| opendevreview | Merged openstack/openstack-ansible master: Define corosync interface for facts gathering https://review.opendev.org/c/openstack/openstack-ansible/+/845894 | 01:23 |
|---|---|---|
| *** chkumar|rover is now known as chandankumar | 05:10 | |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_nova master: Remove AvailabilityZoneFilter from scheduler https://review.opendev.org/c/openstack/openstack-ansible-os_nova/+/846271 | 06:03 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_neutron master: Deprecate allow_overlapping_ips option https://review.opendev.org/c/openstack/openstack-ansible-os_neutron/+/846272 | 06:05 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible master: Install murano client for distro installs https://review.opendev.org/c/openstack/openstack-ansible/+/846162 | 06:20 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_ironic master: Remove [keystone] configuration block https://review.opendev.org/c/openstack/openstack-ansible-os_ironic/+/831544 | 06:25 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_ironic master: Support service tokens https://review.opendev.org/c/openstack/openstack-ansible-os_ironic/+/846033 | 06:26 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_neutron master: Allow to provide custom configuration for VPNaaS https://review.opendev.org/c/openstack/openstack-ansible-os_neutron/+/844041 | 06:49 |
| *** arxcruz is now known as arxcruz|rover | 07:08 | |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible stable/pike: EOL Pike branch https://review.opendev.org/c/openstack/openstack-ansible/+/846279 | 07:37 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible stable/pike: EOL Pike branch https://review.opendev.org/c/openstack/openstack-ansible/+/826956 | 07:39 |
| noonedeadpunk | that is also interesting thread for malloc actuall https://bugs.python.org/issue11849 | 07:44 |
| noonedeadpunk | and tbh I'm not really sure if /etc/environment anyhow affect systemd services... | 07:46 |
| noonedeadpunk | it should though | 07:46 |
| noonedeadpunk | It's also possible that we don't see any improvement with releasing memory as in aio we have set only sigle worker for everything | 07:58 |
| noonedeadpunk | or jsut systemd doesn't care about /etc/environment which is most likely thing that happens | 08:00 |
| *** ysandeep|out is now known as ysandeep|afk | 08:10 | |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible master: Reduce memory consumption in CI https://review.opendev.org/c/openstack/openstack-ansible/+/846228 | 08:14 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible master: Reduce memory consumption in CI https://review.opendev.org/c/openstack/openstack-ansible/+/846228 | 08:16 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_magnum master: Control amount of conductor workers https://review.opendev.org/c/openstack/openstack-ansible-os_magnum/+/846153 | 08:24 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_magnum master: Support service tokens https://review.opendev.org/c/openstack/openstack-ansible-os_magnum/+/846037 | 08:29 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_magnum master: Fixed dest typo in config_template https://review.opendev.org/c/openstack/openstack-ansible-os_magnum/+/845993 | 08:29 |
| opendevreview | Merged openstack/openstack-ansible-os_swift master: Support service tokens https://review.opendev.org/c/openstack/openstack-ansible-os_swift/+/846051 | 09:59 |
| opendevreview | Merged openstack/openstack-ansible-os_sahara master: Support service tokens https://review.opendev.org/c/openstack/openstack-ansible-os_sahara/+/846044 | 10:04 |
| opendevreview | Merged openstack/openstack-ansible-os_cloudkitty master: Support service tokens https://review.opendev.org/c/openstack/openstack-ansible-os_cloudkitty/+/846019 | 10:06 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible master: Stop NetworkManager on centos 9. https://review.opendev.org/c/openstack/openstack-ansible/+/846224 | 10:13 |
| opendevreview | Merged openstack/openstack-ansible-os_mistral master: Support service tokens https://review.opendev.org/c/openstack/openstack-ansible-os_mistral/+/846041 | 10:21 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible stable/xena: Define corosync interface for facts gathering https://review.opendev.org/c/openstack/openstack-ansible/+/846124 | 10:25 |
| opendevreview | Merged openstack/openstack-ansible-os_zun master: Remove mention of haproxy-endpoints role https://review.opendev.org/c/openstack/openstack-ansible-os_zun/+/845803 | 10:39 |
| opendevreview | Merged openstack/openstack-ansible-os_zun master: Support service tokens https://review.opendev.org/c/openstack/openstack-ansible-os_zun/+/846046 | 10:44 |
| opendevreview | Merged openstack/openstack-ansible master: Add mistral-extra to required projects in CI https://review.opendev.org/c/openstack/openstack-ansible/+/846135 | 10:53 |
| opendevreview | Merged openstack/openstack-ansible-os_nova master: Remove AvailabilityZoneFilter from scheduler https://review.opendev.org/c/openstack/openstack-ansible-os_nova/+/846271 | 10:59 |
| opendevreview | Merged openstack/openstack-ansible master: Limit magnum workers in CI https://review.opendev.org/c/openstack/openstack-ansible/+/846151 | 11:00 |
| opendevreview | Merged openstack/openstack-ansible master: Explicitly use collection for haproxy module https://review.opendev.org/c/openstack/openstack-ansible/+/845740 | 11:08 |
| opendevreview | Merged openstack/openstack-ansible master: Reduce m1.magnum flavor https://review.opendev.org/c/openstack/openstack-ansible/+/846142 | 11:08 |
| opendevreview | Merged openstack/openstack-ansible master: Fix deprecations handling in journal logs https://review.opendev.org/c/openstack/openstack-ansible/+/846189 | 11:08 |
| *** ysandeep|afk is now known as ysandeep | 11:09 | |
| noonedeadpunk | I guess I fail to understand how to configure these glibc tunables... | 11:32 |
| noonedeadpunk | ah, triplo said they see improvement _only_ for cinder... | 11:36 |
| *** ysandeep is now known as ysandeep|afk | 11:47 | |
| mgariepy | https://bugzilla.redhat.com/show_bug.cgi?id=2092856 | 11:52 |
| mgariepy | i've seen it yesterday but i don't remember on which patch .. :/ | 11:52 |
| mgariepy | https://storage.bhs.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/zuul_opendev_logs_098/846162/2/gate/openstack-ansible-deploy-aio_ceph-ubuntu-focal/098c037/logs/ara-report/results/3841.html | 11:56 |
| mgariepy | same error for both focal aio_ceph and aio_lxc job on : 846162 | 11:57 |
| noonedeadpunk | that was libvirt bug specificaly, isn't it? | 11:57 |
| noonedeadpunk | and, if you're here, can you review https://review.opendev.org/c/openstack/openstack-ansible-os_magnum/+/846153/2 ?:) | 11:59 |
| mgariepy | https://zuul.opendev.org/t/openstack/build/3454d5bfc3234e77aa68c329a34132e6/log/logs/host/nova-compute.service.journal-22-32-45.log.txt#3992 | 12:02 |
| mgariepy | there | 12:02 |
| mgariepy | for octavia | 12:02 |
| noonedeadpunk | yeah... that's exactly the bug ... | 12:02 |
| mgariepy | https://review.opendev.org/c/openstack/tripleo-quickstart/+/846194 | 12:03 |
| mgariepy | not sure why octavia fails but not nova ones.. | 12:03 |
| mgariepy | ho. the nova doesn't attach a net dev separately ? | 12:03 |
| noonedeadpunk | yup | 12:03 |
| mgariepy | good. | 12:04 |
| noonedeadpunk | or at least not with the test we have | 12:04 |
| noonedeadpunk | I'd say octavia we can backport anytime later then | 12:04 |
| mgariepy | are we ready for the release ? | 12:05 |
| noonedeadpunk | that's good question | 12:07 |
| noonedeadpunk | I'd say things are quite in order now? Wdyt? | 12:08 |
| noonedeadpunk | And we totally can't land much more as deadline is less then in a week | 12:08 |
| mgariepy | gnocchi-metricd seems to have quite a few process on ci / c8s. | 12:11 |
| noonedeadpunk | yeah we limit only wsgi there | 12:12 |
| noonedeadpunk | so might be worth implementing variable to control it as well | 12:12 |
| mgariepy | can we override [metricd]/workers ? | 12:13 |
| noonedeadpunk | yup | 12:13 |
| opendevreview | Merged openstack/openstack-ansible master: Disable service_token requirement by default https://review.opendev.org/c/openstack/openstack-ansible/+/823005 | 12:14 |
| opendevreview | Merged openstack/openstack-ansible stable/pike: EOL Pike branch https://review.opendev.org/c/openstack/openstack-ansible/+/826956 | 12:14 |
| noonedeadpunk | but worth doing like https://review.opendev.org/c/openstack/openstack-ansible-os_magnum/+/846153/2 I guess? | 12:14 |
| mgariepy | do you or should i ? | 12:15 |
| noonedeadpunk | go on :) | 12:15 |
| mgariepy | haha :) ok | 12:16 |
| noonedeadpunk | I'm trying to figure out this memory reduce thing atm.. | 12:16 |
| opendevreview | Marc Gariépy proposed openstack/openstack-ansible-os_gnocchi master: Control amount of metricd workers https://review.opendev.org/c/openstack/openstack-ansible-os_gnocchi/+/846347 | 12:27 |
| noonedeadpunk | jsut in case that's memory usage in sandbox after I enabled glibc malloc tuning https://pasteboard.co/w7L1BEVnyNW2.png | 12:28 |
| noonedeadpunk | playbooks started at 13:00 | 12:28 |
| noonedeadpunk | and it's just one of control planes | 12:28 |
| noonedeadpunk | from other side, rally performs better... | 12:29 |
| mgariepy | what are the steps in the memory drop ? | 12:29 |
| mgariepy | avail. it's not less ram it's more lol | 12:29 |
| mgariepy | never mind :) haha | 12:29 |
| noonedeadpunk | https://paste.openstack.org/show/buJsGoy1AVCuaxy06udh/ | 12:30 |
| mgariepy | wow | 12:30 |
| noonedeadpunk | yeah, it's less. | 12:30 |
| noonedeadpunk | Or well, consumption is higher | 12:30 |
| mgariepy | less free ram | 12:30 |
| noonedeadpunk | yeah | 12:30 |
| noonedeadpunk | but execution seems faster. unless image jsut got cached on all computes in sandbox | 12:31 |
| noonedeadpunk | and also I realized it has most effect on cinder volume after run rally, so it's not bfv test, but just ephemeral | 12:31 |
| opendevreview | Marc Gariépy proposed openstack/openstack-ansible master: Set the max number of gnicchi-metricd workers https://review.opendev.org/c/openstack/openstack-ansible/+/846349 | 12:35 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_octavia master: Set CentOS 9 Stream as NV https://review.opendev.org/c/openstack/openstack-ansible-os_octavia/+/846350 | 12:35 |
| noonedeadpunk | -that's my proposal how to workaround bug you found mgariepy :D ^ | 12:36 |
| mgariepy | ahha :) yep good enough | 12:38 |
| mgariepy | gnocchi metricd workers stuff will be needed for other gnocchi patch to pass i guess | 12:39 |
| mgariepy | rockylinux 8 issue was also memory issue. | 12:39 |
| noonedeadpunk | for gnocchi gnocchi_metricd_workers should be implemented first in role? | 12:41 |
| mgariepy | https://review.opendev.org/c/openstack/openstack-ansible-os_gnocchi/+/846347 | 12:42 |
| mgariepy | might need it to depends on the other patch tho. | 12:43 |
| mgariepy | let me fix that. | 12:43 |
| opendevreview | Marc Gariépy proposed openstack/openstack-ansible-os_gnocchi master: Control amount of metricd workers https://review.opendev.org/c/openstack/openstack-ansible-os_gnocchi/+/846347 | 12:43 |
| noonedeadpunk | how I missed that | 12:43 |
| mgariepy | here you go. | 12:43 |
| mgariepy | that might fix the test for gnocchi | 12:43 |
| noonedeadpunk | nice! | 12:43 |
| noonedeadpunk | I have feeling it will | 12:44 |
| mgariepy | just copied you magnum patch with mostly sed :p | 12:44 |
| mgariepy | maybe we could also set the number of thread to 1 for all the wsgi process ? | 12:51 |
| mgariepy | https://github.com/openstack/openstack-ansible/blob/master/tests/roles/bootstrap-host/templates/user_variables.aio.yml.j2#L96 and so on. | 12:52 |
| *** dviroel|out is now known as dviroel | 13:00 | |
| noonedeadpunk | I'd say we should jsut drop `nova_wsgi_processes_max` as it has no effect when `nova_wsgi_processes` is set | 13:04 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible master: [doc] Update link to Ansible doc in manifesto https://review.opendev.org/c/openstack/openstack-ansible/+/846358 | 13:08 |
| mgariepy | the max should be removed, but also maybe we can set to 1 every process some are set to 2 other to 1.. | 13:13 |
| *** ysandeep|afk is now known as ysandeep | 13:16 | |
| noonedeadpunk | ah, yes | 13:37 |
| noonedeadpunk | that's might be worth doing indeed | 13:37 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_neutron master: Allow to provide custom configuration for VPNaaS https://review.opendev.org/c/openstack/openstack-ansible-os_neutron/+/844041 | 13:38 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_octavia master: Support service tokens https://review.opendev.org/c/openstack/openstack-ansible-os_octavia/+/845913 | 13:40 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_aodh master: Support service tokens https://review.opendev.org/c/openstack/openstack-ansible-os_aodh/+/846010 | 13:40 |
| noonedeadpunk | so overall we have telemetry/murano/masakari left I believe? | 13:42 |
| noonedeadpunk | and ironic | 13:42 |
| mgariepy | i think so. | 13:47 |
| mgariepy | murano might need some push :/ not running in CI. | 13:47 |
| noonedeadpunk | I think it depends on some integrated repo patch? | 13:57 |
| noonedeadpunk | https://review.opendev.org/c/openstack/openstack-ansible/+/846162 | 13:57 |
| mgariepy | yes but why does it block ? because it failed ? | 13:58 |
| mgariepy | anyway it's all green for now. | 13:59 |
| noonedeadpunk | so murano depends on this one. and as it's failed several times murano blocked I believe | 14:00 |
| *** dviroel is now known as dviroel|pto | 14:03 | |
| opendevreview | Merged openstack/openstack-ansible-os_magnum master: Control amount of conductor workers https://review.opendev.org/c/openstack/openstack-ansible-os_magnum/+/846153 | 14:03 |
| opendevreview | Merged openstack/openstack-ansible-os_magnum master: Support service tokens https://review.opendev.org/c/openstack/openstack-ansible-os_magnum/+/846037 | 14:03 |
| mgariepy | service token for gnocchi failed on mem alloc rocky8 again .. | 14:04 |
| *** ysandeep is now known as ysandeep|out | 14:04 | |
| noonedeadpunk | worth rebasing then | 14:10 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_gnocchi master: Support service tokens https://review.opendev.org/c/openstack/openstack-ansible-os_gnocchi/+/846030 | 14:10 |
| opendevreview | Marc Gariépy proposed openstack/openstack-ansible master: Limit the number of wsgi process to 1. https://review.opendev.org/c/openstack/openstack-ansible/+/846385 | 14:12 |
| mgariepy | let's wait on the CI :d | 14:15 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/ansible-config_template master: Publish openstack.config_template on galaxy https://review.opendev.org/c/openstack/ansible-config_template/+/846391 | 14:27 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/ansible-config_template master: Publish openstack.config_template on galaxy https://review.opendev.org/c/openstack/ansible-config_template/+/846391 | 14:32 |
| spatel | random question, how do you guys provide storage service to your public cloud customer? | 14:44 |
| spatel | shared storage i meant? | 14:44 |
| spatel | do you have special network for storage you attach to VM or use tenant network to talk to storage? | 14:44 |
| mgariepy | i guess the answer is : it depends | 14:45 |
| noonedeadpunk | we not atm as it's pita as you rightfully mentioned. As using tenant network to access storage one is super bad idea for me. | 14:45 |
| noonedeadpunk | As then you need ganesha and ensure it's ha deployment which is also meh... | 14:46 |
| noonedeadpunk | or generic driver, which also meh | 14:46 |
| noonedeadpunk | (if we're talking about manila) | 14:46 |
| noonedeadpunk | and under shared storage you meant shared filesystem :) | 14:47 |
| mgariepy | https://blueprints.launchpad.net/nova/+spec/libvirt-virtiofs-attach-manila-shares | 14:47 |
| spatel | i have same problem currently | 14:47 |
| mgariepy | if this gets done it would solve a lot of issues. | 14:48 |
| spatel | I am deploying cephfs with manila to provide shared filesystem to cluster for HPC job but not sure how do i expose filesystem to all VM. | 14:48 |
| mgariepy | do you have a lot of tenants? | 14:48 |
| spatel | Problem is customer doesn't want to spend money on network switches to have dedicated nic attach to vm | 14:48 |
| spatel | I have single tenants at present but i don't think we will have many. | 14:49 |
| spatel | ofc if i route via tenants router then performance will be shit | 14:49 |
| noonedeadpunk | mgariepy: yeah, we wait for this to be implemented :D | 14:49 |
| noonedeadpunk | but afaik nova part hasn't started yet | 14:50 |
| vkmc | noonedeadpunk, hey :) the spec I mentioned is under review for virtiofs in manila is here https://review.opendev.org/c/openstack/nova-specs/+/833669, that's latest about virtiofs on manila | 14:50 |
| spatel | I want virtiofs :) | 14:50 |
| vkmc | and there are several related changes to that | 14:52 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/ansible-config_template master: Add blank lines between functions and classes https://review.opendev.org/c/openstack/ansible-config_template/+/846418 | 14:53 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/ansible-config_template master: Add blank lines between functions and classes https://review.opendev.org/c/openstack/ansible-config_template/+/846418 | 14:54 |
| noonedeadpunk | vkmc: yeah, thanks! I will try to have a look where and how we can help with pushing that forward | 14:56 |
| vkmc | noonedeadpunk++ thanks | 14:58 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_neutron master: Deprecate allow_overlapping_ips option https://review.opendev.org/c/openstack/openstack-ansible-os_neutron/+/846272 | 14:59 |
| noonedeadpunk | vkmc: oh, wow, there're a lot of things already done before blueprint got merged | 15:08 |
| opendevreview | Merged openstack/openstack-ansible stable/xena: Define corosync interface for facts gathering https://review.opendev.org/c/openstack/openstack-ansible/+/846124 | 15:21 |
| jrosser_ | is this good to merge now? https://review.opendev.org/c/openstack/openstack-ansible/+/846145 | 16:02 |
| jrosser_ | want to avoid accidental zuul queue problem | 16:02 |
| noonedeadpunk | jrosser_: yes, should be good | 17:55 |
| opendevreview | Merged openstack/openstack-ansible master: Install murano client for distro installs https://review.opendev.org/c/openstack/openstack-ansible/+/846162 | 17:57 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible master: Stop NetworkManager on centos 9. https://review.opendev.org/c/openstack/openstack-ansible/+/846224 | 17:59 |
| noonedeadpunk | ^ was failing on ceph job | 17:59 |
| noonedeadpunk | so just rebased to save time | 17:59 |
| mgariepy | wath was the ceph job issue ? | 18:00 |
| mgariepy | cinder volume type stuff :/ | 18:00 |
| noonedeadpunk | well most time I was looking these looked like oom? | 18:04 |
| noonedeadpunk | but I haven't checked this specific one though... | 18:04 |
| mgariepy | ok | 18:05 |
| mgariepy | less oom ?https://review.opendev.org/c/openstack/openstack-ansible/+/846385 | 18:05 |
| mgariepy | if we are a couple of Mb short ;) haha | 18:06 |
| noonedeadpunk | there's also actually keystone_httpd_mpm stuff to check. And I guess that actually cinder type can be related to this. As catched several times that we're short in mpm workers | 18:07 |
| noonedeadpunk | or smth is off with mpm | 18:07 |
| mgariepy | https://zuul.opendev.org/t/openstack/build/5d6ab88c4d504b2fb2ab24dbfe8ca99f | 18:08 |
| noonedeadpunk | sounds like oom https://zuul.opendev.org/t/openstack/build/5d6ab88c4d504b2fb2ab24dbfe8ca99f/log/logs/host/haproxy.service.journal-17-14-22.log.txt#2332-2335 | 18:14 |
| noonedeadpunk | hm, but nothing in syslog.... where would you see oom in ubuntu? :D | 18:15 |
| noonedeadpunk | it's /var/log/messages on centosos, that I know. On ubuntu I'd expect to see in syslog, but it's not there... | 18:16 |
| mgariepy | dmesg | 18:17 |
| noonedeadpunk | or dmesg | 18:17 |
| noonedeadpunk | but dmesg is cut and short | 18:17 |
| noonedeadpunk | sounds like somewhere in journal, bu we ge only specific services from there.... | 18:18 |
| mgariepy | cloud we have timestamps in haproxy logs ? | 18:18 |
| noonedeadpunk | well. yes?:) | 18:20 |
| mgariepy | ahha | 18:20 |
| noonedeadpunk | but there're timestamps? | 18:20 |
| noonedeadpunk | or well. not for everything | 18:21 |
| noonedeadpunk | it's eventually all about this line of code https://opendev.org/openstack/openstack-ansible/src/branch/master/scripts/journal_dump.py#L121 | 18:21 |
| jrosser_ | we might need /var/log/kern.log for that? | 18:21 |
| mgariepy | we have it. | 18:22 |
| jrosser_ | the other thing to do is look at the journal of a system that has had an OOM | 18:23 |
| noonedeadpunk | yeah, but for journal you remeber that we select only specific stuff from it? | 18:24 |
| jrosser_ | there are many fields in the journal entries, and we bail out early on logging most junk here https://opendev.org/openstack/openstack-ansible/src/branch/master/scripts/journal_dump.py#L98-L105 | 18:24 |
| noonedeadpunk | or we save full journal somewhere else? | 18:24 |
| jrosser_ | so if there was something we could match on for OOM then we could log that specifically | 18:24 |
| mgariepy | oom depends how it goes. for lack of memory on vm start it does only show in nova-compute logs as it's not killing out of oom | 18:24 |
| jrosser_ | at the cost of some extra log processing we could detect some of these things and write out a message | 18:26 |
| jrosser_ | i think i saw that 'Operation not permitted: b'/tmp/.ansible_tmp_t48psnvtest_list_only.yml' -> b'/tmp/test_list_only.yml' on another config_template patch too | 18:33 |
| noonedeadpunk | yep, saw that | 18:38 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible master: Add timestamp to written journal file https://review.opendev.org/c/openstack/openstack-ansible/+/846439 | 18:42 |
| noonedeadpunk | not sure how good idea is that though ^ | 18:42 |
| noonedeadpunk | seems we also has horizon role broken btw... | 18:55 |
| mgariepy | really ? :/ | 18:56 |
| noonedeadpunk | https://review.opendev.org/c/openstack/openstack-ansible-os_horizon/+/844815 | 18:56 |
| noonedeadpunk | fails on tempest with urllib.error.HTTPError: HTTP Error 500: Internal Server Error | 18:57 |
| noonedeadpunk | wf is that https://zuul.opendev.org/t/openstack/build/94a1a9a2f9394d1fbaea0674e20d8dcd/log/logs/host/httpd.service.journal-07-45-36.log.txt#2079 | 18:58 |
| noonedeadpunk | I think it needs an AIO... | 18:59 |
| mgariepy | https://zuul.opendev.org/t/openstack/build/94a1a9a2f9394d1fbaea0674e20d8dcd/log/logs/etc/host/httpd/conf/magic.txt#386 | 19:04 |
| noonedeadpunk | um | 19:05 |
| noonedeadpunk | that looks like it shouldn't be there | 19:06 |
| noonedeadpunk | I like comment on L376 there though | 19:07 |
| mgariepy | yep | 19:08 |
| noonedeadpunk | anyway | 19:12 |
| * noonedeadpunk spawning horizon c9s aio | 19:12 | |
| mgariepy | i think c9s sends broken mod_mime_magic config | 19:14 |
| noonedeadpunk | yeah, I guess it does... | 19:20 |
| noonedeadpunk | we also have issue in parallel clone script https://zuul.opendev.org/t/openstack/build/34004202b5224784b4b276c943a0fe7c/log/job-output.txt#4661 | 19:20 |
| mgariepy | master/main issue | 19:21 |
| noonedeadpunk | yeah, but seems it clones not respecting branch provided | 19:21 |
| mgariepy | it's recovered by the role | 19:21 |
| noonedeadpunk | well, yes | 19:22 |
| jrosser_ | https://github.com/ceph/ceph-ansible/commit/bcedff95bd0f7b3897a56f5ced78865072597cd6 | 19:22 |
| mgariepy | The project is still maintained for the time being but it is encouraged to migrate to ``cephadm``. | 19:24 |
| noonedeadpunk | oh, they've dropped deprecation notice | 19:24 |
| mgariepy | nice :D | 19:24 |
| jrosser_ | https://opendev.org/openstack/openstack-ansible-plugins/src/branch/master/plugins/modules/git_requirements#L217 | 19:24 |
| jrosser_ | i wonder why it's done like that | 19:25 |
| mgariepy | git blame does point to noonedeadpunk :P | 19:26 |
| noonedeadpunk | also I think we should revert https://review.opendev.org/c/openstack/openstack-ansible/+/843704 ? | 19:26 |
| noonedeadpunk | mgariepy: yeah, I have some guesses why :) | 19:26 |
| noonedeadpunk | jrosser_: I think because we were assuming that trackbranch is not required thing to have | 19:31 |
| noonedeadpunk | but likely we don't need to define branch there at all | 19:31 |
| jrosser_ | i wonder if this is really correct https://opendev.org/openstack/openstack-ansible-plugins/src/branch/master/plugins/modules/git_requirements#L175 | 19:33 |
| jrosser_ | like what actually is special about master in this context | 19:33 |
| jrosser_ | rather than just "any branch name, rather than a hash" | 19:33 |
| noonedeadpunk | fair | 19:33 |
| jrosser_ | version hash is a specific non movable point | 19:33 |
| jrosser_ | and branch_name is a thing that might need pull as it could easily be moving forward | 19:34 |
| noonedeadpunk | dropping https://opendev.org/openstack/openstack-ansible-plugins/src/branch/master/plugins/modules/git_requirements#L217 works for this specific issue fwiw | 19:37 |
| mgariepy | i guess it will use the default branch ? | 19:37 |
| noonedeadpunk | yeah | 19:38 |
| noonedeadpunk | I think it let git to decide | 19:38 |
| noonedeadpunk | jrosser_: will you push this then? | 19:39 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible master: Revert "Add release note about Ceph-Ansible deprecation" https://review.opendev.org/c/openstack/openstack-ansible/+/846259 | 19:40 |
| jrosser_ | yep i can do that | 19:42 |
| opendevreview | Jonathan Rosser proposed openstack/openstack-ansible-plugins master: Let git choose the branch when cloning a repo https://review.opendev.org/c/openstack/openstack-ansible-plugins/+/846440 | 19:45 |
| jrosser_ | i'll just fix the commit message a bit | 19:47 |
| opendevreview | Jonathan Rosser proposed openstack/openstack-ansible-plugins master: Let git choose the branch when cloning a repo https://review.opendev.org/c/openstack/openstack-ansible-plugins/+/846440 | 19:49 |
| jrosser_ | changing the other use of "master" is a little more complicated | 19:52 |
| jrosser_ | it has to treat branches (assumed master), hash and tags differently | 19:53 |
| opendevreview | Merged openstack/openstack-ansible-os_octavia master: Set CentOS 9 Stream as NV https://review.opendev.org/c/openstack/openstack-ansible-os_octavia/+/846350 | 19:59 |
| mgariepy | have a nice weekend guys. | 20:02 |
| noonedeadpunk | jrosser_: I can recall there was some weirdenss complexity around these master things... | 20:13 |
| noonedeadpunk | but this indeed seems to safe to drop | 20:13 |
| noonedeadpunk | I hope to merge outstanding stuff during weekends... | 20:16 |
| * damiandabrowski[m] is going to spend a moment on reviews | 20:17 | |
| jrosser_ | repo.tags[] is a thing so we can maybe simplify the logic here | 20:17 |
| jrosser_ | it should be possible to be certain if required_version is a tag rather than that be the 'else' case | 20:18 |
| noonedeadpunk | damiandabrowski[m]: some don't have +W on purpose as dependant are not merged | 20:24 |
| noonedeadpunk | I will add +w on these once all dependencies will be merged | 20:24 |
| damiandabrowski[m] | yeah, but i was wondering what happened here: https://review.opendev.org/c/openstack/openstack-ansible-os_murano/+/846163 | 20:24 |
| jrosser_ | the patch below it is not merged yet | 20:25 |
| noonedeadpunk | it was witing for https://review.opendev.org/c/openstack/openstack-ansible/+/846162 at first | 20:25 |
| noonedeadpunk | and now I guess for https://review.opendev.org/c/openstack/openstack-ansible-os_murano/+/846147/2 to pass check pipeline | 20:25 |
| damiandabrowski[m] | dependent change is merged, this change had workflow+1 but i couldn't see zuul job | 20:25 |
| noonedeadpunk | It's rebased on top of another | 20:26 |
| damiandabrowski[m] | ahh i haven't seen relation chain.. | 20:26 |
| damiandabrowski[m] | sorry | 20:26 |
| jrosser_ | we have a loooong running log upload right at the top of the gate queue :/ | 20:26 |
| noonedeadpunk | doh... | 20:27 |
| noonedeadpunk | I'm thinking what we can reduce in terms of logs collection for jobs that passed... | 20:29 |
| noonedeadpunk | as we still need ara/journald/etc | 20:29 |
| jrosser_ | ARA is the largest number of files by a looong way | 20:30 |
| noonedeadpunk | and what we can drop is really minor things that won't help likely | 20:30 |
| noonedeadpunk | well... maybe indeed drop ara for passed then? | 20:31 |
| jrosser_ | i was looking at random web frameworks to try to wrap up the whole static ara site in a bundle | 20:31 |
| noonedeadpunk | it has some insigths in terms of what exactly was ran | 20:31 |
| jrosser_ | so it would be just a small number of files | 20:32 |
| jrosser_ | but man these tools are all obtuse and wierd | 20:32 |
| opendevreview | Merged openstack/openstack-ansible master: Set the max number of gnicchi-metricd workers https://review.opendev.org/c/openstack/openstack-ansible/+/846349 | 20:32 |
| opendevreview | Merged openstack/openstack-ansible stable/xena: Install rally only when rally_install is set https://review.opendev.org/c/openstack/openstack-ansible/+/845684 | 20:32 |
| jrosser_ | oh phew the gate logs uploaded before timeout | 20:33 |
| noonedeadpunk | timeout is 30mins... | 20:34 |
| damiandabrowski[m] | \o/ | 20:34 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-lxc_hosts master: Do not install COPR repo for CentOS LXC https://review.opendev.org/c/openstack/openstack-ansible-lxc_hosts/+/843672 | 20:40 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_nova master: Change default mode while creating directories https://review.opendev.org/c/openstack/openstack-ansible-os_nova/+/824774 | 20:44 |
| opendevreview | Merged openstack/openstack-ansible master: Update magnum tags and images for CI https://review.opendev.org/c/openstack/openstack-ansible/+/846145 | 20:51 |
| opendevreview | Merged openstack/openstack-ansible master: Limit the number of wsgi process to 1. https://review.opendev.org/c/openstack/openstack-ansible/+/846385 | 20:51 |
| opendevreview | Merged openstack/openstack-ansible master: Revert "Add release note about Ceph-Ansible deprecation" https://review.opendev.org/c/openstack/openstack-ansible/+/846259 | 20:51 |
| jrosser_ | noonedeadpunk: are you happy with the vpnaas config patch | 20:51 |
| jrosser_ | i don't really have much insight on that | 20:51 |
| noonedeadpunk | well, I haven't tested it but it looks reasonable | 20:52 |
| noonedeadpunk | and I believe we should spend time in testing/implementing same on production | 20:53 |
| noonedeadpunk | as now it's indeed a PITA to understand what's going on with connection | 20:54 |
| opendevreview | Jonathan Rosser proposed openstack/openstack-ansible master: [doc] Update link to Ansible doc in manifesto https://review.opendev.org/c/openstack/openstack-ansible/+/846358 | 20:54 |
| noonedeadpunk | ofc maybe I'd leverage config_template instead... | 20:55 |
| noonedeadpunk | and put content and not file... | 20:55 |
| noonedeadpunk | but well | 20:55 |
| noonedeadpunk | this can be adjusted | 20:55 |
| jrosser_ | i think this is why i've been hesitating to approve the patch | 20:55 |
| jrosser_ | it didnt feel very "osa like" | 20:55 |
| noonedeadpunk | I agree here. But we have smth common for ceilometer | 20:56 |
| noonedeadpunk | which grow up in kind of mess to be fair | 20:56 |
| damiandabrowski[m] | i have to go, have a great weekend guys! | 20:56 |
| jrosser_ | o/ | 20:56 |
| noonedeadpunk | thanks o/ | 20:56 |
| noonedeadpunk | https://opendev.org/openstack/openstack-ansible-os_ceilometer/src/branch/master/defaults/main.yml#L297-L310 | 20:57 |
| noonedeadpunk | you too! | 20:57 |
| noonedeadpunk | so probably because of that I kind of - meh, if proposed this way it might be fine... | 20:57 |
| noonedeadpunk | But let's re-write it then at once? I will be able to do that only tomorrow (hopefully) | 20:59 |
| noonedeadpunk | For me - kind of fine, can iterate later if needed. But I;'m also fine to make it more smart :) | 20:59 |
| noonedeadpunk | anyway, will check out now too | 21:00 |
| opendevreview | Merged openstack/openstack-ansible-os_murano master: Delegate package import to setup host https://review.opendev.org/c/openstack/openstack-ansible-os_murano/+/846147 | 22:47 |
| opendevreview | Merged openstack/openstack-ansible-os_murano master: Support service tokens https://review.opendev.org/c/openstack/openstack-ansible-os_murano/+/846163 | 22:47 |
| opendevreview | Merged openstack/openstack-ansible-os_murano master: Use global service variables https://review.opendev.org/c/openstack/openstack-ansible-os_murano/+/846165 | 22:47 |
| opendevreview | Merged openstack/openstack-ansible-os_murano master: Remove mention of haproxy-endpoints role https://review.opendev.org/c/openstack/openstack-ansible-os_murano/+/845802 | 22:52 |
| opendevreview | Merged openstack/openstack-ansible-os_neutron master: Deprecate allow_overlapping_ips option https://review.opendev.org/c/openstack/openstack-ansible-os_neutron/+/846272 | 23:00 |
| opendevreview | Merged openstack/openstack-ansible master: Stop NetworkManager on centos 9. https://review.opendev.org/c/openstack/openstack-ansible/+/846224 | 23:21 |
| opendevreview | Merged openstack/openstack-ansible master: [doc] Update link to Ansible doc in manifesto https://review.opendev.org/c/openstack/openstack-ansible/+/846358 | 23:21 |
| opendevreview | Merged openstack/openstack-ansible-os_mistral master: Add mistral-extra in the mistral venv https://review.opendev.org/c/openstack/openstack-ansible-os_mistral/+/844049 | 23:32 |
| opendevreview | Merged openstack/openstack-ansible-os_masakari master: Remove mention of haproxy-endpoints role https://review.opendev.org/c/openstack/openstack-ansible-os_masakari/+/845798 | 23:47 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
