View the original community article here
Last tested: Aug 26, 2020
If you see logs like this that just stop immediately after when starting up:
2020-08-26 15:15:44.974 +0000 [INFO|007e6|main] :: Starting Looker in clustered environment: Version 7.10.31-fb654d
2020-08-26 15:16:07.300 +0000 [INFO|007e6|db:*] :: Initializing database adapter for mysql
2020-08-26 15:16:07.308 +0000 [INFO|007e6|db.connection] :: Extended looker connection with validation timeout of 3600 seconds
2020-08-26 15:16:07.312 +0000 [INFO|007e6|db.connection] :: Created pool for name: looker max connections: 72 pools for this connection: 1
This means the customer is using Aurora for their internal database. Upgrading from version 2.06 to 2.07 causes this issue to occur.
Aurora is NOT supported as an internal database, so the solution is to switch to a standard MySQL internal db. Docs that call out what is supported here: https://docs.looker.com/setup-and-management/tutorials/clustering#mysql_application_database
As of August 31, 2020, Amazon updated all Aurora instances to 2.07, rendering them unusable as an internal db. Note that this has no effect on using Aurora as an actual db connection, only as an internal db
This content is subject to limited support.