[02:53:15] Anyone here know how to get https://meetbot.toolforge.org/ back up? [02:53:20] ( bd808 perhaps? ) [02:53:31] I played with it on Oct 16 (see backscroll) but no luck [02:54:34] in particular so that office hours meeting logs are served and accessible [02:55:14] e.g. https://meetbot.toolforge.org/wikimedia-office/2018/wikimedia-office.2018-01-31-22.00.html [02:56:37] maybe worth turning into a light production service, e.g. backed up and served from swift, with a simple file browser in front, from a CORS/CSP restricted domain [02:57:12] or... come to think of it, we have this wiki thing, maybe we shoudl store it there [02:57:54] bunch of pages on metawiki should do the trick [07:36:04] Krinkle: seems like lighttpd is down? couldn't figured out so I set up https://tools-static.wmflabs.org/meetbot/ as a temporary solution [09:04:38] Hello, Im using "wb" cli to request wikidata API. Sadly, all my complexes requests finished with an TimeoutErrorException. Is there a way to avoid this limitation ? I already tried to paginate with LIMIT, ORDER & OFFSET, but it does not help. Any suggestions ? [09:34:34] !help [09:34:34] If you don't get a response in 15-30 minutes, please create a phabricator task -- https://phabricator.wikimedia.org/maniphest/task/edit/form/1/?projects=wmcs-kanban [09:40:44] Marc6: perhaps send an email to cloud@l.w.o [10:11:07] Ok merci arturo [10:11:22] thank you [10:11:30] 👍 [13:35:21] !log admin added dcar as projectadmin & user (T266068) [13:35:24] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [13:35:24] T266068: Onboard David Caro to Wikimedia Foundation as SRE in Cloud Services - https://phabricator.wikimedia.org/T266068 [13:35:32] !log tools added dcar as projectadmin & user (T266068) [13:35:34] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools/SAL [13:35:41] !log toolsbeta added dcaro as projectadmin & user (T266068) [13:35:43] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Toolsbeta/SAL [13:35:56] !log tools (typo: dcaro) [13:35:57] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools/SAL [13:36:06] !log admin (typo: dcaro) [13:36:07] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [13:56:45] !log tools.admin added dcaro as maintainer (T266068) [13:56:48] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools.admin/SAL [13:56:49] T266068: Onboard David Caro to Wikimedia Foundation as SRE in Cloud Services - https://phabricator.wikimedia.org/T266068 [15:25:36] !log planet deleting project as per daniel's note in the 2020 purge page [15:25:38] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Planet/SAL [15:56:45] !log tools.quickcategories made tool readonly to avoid issues during ToolsDB maintenance [15:56:48] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools.quickcategories/SAL [16:06:25] Hello cloud folks. I am need of a VM that itself can run the kvm hypervisor (i.e, VM nesting). Is that possible via WMCS? [16:07:44] inception! [16:09:20] totally [16:09:53] dancy: I don't think there is anything preventing you from doing it, other than a likely poor performance on the inner VM [16:10:47] I have attempted to get the kvm_intel kernel module to load but it doesn't work in the VM that I'm testing on. [16:10:57] But it does work on my personal system. [16:11:13] oh the virtual CPU might lack that flag/feature [16:11:48] nod.. I think that's the issue as well. The L1 VM needs to be started with non-default `-cpu` flag [16:12:22] `-cpu host` works for me. [16:12:41] that we don't support AFAIK. We might need to patch nova for that (our nova config, that's it) [16:13:08] dancy: would you like to explore the config and send a patch for us to evaluate? [16:13:14] Sure! [16:13:20] great! [16:13:49] If you can link me to a starting point I can go from there. [16:14:09] first thing I would do is to read nova documentation on how to modify the CPU flags [16:14:26] alright. Googling. [16:15:52] https://docs.openstack.org/nova/latest/admin/configuration/hypervisor-kvm.html [16:16:01] specifically https://docs.openstack.org/nova/latest/admin/configuration/hypervisor-kvm.html#nested-guest-support [16:17:53] !log tools.quickcategories tool is read-write again, ToolsDB maintenance was rescheduled (again) [16:17:55] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools.quickcategories/SAL [16:19:56] dancy: once you've studied the thing, the file that needs patching seems to be this one: modules/openstack/templates/rocky/nova/compute/nova-compute.conf.erb in the operations/pupppet.git repo. read more on how to submit patches for that repo here https://wikitech.wikimedia.org/wiki/Puppet_coding [16:20:17] Will do. Thanks Arturo! [16:20:39] 👍 thanks you! [16:56:24] arturo: Can you run `cat /sys/module/kvm_intel/parameters/nested` on cloudvirt1013.eqiad.wmnet and lemme know what it reports? [16:57:12] dancy: [16:57:16] aborrero@cloudvirt1013:~ $ cat /sys/module/kvm_intel/parameters/nested [16:57:16] N [16:57:28] dancy: I'm curious about your use case. What awesome thing are you hoping to build/test with nested kvm instances? [17:01:04] This is the train-dev environment. Primarily for developing/testing scap. We have some slides: https://docs.google.com/presentation/d/1BqLYKXZ51OJ1dlQysRrhdinKNdhlj4qU0Zui0RaOWYE/edit?usp=sharing [17:01:04] Another use case is https://phabricator.wikimedia.org/T250808 [17:03:00] That first slide deck is describing MediaWiki-Vagrant when I read it ;) [17:05:48] Nod. Similar [17:06:05] One of these days we may add a"why not MediaWiki-Vagrant?" slide. [17:07:29] mw-v is a single vm setup today, but I have made multi-vm setups for it before (like when I was developing scap) [17:24:10] In nova-compute.conf.erb I notice that the cpu_mode/cpu_model settings are within the <% if @enable_nova_rbd %>/<% if @enable_nova_rbd %> which doesn't seem appropriate. [17:24:41] Redo: In nova-compute.conf.erb I notice that the cpu_mode/cpu_model settings are within the <% if @enable_nova_rbd %>/<% end -%> block which doesn't seem appropriate. [17:27:28] so I'll be including a mod for that in my commit which adds nested guest virt [17:33:21] Also, how do we test this without disrupting a bunch of systems? [17:34:23] dancy: it can probably be tested with some manual work in the OpenStack testing cluster that the WMCS team has in codfw [17:34:43] That would be great. [17:35:38] I think we still have a way to cherry-pick puppet changes there. If we don't then we can probably use some hiera guards to control where the changes end up being used first [17:36:13] Alright. Preparing the commit now. [17:39:23] https://gerrit.wikimedia.org/r/c/operations/puppet/+/638146