Using same S3 & Redis instance for multiple proxy clusters?

We have several clusters where we run Caddy v.2 as a reverse proxy. Now we are making changes to move from a single Caddy node setup to a cluster setup with multiple nodes.

Planing to use s3 storage module. It is already tested and looks OK for us.

Manipulating with storage “prefix” config it is possible and safe to use same S3 bucket for different clusters.

But in the same time module documentation said that it uses single node Redis as the lock.

So the question is:

Is it safe to use same Redis instance and db for different Caddy clusters or this will cause any issues?

If each cluster is managing a different sets of domains, it should probably be fine, but why not just set a different DB number if you have more than one cluster? Avoids the problem entirely.

As an aside, why not use GitHub - gamalan/caddy-tlsredis: Redis Storage using for Caddy TLS Data instead if you’re going to involve Redis anyways? I honestly don’t understand the point of using the S3 module when it still needs Redis for locking.

This topic was automatically closed after 30 days. New replies are no longer allowed.