[05:59:02] <_joe_> hi, I kinda lost track of where we are with the pki project [05:59:47] <_joe_> it was made available in deployment-prep, but I don't know the state in production [07:20:12] jbond42: o/ around? [07:45:31] _joe_: j.bond will surely answer more in detail later, but AFAIK he was planning to use debmonitor as a testbed for its first use in production very soon [08:31:28] _joe_: we have it in prouction now but now real clients, i hope to have at least some debmonitor useres using the new pki infrastructure by the end of the week, let me know if you need any more info [13:15:10] volans: quick review https://gerrit.wikimedia.org/r/c/operations/software/spicerack/+/676367 [13:15:49] dcaro: looking [13:16:24] for some definition of "quick"... :-P (+257 -2) [13:17:05] 210 of those are tests, so 47 lines of code [13:17:38] without comments and empty lines ~30 lines of code [13:24:31] dcaro: {done} [14:08:41] thanks [15:16:29] Hi SREs. I'm looking for someone to +2 https://gerrit.wikimedia.org/r/c/operations/puppet/+/679522 (enables a scap l10n-related feature flag in beta cluster) [15:20:03] dancy: see my cr comment [15:21:00] Saw that. Happy to make the change but can you give me some insight into what's up w/ eqiad1.wikimedia.cloud vs the wmflabs sectiosn? [15:21:56] old VMs have their names in .wmflabs, new ones are in .wikimedia.cloud, see https://wikitech.wikimedia.org/wiki/News/Phasing_out_the_.wmflabs_domain for more details [15:22:15] * dancy reads [15:23:59] ok. making updates now. [15:25:26] I notice that pybal_interface and canary_dashboard_url are missing from the wmflabs section. Seems like the two sections should be identical? [15:27:32] the config for wmflabs is somewhat split between wmflabs and eqiad.wmflabs, but the config that ultimately gets applied to hosts should be identical, yes [15:28:37] alright. new patchset pushed. [15:29:23] thanks! [15:29:44] dcaro,arturo - o/ not sure if you saw my ping in https://phabricator.wikimedia.org/T279342, but the more logging kafka brokers we replace the more rsyslog traffic wmcs hosts drop [15:29:49] without a change in the firewall rules [15:30:21] elukey: oh yes, saw the comment, asked something to XioNoX and the got distracted with other stuff [15:30:36] elukey: it is not exactly clear what rules are expected where [15:31:34] arturo: in homer_public the wmcs filters mention the old logstash kafka brokers (ips I mean), it is sufficient to add kafka-logging100{1,2,3} to them [15:31:49] you can copy/paste from what I added today [15:32:03] ok thanks [15:33:13] dancy: I can have a look [15:33:41] Thanks effie! [15:37:42] np :) [15:38:00] elukey: I think this is it: https://gerrit.wikimedia.org/r/c/operations/homer/public/+/679862 [15:38:12] dancy: you will return the favour one day [15:38:15] one day ... [15:38:35] :p [15:38:58] I hope so. [15:39:35] oh I was saying that with a Godfather accent [15:39:41] do you still hope so? [15:42:49] I'll always be loyal to the family [16:11:46] elukey: will follow up tomorrow [16:16:11] ack! [18:01:17] Hi all, I have a change https://gerrit.wikimedia.org/r/c/operations/puppet/+/670013 I'm planning on deploying, which will allow roles with apostrophes to be printed upon login (message of the day). It's small but will affect every host, so I wanted to announce it here to see if anybody has concerns before deploying [18:10:41] razzi: looks good! [18:15:20] shell escaping is such a pain in the butt [18:26:55] agreed. [18:30:22] Alright here goes! [19:02:37] Seems like everything is working with the motd change