Redis no longer accepting connections

Questions and Answers Deployment
1

INFORMATION

Relevant information to this issue:

  • databases: Redis version 7
  • Programming language and version: Ruby 3.1.2

ISSUE

We have preview deployments configured in qovery that were working without issue up until 10/12. Then any new deployments after that failed to connect to the redis database container in qovery. We have the alias setup for the internal redis database url and are receiving the following error.

__connect_nonblock': Address not available - connect(2) for [::1]:6379 (RedisClient::CannotConnectError)

Things we have tried to fix that did not work:

  1. making redis public and switching to the public url for the env.
  2. hard coding a connection
  3. redeploying the entire application

Describe step by step how to reproduce the issue
Try manually connect to the redis container

irb(main):003:0> redis.set("testkey", "hello world")
/usr/local/lib/ruby/3.1.0/socket.rb:61:in `connect_internal': Operation timed out - user specified timeout (Redis::CannotConnectError)          
/usr/local/lib/ruby/3.1.0/socket.rb:61:in `connect_internal': Operation timed out - user specified timeout (RedisClient::CannotConnectError)    
/usr/local/lib/ruby/3.1.0/socket.rb:61:in `connect_internal': Operation timed out - user specified timeout (RedisClient::CannotConnectError)    
/usr/local/lib/ruby/3.1.0/socket.rb:61:in `connect_internal': Operation timed out - user specified timeout (Errno::ETIMEDOUT)                   
irb(main):004:0>
2

Hi @mwkp , can you try to stop your Redis instance and start it again? (please do not restart).

3

For us this turned out to be a resource issue with the cluster. If anyone else has this problem take a deep dive into the cluster logs and you may see some resource issues.