Thursday, 2026-07-02

opendevreviewOpenStack Proposal Bot proposed openstack/openstack-ansible master: Imported Translations from Zanata  https://review.opendev.org/c/openstack/openstack-ansible/+/99577704:28
noonedeadpunkSo https://review.opendev.org/c/openstack/openstack-ansible/+/995111 seems to be working really well07:33
noonedeadpunkI have additionally tested it with https://review.opendev.org/c/openstack/openstack-ansible-os_freezer/+/983870 where I change tempest plugin, and I see behavior changes in tempest for surte07:33
opendevreviewMerged openstack/openstack-ansible master: Imported Translations from Zanata  https://review.opendev.org/c/openstack/openstack-ansible/+/99577707:52
opendevreviewDmitriy Rabotyagov proposed openstack/openstack-ansible-os_freezer master: Remove v1 API  https://review.opendev.org/c/openstack/openstack-ansible-os_freezer/+/99171608:37
opendevreviewDmitriy Rabotyagov proposed openstack/openstack-ansible-ops master: Add boilerplate for observability collection  https://review.opendev.org/c/openstack/openstack-ansible-ops/+/99578708:55
opendevreviewDmitriy Rabotyagov proposed openstack/openstack-ansible-ops master: Add boilerplate for observability collection  https://review.opendev.org/c/openstack/openstack-ansible-ops/+/99578709:01
mgariepynoonedeadpunk: do we merge the Depends-On fix then ?10:29
noonedeadpunkmgariepy: we need https://review.opendev.org/c/openstack/openstack-ansible/+/995102/2 first :)10:30
noonedeadpunkor well10:30
noonedeadpunkwe can rebase as it's not "required" 10:30
noonedeadpunkbut it was hard for me to see the content without data being collected 10:30
opendevreviewDmitriy Rabotyagov proposed openstack/openstack-ansible master: Respect Depends-On for service projects  https://review.opendev.org/c/openstack/openstack-ansible/+/99511110:30
noonedeadpunkI just realized we can decouple these10:30
mgariepyokie.10:32
mgariepyvictorialogs still holding strong.10:32
noonedeadpunknice10:32
noonedeadpunkI'm just now got into looking on adding observability as a collection10:32
mgariepy4.35B, ~240gb on disk.10:32
noonedeadpunkreally nice result10:32
mgariepyfull text search on last 30 days is like 15 sec result.10:33
noonedeadpunkok, that's what I was worried about10:33
mgariepyall this on a 8 core 32gb vm.10:33
mgariepythe way it works is really clever ;)10:34
noonedeadpunkit sounds like a very nice alternative to opensearch then10:34
mgariepyand you also do need to pre-hash all your log.10:34
mgariepydo not**10:34
noonedeadpunkyeah, pre-hashing is unrealistic in world with dozen different sources10:35
mgariepyit's really impressive.10:35
mgariepyi do injest between 2M and 10M logs per hours.10:36
mgariepyif you need to extract some data from your query, it's also very easy.10:37
mgariepyand you can search on any field, not just limited to a few fields.10:39
noonedeadpunkthat's what fulltext mean?:)10:39
mgariepyabout 9.6TB down to ~240GB10:40
mgariepyi wonder what else is out there that we don't know about:) haha10:40
mgariepyour org might be ditching opensearch soonish.10:41
mgariepyopenstack + ceph is the most chatty cluster tho lol.10:42
noonedeadpunkOh, yes, sure10:44
mgariepyabout 330 host does generate a lot of data :)10:44
noonedeadpunkWell, switches can be also quite chatty I'd say10:44
noonedeadpunkbut nothing compared to these 210:44
mgariepyand i know i'm missing some logs10:44
mgariepyanyway. it's kinda nice to be able to track what happens to a single file in object store since a certain date :D10:46
mgariepyho. 380 hosts, 330 was only the last 5 minutes haha10:47
noonedeadpunkhave you also played with alerting on logs there?10:48
noonedeadpunkliek rate of errors from a service and things like that?10:48
mgariepynot yet, but i started building some grafana dashboard 10:48
mgariepyi do have a small graph for traceback count per service.10:49
mgariepybut my goal is to have all this at some point. but it might be more like a medium goal.10:50
mgariepya nice thing is that we can have dashboard for both logs and metrics at the same place.10:51
noonedeadpunkI'd argue you can do that with opensearch as well?10:53
mgariepyour opensearch cluster wasn't able to follow up my logs much.10:54
mgariepyand the pre-hashing thing was an issue.10:54
noonedeadpunkyes, right, scaling is bigger problem indeed10:54
mgariepywithout my cloud in it, i think when the openstack cluster is down for maintenance it takes a few hours to sync back the logs.10:55
mgariepyfew here is 12h+10:56
mgariepyand maintenance is also a log longer than a single server for victorialogs :) haha10:57
opendevreviewDmitriy Rabotyagov proposed openstack/openstack-ansible-ops master: Grafana: fix db creation, remove built-in plugin, refresh and introduce install docs  https://review.opendev.org/c/openstack/openstack-ansible-ops/+/98487817:17
opendevreviewDmitriy Rabotyagov proposed openstack/openstack-ansible-ops master: Grafana: fix db creation, remove built-in plugin, refresh and introduce install docs  https://review.opendev.org/c/openstack/openstack-ansible-ops/+/98487817:21
opendevreviewDmitriy Rabotyagov proposed openstack/openstack-ansible-ops master: Grafana: fix db creation, remove built-in plugin, refresh and introduce install docs  https://review.opendev.org/c/openstack/openstack-ansible-ops/+/98487818:09
opendevreviewDmitriy Rabotyagov proposed openstack/openstack-ansible-ops master: Grafana: fix db creation, remove built-in plugin, refresh and introduce install docs  https://review.opendev.org/c/openstack/openstack-ansible-ops/+/98487818:48
opendevreviewDmitriy Rabotyagov proposed openstack/openstack-ansible-ops master: Grafana: fix db creation, remove built-in plugin, refresh and introduce install docs  https://review.opendev.org/c/openstack/openstack-ansible-ops/+/98487818:49
opendevreviewDmitriy Rabotyagov proposed openstack/openstack-ansible-ops master: Grafana: fix db creation, remove built-in plugin, refresh and introduce install docs  https://review.opendev.org/c/openstack/openstack-ansible-ops/+/98487818:51
opendevreviewDmitriy Rabotyagov proposed openstack/openstack-ansible-os_freezer master: Add support for centralized scheduler deployment  https://review.opendev.org/c/openstack/openstack-ansible-os_freezer/+/98387018:52
opendevreviewJimmy McCrory proposed openstack/openstack-ansible-rabbitmq_server master: Set RABBITMQ_MAX_NUMBER_OF_PROCESSES variable  https://review.opendev.org/c/openstack/openstack-ansible-rabbitmq_server/+/99588319:01
opendevreviewDmitriy Rabotyagov proposed openstack/openstack-ansible-ops master: Grafana: fix db creation, remove built-in plugin, refresh and introduce install docs  https://review.opendev.org/c/openstack/openstack-ansible-ops/+/98487819:25
opendevreviewDmitriy Rabotyagov proposed openstack/openstack-ansible-ops master: Grafana: fix db creation, remove built-in plugin, refresh and introduce install docs  https://review.opendev.org/c/openstack/openstack-ansible-ops/+/98487820:29
opendevreviewDmitriy Chubinidze proposed openstack/openstack-ansible-openstack_hosts master: Add tar and unzip to distro packages  https://review.opendev.org/c/openstack/openstack-ansible-openstack_hosts/+/99589820:47
opendevreviewDmitriy Chubinidze proposed openstack/openstack-ansible-openstack_hosts master: Add tar and unzip to distro packages  https://review.opendev.org/c/openstack/openstack-ansible-openstack_hosts/+/99589820:48
*** zseguin_ is now known as zseguin23:53

Generated by irclog2html.py 4.1.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!