[09:38:40] <XioNoX>	 fyi, follow up from Thursday's patch, https://gerrit.wikimedia.org/r/c/operations/homer/public/+/904544 is needed to, will be noop on the staging cluster
[10:02:53] <XioNoX>	 all done and working as expected https://phabricator.wikimedia.org/T328523#8749665
[10:17:49] <XioNoX>	 if someone could let me know if they see any issue with the staging cluster, otherwise I'll continue with mlstage + aux - https://gerrit.wikimedia.org/r/905170 this afternoon
[13:20:11] <XioNoX>	 rolling it to mlstage and aux
[13:23:31] <claime>	 ack
[13:28:31] <XioNoX>	 looks all fine for what I see
[13:29:11] <XioNoX>	 I'll need to sync up with a k8s pro before tacking DSE/Wiki/ML
[13:29:17] <XioNoX>	 tackling*
[13:45:05] <elukey>	 ack we can definitely use something like dse or ml-serve for a more in depth test
[13:45:31] <elukey>	 I am totally ignorant about the task, will try to read it to see if I can watch/check specific things when you roll out
[13:45:41] <elukey>	 but I guess mostly calico/bgp-related right?
[13:55:45] <XioNoX>	 elukey: yeah exactly
[13:55:56] <XioNoX>	 the main thing to check is if there are communication issue between the nodes
[13:56:27] <elukey>	 make sense yes
[13:58:35] <XioNoX>	 but if it works for all the other clusters no reason it doesn't for those :)
[14:52:30] <akosiaris>	 XioNoX: I can run my test suite tomorrow against a cluster
[14:52:35] <akosiaris>	 "test suite"
[14:52:49] <akosiaris>	 I guess preferably one with >1 nodes per row ? 
[14:59:26] <XioNoX>	 akosiaris: yep
[14:59:28] <XioNoX>	 thanks!
[14:59:35] <XioNoX>	 akosiaris: can I ping you so we sync up?
[15:00:00] <akosiaris>	 yup
[15:43:16] <elukey>	 akosiaris: one thing that I forgot to ask during the last meeting is if we have plans for smaller projects like Docker runtime replacement 
[15:58:52] <ottomata>	 o/
[15:59:29] <ottomata>	 anyone know why regular k8s pod/container metrics are not available in prometheus for all namespaces in dse-k8s-eqiad?
[15:59:29] <ottomata>	 simplest dashboard example here: https://grafana.wikimedia.org/goto/PAb1YVYVz?orgId=1
[15:59:29] <ottomata>	 https://www.irccloud.com/pastebin/2pxrbAPv/
[16:00:19] <ottomata>	 e.g. flink-operator is a namespace and has running pods, but label_values(container_start_time_seconds,namespace) apparently does not return that namespace 
[16:15:46] <akosiaris>	 I 've answered already in -serviceops fwiw 
[16:16:02] <akosiaris>	 for everyone else's convenience :-)
[16:43:40] <ottomata>	 ^ ty
[16:43:57] <ottomata>	 for future, is it better to ask these kinds of questions here in k8s-sig?