January 27th-31st - No deploys at all (including train), All-Hands
Need some help (from CPT? Maybe?) looking at jobqueue on beta cluster which is evidently running slowly: phab:T241448
SRE:
The eqiad kubernetes cluster will be reinitialized on Tuesday Jan 14th 2020. The duration is expected to be up to 4 hours during EU morning. All requests will have been routed beforehand to codfw to avoid causing issues to users. Please refrain from trying to deploy to that cluster during that time period.
OKRs in final drafts, reach out if your team requires SRE involvement
Main focus is on releasing the first version of topic matching for suggested edits module. Initial release utilizing morelike targeted for before All Hands.
Core: MW Job consumers sometimes pause for several minutes phab:T224425
Updates:
Re-import OSM data at eqiad and codfw to temporarily fix current OSM replication issues (including reducing replication interval, which might fix the issue long term) - phab:T239728
Migrate and redeploy discovery platform analytics jobs with kerberos integration - phab:T241011
Research on recommdendation-api migration to kubernetes. Working on it
Product infrastructure on creating a buster nodejs image. Working on it
Release Engineering - ServiceOps: Rolling out PHP 7.2.26 to CI is blocked on creating a PHP 7.2.2.6 package phab:T241224
Updates:
The eqiad kubernetes cluster will be reinitialized on Tuesday Jan 14th 2020. The duration is expected to be up to 4 hours during EU morning. All requests will have been routed beforehand to codfw to avoid causing issues to users. Please refrain from trying to deploy to that cluster during that time period.
OKRs in final drafts, reach out if your team requires SRE involvement