait, it really installs osa_toolkit scripts | 16:05 |
| noonedeadpunk | or we can actually rename test-requirements.txt to smth like linters-requirements.txt | 16:06 |
| jrosser | i think it is used by pip https://zuul.opendev.org/t/openstack/build/adcf0a1b64a74eb090f4578a2b97e4ef/log/job-output.txt#3330-3346 | 16:07 |
| noonedeadpunk | and use test-requirements as we should for CI only | 16:07 |
| jrosser | the osa_toolkits stuff needs to be in the runtime venv, i think thats why it's doing this | 16:08 |
| noonedeadpunk | and pip can say openstack-ansible version, huh | 16:08 |
| jrosser | yup | 16:08 |
| noonedeadpunk | so it's kind of classic requirements.txt... | 16:09 |
| jrosser | turns out so, yes :) | 16:09 |
| jrosser | perhaps i'm just doing it wrong and the journal collection stuff somehow needs to be in the osa_toolkit stuff | 16:10 |
| jrosser | and be properly installed | 16:10 |
| jrosser | noonedeadpunk: btw i think the '2' on those neutron job names was because the same ones exist(ed) in the integrated repo? | 16:13 |
| jrosser | so needs a bit of switch around to remove the duplicates first, then drop the 2 | 16:14 |
| noonedeadpunk | well, on master they're dropped | 16:15 |
| noonedeadpunk | https://codesearch.opendev.org/?q=openstack-ansible-deploy-aio_metal_ovn-ubuntu-focal&i=nope&files=&excludeFiles=&repos= | 16:15 |
| noonedeadpunk | journal collection is kind of ci thing only that will be close to never used in deployments? | 16:16 |
| jrosser | yes thats right | 16:20 |
| jrosser | theres two ways to approach it | 16:21 |
| jrosser | either we need to get all the right python libs installed on the host with distro packages, and run the journal collection with the system python | 16:21 |
| jrosser | or we need the python libs in the ansible-runtime venv | 16:22 |
| spatel | noonedeadpunk i did this in user_variables nova_compute_kvm_distro_packages: qemu-kvm-ev-2.12.0-18 | 16:22 |
| spatel | now getting error {"changed": false, "failed": true, "msg": "No package matching 'q' found available | 16:23 |
| spatel | should i be putting that in quotes ? | 16:23 |
| jrosser | spatel: should that varaible be a list? | 16:24 |
| jrosser | that error suggests its trying to install ['q', 'e', 'm', 'u',.........] | 16:25 |
| noonedeadpunk | agree, that should be a list | 16:26 |
| spatel | This is what happening, i am trying to install new compute node and getting dependency error for qemu-kvm-ev (because its trying to install latest version) so i am trying to use specific version qemu-kvm-ev-2.12.0-18 | 16:26 |
| noonedeadpunk | jrosser: I think libs in ansible-runtime venv are ok, the question is just where to put them to affect only CI deployments and not install for all | 16:26 |
| spatel | hmm | 16:27 |
| jrosser | spatel: you should maybe be setting some kind of yum thing like we do for apt package pins | 16:27 |
| jrosser | so get a specific version | 16:27 |
| spatel | Do you guys put specific version pin on apt package? | 16:28 |
| jrosser | sometimes | 16:28 |
| jrosser | if you mess with the package list and make it install a specific version, if you ever run an update on that host by some other means you might get an unwelcome surpruse | 16:28 |
| spatel | Agreed! | 16:29 |
| spatel | jrosser you are saying this method won't work ? nova_compute_kvm_distro_packages: qemu-kvm-ev-2.12.0-18 ? | 16:30 |
| jrosser | it may perhaps work that one time you run it | 16:30 |
| jrosser | it depends on how yum works, if it considers that to be an absolute version that can't be upgraded | 16:31 |
| spatel | That package comes from external repo rdo-qemu-ev.repo | 16:31 |
| jrosser | or if you were to just globally update the packages on the host at some later point it might see a later version in the repo and update it for you | 16:32 |
| jrosser | that would be a good experiment to do | 16:32 |
| jrosser | you would also need to override the whole variable https://opendev.org/openstack/openstack-ansible-os_nova/src/branch/stable/victoria/vars/redhat.yml#L61-L69 | 16:33 |
| jrosser | ^ list :) | 16:33 |
| jrosser | noonedeadpunk: so we need to leave requirements.txt just as it is, for the toolkit install | 16:34 |
| noonedeadpunk | yep | 16:34 |
| jrosser | then test-requirements is kind of wierd becasue that doesnt use the ansible-runtime venv at all | 16:35 |
| noonedeadpunk | yep, they are... | 16:35 |
| spatel | jrosser this is what its doing - http://paste.openstack.org/show/803573/ | 16:36 |
| noonedeadpunk | but where does linters got installed... | 16:36 |
| spatel | but why??? | 16:36 |
| * jrosser looks at the ancient version of virtualenv OMG https://opendev.org/openstack/openstack-ansible/src/branch/master/test-requirements.txt#L9 | 16:36 | |
| jrosser | spatel: paste what you have overidden the variable to | 16:37 |
| spatel | nova_compute_kvm_distro_packages: "qemu-kvm-ev-2.12.0-18.el7_6.5.1.x86_64". | 16:37 |
| spatel | in user_variables.yml | 16:37 |
| jrosser | jrosser> spatel: should that varaible be a list? | 16:37 |
| jrosser | noonedeadpunk> agree, that should be a list | 16:37 |
| spatel | yes that variabled in vars/redhat-7.yml | 16:37 |
| spatel | its qeens version of oprnestack | 16:38 |
| spatel | queens* | 16:38 |
| jrosser | look at the formatting in the vars file | 16:38 |
| jrosser | it's lines all starting with '-' | 16:38 |
| jrosser | you've specified a single string as a list, so it iterates over the chars in the string, as a list | 16:39 |
| jrosser | and like i said | 16:39 |
| spatel | This is what i found in vars/redhat-7.yml file - http://paste.openstack.org/show/803574/ | 16:39 |
| jrosser | ? yes! its a list :) | 16:40 |
| spatel | hmm it me try with "-" | 16:40 |
| openstackgerrit | Dmitriy Rabotyagov proposed openstack/openstack-ansible stable/ussuri: Remove OVN jobs from the integrated repo https://review.opendev.org/c/openstack/openstack-ansible/+/780616 | 16:40 |
| spatel | jrosser :) you nailed it, can you show me how do you pin packages in apt or how many package you pin? etc | 16:42 |
| openstackgerrit | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_neutron stable/ussuri: Use integrated OVN test https://review.opendev.org/c/openstack/openstack-ansible-os_neutron/+/780617 | 16:43 |
| noonedeadpunk | spatel: ie https://opendev.org/openstack/openstack-ansible-rabbitmq_server/src/branch/master/tasks/install_apt.yml#L15-L30 | 16:44 |
| *** jamesdenton has quit IRC | 16:46 | |
| *** jamesdenton has joined #openstack-ansible | 16:47 | |
| jrosser | spatel: here is an example for ceph http://paste.openstack.org/show/803576/ | 16:47 |
| jrosser | it says "i only want packages from ceph.com, and i want this specific version" | 16:48 |
| spatel | interesting, i didn't know about this feature - /etc/apt/preferences.d/ | 16:48 |
| noonedeadpunk | ok, so tempest tests for centos7 on train are completely broken | 16:49 |
| jrosser | oh dear | 16:49 |
| spatel | Do you think i should do same for KVM/Libvirt etc too? | 16:49 |
| jrosser | i fixed some stuff last week i think in os_tempest | 16:49 |
| noonedeadpunk | http://paste.openstack.org/show/803577/ | 16:49 |
| noonedeadpunk | it's py2.7 in tempest venv O_o | 16:50 |
| noonedeadpunk | haven't we switched centos 7 to py3? or it was in U? | 16:50 |
| jrosser | oh hah "fixed!" https://review.opendev.org/c/openstack/openstack-ansible-os_tempest/+/779696 | 16:50 |
| jrosser | train was still dual-stacked py2/py3 | 16:50 |
| noonedeadpunk | how to make it pass.... | 16:51 |
| noonedeadpunk | `Cloning https://opendev.org/openstack/cinder-tempest-plugin (to revision False)` hm | 16:53 |
| noonedeadpunk | https://662fbc83c91c32a8789e-45518917cf8baf33fe991d0324b9a061.ssl.cf2.rackcdn.com/779696/2/check/openstack-ansible-deploy-aio_metal-debian-buster/4ca234b/logs/ara-report/result/2a3f4aaa-0d8e-451f-85c3-8f1ba7697d2d/ | 16:53 |
| jrosser | hrrm that doesnt look right | 16:53 |
| noonedeadpunk | ok, missed brackets I guess | 16:53 |
| openstackgerrit | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_tempest stable/train: Pin neutron, keystone and cinder tempest plugins to py2 compatible versions where needed https://review.opendev.org/c/openstack/openstack-ansible-os_tempest/+/779696 | 16:54 |
| jrosser | ah round the == condition | 16:54 |
| noonedeadpunk | yeah | 16:54 |
| jrosser | donh /o\ | 16:54 |
| jrosser | there should really be some huge warning when that happens | 16:55 |
| noonedeadpunk | can't agree more | 16:56 |
| noonedeadpunk | regarding ovn integrated jobs - my plan to drop them from integrated repo down to U | 16:56 |
| * noonedeadpunk don't like `2` in job name | 16:57 | |
| noonedeadpunk | and re-add to just neutron | 16:57 |
| seanp | Is there something I should look for specifically when lauching an instance from an ISO and glance seems to write the contents of the ISO to the volume, then boot from it? It doesn't seem to want to mount the ISO on the instance | 16:58 |
| seanp | there are no apparent errors in the syslog on the controller or the compute node | 16:59 |
| noonedeadpunk | in prior to train that didn't work for sure. You had to create an ephemeral drive from ISO | 17:00 |
| noonedeadpunk | I think this hs been fixed in newer cinder | 17:00 |
| jamesdenton | noonedeadpunk the '2' in the name was due to zuul complaining about conflicting job names, even though the patch it depended on had removed those jobs. The os_neutron patch just merged too quickly before i had a chance to correct it | 17:00 |
| seanp | is that something I would need to add to the flavor? like a 10G ephermeral disk or something? | 17:00 |
| noonedeadpunk | jamesdenton: yeah, we need to remove these jobs from all branches where they're referenced | 17:01 |
| jamesdenton | i see | 17:01 |
| noonedeadpunk | seanp: well, kind of. or allow to create vms with zero-disk flavors in policy, then emphemeral size would be equall to image size | 17:03 |
| noonedeadpunk | `os_compute_api:servers:create:zero_disk_flavor` inpolicy iirc | 17:03 |
| noonedeadpunk | https://docs.openstack.org/nova/latest/configuration/policy.html | 17:03 |
| seanp | Thanks, I will read up on that | 17:04 |
| LowKey | Hi, i just installed gnocchi and keep getting gnocchi error reading communication on syslog and also haproxy seem very busy with it, not sure this is normal or not? the load server seem low, but this make the syslog so busy.. http://paste.openstack.org/show/BJ0ckUjgjxabjNTAZ4RC/ | 17:04 |
| *** rpittau is now known as rpittau|afk | 17:04 | |
| spatel | jrosser this trick for me. - yum versionlock add qemu-kvm-ev-2.12.0-18.el7_6.5.1.x86_64 | 17:05 |
| jrosser | cool, that should do it | 17:05 |
| spatel | you are right i should lock them in yum and instead of messing with variabled | 17:06 |
| spatel | variables* | 17:06 |
| openstackgerrit | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_neutron stable/victoria: Add integrated tests various Neutron deployment scenarios https://review.opendev.org/c/openstack/openstack-ansible-os_neutron/+/780457 | 17:08 |
| *** mubix_ has joined #openstack-ansible | 17:09 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_blazar master: Use ansible_facts[] instead of fact variables https://review.opendev.org/c/openstack/openstack-ansible-os_blazar/+/780596 | 17:09 |
| *** jonher_ has joined #openstack-ansible | 17:09 | |
| *** jonher has quit IRC | 17:10 | |
| *** jonher_ is now known as jonher | 17:10 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_barbican master: Use ansible_facts[] instead of fact variables https://review.opendev.org/c/openstack/openstack-ansible-os_barbican/+/780595 | 17:10 |
| openstackgerrit | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_zun master: [reno] Stop publishing release notes https://review.opendev.org/c/openstack/openstack-ansible-os_zun/+/772053 | 17:10 |
| *** dasp_ has joined #openstack-ansible | 17:10 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_aodh master: Use ansible_facts[] instead of fact variables https://review.opendev.org/c/openstack/openstack-ansible-os_aodh/+/780594 | 17:11 |
| *** evrardjp_ has joined #openstack-ansible | 17:11 | |
| *** evrardjp has quit IRC | 17:14 | |
| *** mubix has quit IRC | 17:14 | |
| *** dasp has quit IRC | 17:14 | |
| *** mubix_ is now known as mubix | 17:15 | |
| seanp | I got a slight clarification on this issue. It seems that perhaps OpenStack doesn't know that I want to boot from an ISO, it is treating the image as if I had uploaded a qcow2 image and is assigning that (+ whatever extra available disk for what I selected) as the VM disk | 17:16 |
| seanp | What I'm trying to accomplish is to boot from an ISO, as one would on a "traditional" KVM virtual machine | 17:17 |
| seanp | The person who engineered the production environment where I work said that they had to do some sort of hackery to make it work the way I envision | 17:18 |
| jrosser | seanp: is it something special on the ISO you wouldnt be able to get a normal cloud image for? | 17:42 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_ceilometer master: Use ansible_facts[] instead of fact variables https://review.opendev.org/c/openstack/openstack-ansible-os_ceilometer/+/780635 | 17:48 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_cloudkitty master: Use ansible_facts[] instead of fact variables https://review.opendev.org/c/openstack/openstack-ansible-os_cloudkitty/+/780636 | 17:57 |
| *** jonher has quit IRC | 17:58 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_designate master: Use ansible_facts[] instead of fact variables https://review.opendev.org/c/openstack/openstack-ansible-os_designate/+/780638 | 17:58 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_glance master: Remove remaining suse reference https://review.opendev.org/c/openstack/openstack-ansible-os_glance/+/780639 | 18:00 |
| seanp | jrosser: No, I just (at the time) wasn't thinking like that; about having a pre-installed image and booting to that like the CirrOS test image | 18:00 |
| jrosser | you should be able to find those for all the major OS | 18:01 |
| seanp | I just came across this: https://docs.openstack.org/image-guide/obtain-images.html | 18:01 |
| jrosser | one thing that does work like you describe is when you "rescue" an already existing instance | 18:01 |
| jrosser | that reboots the instance with an extra disk attached that it should boot off instead | 18:01 |
| jrosser | and that can be a plain OS image or system-rescue-cd or something | 18:02 |
| seanp | Ahh, ok cool. I will check that out as well | 18:02 |
| jrosser | you do whatever fixing is needed then 'unrescue' to detach the extra disk and reboot the original | 18:02 |
| jrosser | useful if you lose your ssh keys or something | 18:02 |
| seanp | Neat. I guess I was just confused because the system I use primarily lets you choose and image and boot to it, but also has the pre-installed cloud images available as well | 18:03 |
| *** masterpe has quit IRC | 18:03 | |
| jrosser | You can also boot a cloud image and install some stuff on it, then create a new glance image from a snapshot of that | 18:04 |
| *** masterpe has joined #openstack-ansible | 18:05 | |
| *** jonher has joined #openstack-ansible | 18:06 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_gnocchi master: Use ansible_facts[] instead of fact variables https://review.opendev.org/c/openstack/openstack-ansible-os_gnocchi/+/780640 | 18:06 |
| *** csmart has quit IRC | 18:08 | |
| *** manti has quit IRC | 18:08 | |
| *** fridtjof[m] has quit IRC | 18:08 | |
| *** masterpe has quit IRC | 18:09 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_heat master: Use ansible_facts[] instead of fact variables https://review.opendev.org/c/openstack/openstack-ansible-os_heat/+/780641 | 18:10 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_ironic master: Use ansible_facts[] instead of fact variables https://review.opendev.org/c/openstack/openstack-ansible-os_ironic/+/780642 | 18:13 |
| *** andrewbonney has quit IRC | 18:13 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_magnum master: Use ansible_facts[] instead of fact variables https://review.opendev.org/c/openstack/openstack-ansible-os_magnum/+/780643 | 18:17 |
| *** yasemind has quit IRC | 18:19 | |
| *** jonher has quit IRC | 18:19 | |
| *** jonher has joined #openstack-ansible | 18:20 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_manila master: Use ansible_facts[] instead of fact variables https://review.opendev.org/c/openstack/openstack-ansible-os_manila/+/780646 | 18:22 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_heat master: Use ansible_facts[] instead of fact variables https://review.opendev.org/c/openstack/openstack-ansible-os_heat/+/780641 | 18:24 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_masakari master: Use ansible_facts[] instead of fact variables https://review.opendev.org/c/openstack/openstack-ansible-os_masakari/+/780649 | 18:26 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_mistral master: Use ansible_facts[] instead of fact variables https://review.opendev.org/c/openstack/openstack-ansible-os_mistral/+/780651 | 18:27 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_murano master: Use ansible_facts[] instead of fact variables https://review.opendev.org/c/openstack/openstack-ansible-os_murano/+/780653 | 18:30 |
| *** jamesdenton has quit IRC | 18:31 | |
| *** jamesdenton has joined #openstack-ansible | 18:32 | |
| *** manti has joined #openstack-ansible | 18:37 | |
| *** spatel has quit IRC | 18:43 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_neutron master: Use ansible_facts[] instead of fact variables https://review.opendev.org/c/openstack/openstack-ansible-os_neutron/+/777650 | 18:48 |
| *** spatel_ has joined #openstack-ansible | 18:56 | |
| *** partlycl- has joined #openstack-ansible | 19:09 | |
| *** owalsh has joined #openstack-ansible | 19:10 | |
| *** partlycloudy has quit IRC | 19:11 | |
| *** owalsh_ has quit IRC | 19:11 | |
| *** gary_perkins_ has quit IRC | 19:11 | |
| *** gary_perkins has joined #openstack-ansible | 19:11 | |
| *** tbarron has quit IRC | 19:11 | |
| *** partlycl- has quit IRC | 19:13 | |
| *** csmart has joined #openstack-ansible | 19:14 | |
| *** fridtjof[m] has joined #openstack-ansible | 19:14 | |
| *** masterpe has joined #openstack-ansible | 19:14 | |
| *** partlycloudy has joined #openstack-ansible | 19:16 | |
| *** lemko9 has joined #openstack-ansible | 19:24 | |
| *** lemko has quit IRC | 19:25 | |
| *** lemko9 is now known as lemko | 19:25 | |
| *** djhankb0 has joined #openstack-ansible | 19:30 | |
| *** crazzy has quit IRC | 19:31 | |
| *** spatel_ has quit IRC | 19:32 | |
| *** djhankb has quit IRC | 19:32 | |
| *** djhankb0 is now known as djhankb | 19:32 | |
| *** seanp has quit IRC | 20:03 | |
| *** spatel_ has joined #openstack-ansible | 20:07 | |
| spatel_ | noonedeadpunk do you running zabbix right? | 20:07 |
| spatel_ | my zabbix not able to talk agent on port 10050 | 20:08 |
| spatel_ | do i need to add 10050 on iptables rules on compute nodes | 20:09 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_tempest stable/train: Pin neutron, keystone and cinder tempest plugins to py2 compatible versions where needed https://review.opendev.org/c/openstack/openstack-ansible-os_tempest/+/779696 | 20:56 |
| *** macz_ has quit IRC | 20:56 | |
| *** odyssey4me has quit IRC | 20:58 | |
| *** macz_ has joined #openstack-ansible | 20:58 | |
| *** odyssey4me has joined #openstack-ansible | 20:58 | |
| spatel_ | admin0 ? | 21:05 |
| spatel_ | around | 21:05 |
| spatel_ | what is "migrate instance" vs "live migrate instance" in openstack horizon? | 21:07 |
| admin0 | migrate = pause and move | 21:12 |
| admin0 | live = make a copy, sync ram, switch when ram is synced | 21:12 |
| *** spatel_ has quit IRC | 21:21 | |
| admin0 | if your instance is very very busy, live will fail coz the ram is never in a state where it can be properly synced | 21:46 |
| *** seanp has joined #openstack-ansible | 22:06 | |
| *** seanp has quit IRC | 22:06 | |
| *** seanp has joined #openstack-ansible | 22:07 | |
| *** macz_ has quit IRC | 23:33 | |
| *** MrClayPole has quit IRC | 23:40 | |
| *** macz_ has joined #openstack-ansible | 23:44 | |
| *** MrClayPole has joined #openstack-ansible | 23:46 | |
| *** macz_ has quit IRC | 23:48 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!