[10:29:23] <Krinkle>	 traffic switch looks quite smooth so far: https://grafana.wikimedia.org/d/000000066/resourceloader?orgId=1&from=1598819577873&to=1598947824621
[14:26:09] <Krinkle>	 navtiming journal on webperf1001.eqiad:
[14:26:15] <Krinkle>	 > Sep 01 14:03:15 webperf1001 python[1357]: 2020-09-01 14:03:15,213 [INFO] (is_master:232) eqiad is not the master datacenter, disabling consumer
[14:26:15] <Krinkle>	 > Sep 01 14:03:15 webperf1001 python[1357]: 2020-09-01 14:03:15,214 [INFO] (run:755) No longer running in the master datacenter
[14:26:15] <Krinkle>	 > Sep 01 14:03:15 webperf1001 python[1357]: 2020-09-01 14:03:15,214 [INFO] (run:756) Stopping consuming
[14:26:15] <Krinkle>	 > Sep 01 14:03:26 webperf1001 python[1357]: 2020-09-01 14:03:26,226 [INFO] (run:714) Checking whether datacenter has been promoted
[14:26:38] <Krinkle>	 navtiming journal on webperf2001.codfw:
[14:26:43] <Krinkle>	 > Sep 01 14:03:12 webperf2001 python[1471]: 2020-09-01 14:03:12,314 [INFO] (run:714) Checking whether datacenter has been promoted
[14:26:43] <Krinkle>	 > Sep 01 14:03:12 webperf2001 python[1471]: 2020-09-01 14:03:12,316 [INFO] (is_master:228) codfw is the master datacenter, going live
[14:26:43] <Krinkle>	 > Sep 01 14:03:12 webperf2001 python[1471]: 2020-09-01 14:03:12,316 [INFO] (run:717) Starting Kafka connection to brokers
[14:27:36] <Krinkle>	 that's the first time our services are following the switchover and looks to be working fine.
[14:28:06] <Krinkle>	 Mostly thanks to what Ian implemented (overwrite and deterministic buffer, so that they don't need to be perfectly aligned)
[14:30:20] <godog>	 neat
[14:31:09] <Krinkle>	 automatically based on polling the MW config for etcd, which is not perfect, but works for now (it's not related to MW in any way)