On Friday, November 2 at 11:22AM EST, migrations were deployed to run on our PostgresQL database. One of the migrations added an index to a new column on a large table. This index was not added concurrently and at 11:33AM EST caused an outage due to a locked table. During this time, we paused the checkout queue to prevent user checkout errors for the duration of the outage. At 11:40AM EST, the migrations were completed and the platform outage was resolved. We are currently investigating further actions to prevent this happening in the future and are in the process of adding automated safety checks on possibly dangerous migrations.