Our great sponsors
-
-
-
InfluxDB
Power Real-Time Data Analytics at Scale. Get real-time insights from all types of time series data with InfluxDB. Ingest, query, and analyze billions of data points in real-time with unbounded cardinality.
-
No you didn't. They're doing what is often referred as "online schema change" using https://github.com/github/gh-ost (but the concept is the same than percona's pt-online-schema-change, or https://github.com/soundcloud/lhm.
No you didn't. They're doing what is often referred as "online schema change" using https://github.com/github/gh-ost (but the concept is the same than percona's pt-online-schema-change, or https://github.com/soundcloud/lhm.
I work on Canvas, and we mostly use straight rails migrations. We have some ActiveRecord extensions, linter rules, and careful manual review steps to ensure we do our migrations with minimal locking and other important things to avoid knocking over production databases, and we tag migrations as "predeploy" or "postdeploy" so they run at the correct time relative to when the code is deployed. But we have automation that runs predeploy migrations (just with rake db:migrate:predeploy) across hundreds of databases (and thousands of postgres schemas) before we deploy, and we run the post deploy migrations also automatically after the deploy (with rake db:migrate).