[07:13:25] <elukey>	 o/
[07:14:06] <elukey>	 FYI for the on-callers: with https://gerrit.wikimedia.org/r/1290808 I flipped the /ml prefix in the Docker Registry to a new S3 backend, that is currently hosting the same images that the same prefix holds on Swift. 
[07:14:55] <elukey>	 I tested a docker pull and everything looks good, so the change should be transparent. If you see any issue in image pulling from the ml-serve k8s clusters and I am not around, you can simply revert and run puppet on registry2*
[07:25:26] <marostegui>	 thanks elukey 
[14:16:33] <bjensen>	 is gitlab being oddly slow today, or is that me?
[14:17:24] <sukhe>	 bjensen: yeah, it's been flapping
[14:17:24] <sukhe>	 10:16:56 <+jinxer-wm> FIRING: [2x] ProbeDown: Service gitlab1004:443
[14:17:32] <bjensen>	 sukhe: ah, thanks
[14:21:48] <XioNoX>	 hnowlan: new oncall slots look great! and the fake "no-oncall" meetings worked fine
[14:48:40] <hnowlan>	 XioNoX: glad to hear it!