Closed
Milestone
Jun 13, 2017–Jun 20, 2017
WoW ending 2017-06-20
Workout of the Week
Display by
Error fetching burnup chart data
Burndown chart
Remaining
Burnup chart
Total
Completed
Unstarted Issues (open and unassigned)
0
Ongoing Issues (open and assigned)
0
Completed Issues (closed)
46
- LFS mounts are bogus
- 2017-06-15: Pages Outage Post Mortem
- GitLab Pages not getting updated even after a successful deploy
- Mount all file shares for blessed/deploy nodes
- ES unresponsive for some minutes
- staging.gitlab.com is down
- Grafana private to public dashboards sync is broken
- Unable to login to staging due to Sentry and repository issues
- New ENV variables for version.gitlab.com app
- Add Log Rotation for version.gitlab.com
- Clean up old resources after the maintenance
- Failover Redis to the new nodes
- Move the IP address for GitLab.com and GitLab Pages
- Move Redis to the GitLabProd virtual network
- Evaluate ES requirements for GitLab.com search functionality
- add ability to clean up prometheus inventory on host/role/file removal
- ES log cluster heavy GC pauses and stalled heap memory usage
- Document vault caveats and findings
- Remove old ES dashboards from Grafana
- Decommission unused ES cluster
- Improve ElasticSearch visibility in Grafana
- Create an alert for staging.gitlab.com SSL certificate expiration
- Renew staging.gitlab.com SSL certificate
- Automate DO identities creation for pipelines
- Hold grafana version in the gitlab-grafana cookbook
- EXPLAIN ANALYZE query
- Put nfs-pages-01 into production
- Clarify _which_ issue is tested on stats.pingdom.com
- Perform various changes to PostgreSQL of which some require downtime
- Define sidekiq queues behavior
- Move the deploy nodes to the new virtual networks
- Prioritisation of Performance & Availability issues
- Run a test with more app servers?
- Document how to make dashboards
- What to do with University.gitlab.com?
- Create new file-storage1 server
- Move the load balancers from HAProd to GitLabProd
- Set PostgreSQL "random_page_cost" to 2 instead of 4
- Lock Redis and PostgreSQL roles to their current version of gitlab, separate from gitlab-cluster-base
- Reduce PostgreSQL max_connections from 3000 to 300
- Increase max_locks_per_transaction from 64 to 128 for PostgreSQL
- Set log_temp_files to 0 for PostgreSQL
- Enable log_checkpoints for PostgreSQL
- Reduce log_min_duration_statement from 2000 to 1000
- Remove limitation of only syncing dashboards that are labeled as public
- Move altssh.gitlab.com to the ARM environment
Loading
Loading
Loading