[09:28:49] <_joe_>	 cross-posting from the k8s-sig channel: I've merged a change that should make CI for the operations/deployment-charts repository about 80-90% faster. Please let me know if you notice any weird behaviour, and please check the output of CI even when it passes.
[12:11:29] <marostegui>	 phabricator is down?
[12:11:46] <taavi>	 yep, looks like the same thing as yesterday with gitlab maintenance
[12:11:53] <marostegui>	 jelto: ^
[12:12:06] <jelto>	 yes that's expected, see my message from 11:00 here or in gitlab/releng channel
[12:12:55] <jelto>	 should be back in 5 minutes
[13:08:39] <XioNoX>	 jelto: is there a task about improving that dependency?
[13:09:01] <jelto>	 yes it's tracked in T333347 and should be removed soon
[13:09:02] <stashbot>	 T333347: Phabricator does not degrade properly when GitLab is unavailable - https://phabricator.wikimedia.org/T333347
[13:09:07] <XioNoX>	 thx!
[14:28:59] <ottomata>	 cwhite: you okay if I apply this for kafka logging clusters? https://phabricator.wikimedia.org/T334733#8823904
[14:31:18] <cwhite>	 ottomata: cc herron looks ok by me.  maybe start in codfw, watch for problems, then move on to eqiad?
[14:31:48] <herron>	 +1
[14:42:13] <ottomata>	 k ty, yeah shoudl be no problems, its just granting access to transactional writes to anonymous users;  no one is using transactional writes anyway
[14:43:11] <ottomata>	 doing
[14:46:09] <ottomata>	 done.
[14:55:17] <cwhite>	 ottomata: saw when that happened.  seems logstash experienced timeouts but recovered and reconnected.
[14:55:56] <cwhite>	 logstash automatically performs a rediscovery
[14:59:51] <ottomata>	 really? that's strange.
[15:00:10] <ottomata>	 is logstash using transactional writes?
[15:00:16] <ottomata>	 no, because those are consumers.
[15:00:17] <ottomata>	 hm
[15:13:34] <cwhite>	 link for reference: https://logstash.wikimedia.org/goto/9f41ef13154d18879d72a0d781d77544