[04:57:02] !log codesearch restarting to deploy T264557 [04:57:06] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Codesearch/SAL [04:57:06] T264557: Add codesniffer to codesearch - https://phabricator.wikimedia.org/T264557 [12:09:07] Is the grid engine for toolforge going away in the future in general, or just for web services? [12:09:34] i.e. are batch jobs also supposed to run on k8s in the future? [12:14:10] bennofs: we are in the middle of figuring that out now. My take is that yes, for webservices, k8s is the way to go [12:14:42] other grid use cases will be evaluated as well [12:24:12] i know for web. but I'd be interested in the non-web case [14:24:03] bennofs: if you're starting a new service today you're better off using k8s unless you have very specific weird requirements. k8s is generally a better and more actively maintained product. [16:05:52] bennofs: also, it is possible to use k8s for a jobs service now, but it sucks. We are hoping to improve that experience before that becomes a recommended practice T251917 [16:05:52] T251917: Design the Jobs service in k8s - https://phabricator.wikimedia.org/T251917 [16:06:08] Grid's not going anywhere until we can make something like that actually good [17:40:17] !log admin `service uwsgi-labspuppetbackend restart` on cloud-puppetmaster-03 (T264649) [17:40:19] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [17:40:20] T264649: openstack-browser not showing any puppet classes - https://phabricator.wikimedia.org/T264649 [17:44:29] !log tools.openstack-browser Purged cache of puppet roles for T264649 [17:44:31] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools.openstack-browser/SAL [18:44:43] !log planet shutting down instance pk8s - not in use since 2019 [18:44:45] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Planet/SAL [18:46:05] !log planet marked project for deletion in 2020 purge [18:46:06] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Planet/SAL [18:50:43] what's the difference between "m1.large" and "g2.cores4.ram8.disk80"? I see that one of the two cvn m1.large instances has retroactively become "g2.cores4.ram8.disk80" but not the other. just making sure that's all fine and as intended :) [18:52:05] I think that's the new naming schema, I think it was in a blog post... [18:53:03] Krinkle: the g2.* names are bound to Ceph, the old m* names are not [18:53:34] As I move VMs to ceph storage they're also getting 'resized' to a g2-style flavor that has the same stats as their previous flavor. [18:54:47] https://lists.wikimedia.org/pipermail/cloud/2020-September/001209.html [18:57:15] andrewbogott: thx, [18:57:28] yeah, I checked the two blogs, but missed this mail [18:57:29] thanks [19:00:39] I don't see cloudvirt1030 or cvn-app8 on the list. [19:00:53] but I suppose that'll happen later this month, np [19:01:26] Krinkle: yeah, going to take a breather after tomorrow's move and check capacity and such. [21:58:12] !log cloudinfra setting "mtail::from_component: true" on both mx-out servers to make puppet work again [21:58:15] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Cloudinfra/SAL [22:28:55] !log shinken deleted the shinken module [22:28:55] mutante: Unknown project "shinken" [22:29:06] !log monitoring deleted the shinken module [22:29:07] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Monitoring/SAL