[bitnami/redis-cluster] No alternative to `REDISCLI_AUTH` that was deleted a few months ago

[Bengrunt]

Name and Version

bitnami/redis-cluster

What architecture are you using?

amd64

What steps will reproduce the bug?

We previously deployed redis-cluster on a Swarm cluster using the REDISCLI_AUTH variable (on a 6.2 image, several months back) and it was working fine.

When trying to upgrade our deployment today, using the latest 7.2 image, I was hit by the fact that the creator node cannot connect to the redis nodes now and this fails to setup the cluster indefinitely.

It seems that the REDISCLI_AUTH environment variable was removed by #39867 in the bitnami/redis-cluster/**/debian-11/rootfs/opt/bitnami/scripts/redis-cluster-env.sh script.

First of all, the fact that REDISCLI_AUTH deletion was not reflected in the provided docker-compose.yml file is quite misleading.

Secondly, I'm wondering how I'm supposed to pass on the password value to bootstrap my cluster without this variable now ?

Thanks!

What is the expected behavior?

No response

What do you see instead?

The creator node loops indefinitely trying to connect to itself (and I guess the issue would be the same with subsequent nodes):

Could not connect to Redis at redis-node-0:6379: Connection refused
Node redis-node-0 not ready, waiting for all the nodes to be ready...
Could not connect to Redis at redis-node-0:6379: Connection refused
Node redis-node-0 not ready, waiting for all the nodes to be ready...
Could not connect to Redis at redis-node-0:6379: Connection refused
Node redis-node-0 not ready, waiting for all the nodes to be ready...
Could not connect to Redis at redis-node-0:6379: Connection refused
Node redis-node-0 not ready, waiting for all the nodes to be ready...
Could not connect to Redis at redis-node-0:6379: Connection refused
Node redis-node-0 not ready, waiting for all the nodes to be ready...
...

Additional information

No response

[github-actions[bot]]

This Issue has been automatically marked as "stale" because it has not had recent activity (for 15 days). It will be closed if no further activity occurs. Thanks for the feedback.

[CeliaGMqrz]

Hi @Bengrunt

Sorry for the delay. Thanks for reporting this issue.

I understand that the variable has been removed to prevent issues in case no password was provided in our chart. However, the variable is still present in our docker-compose as it is officially supported by Redis.

During my tests, I was able to successfully set the password using the variable and log in with redis-cli. Therefore, I have not been able to reproduce the issue. The variable is behaving as expected, so you should be able to use the variable in the docker-compose. I would need more information to understand what is going on in your environment.

Thanks for your feedback.

[github-actions[bot]]

This Issue has been automatically marked as "stale" because it has not had recent activity (for 15 days). It will be closed if no further activity occurs. Thanks for the feedback.

[Bengrunt]

Hello @CeliaGMqrz

Sorry for the delay, I had to wait to spare a little bit of time to do some more tests, from a clean environment.
It appears that I must have messed up something when updating my redis.conf file during the upgrade.

The daemonize yes option was what was causing this issue.
I was able to deploy and create the cluster successfully, using the REDISCLI_AUTH variable, with daemonize no.

Sorry for this.
I will close this issue.

Best regards