We upgraded our two-container multisite, including a move from Postgres 13 to 15, and now gunicorn complains that it cannot connect to redis, saying . Error connecting to Redis on data:6379 (Redis::TimeoutError) .
Redis is running. The port is open; I can connect to port 6379 from the web_only container using, for example, ssh. But gunicorn can’t. We’re running Docker 20. Any ideas?
Occasionally with this kind of inexplicable Docker network connection issue disconnecting all containers from the network, destroying the network, re-creating it and connecting the containers up again can solve it.
Sometimes it’s this PLUS a rebuild of the containers.