This message was deleted.

One common approach is to use a shared redis or memcached (or any other similar type of service). You probably want to aim for network+memory instead of network+disk for that kind of thing (depending on your needs ofc).

I would prefer a memory approach, ya.. It's a non-issue if the caching suddenly gets whiped, but they need to be equal over all running pods.

In that case, I would look at deploying a redis instance in the cluster and use that. You could optionally go for a HA setup using Sentinel, but that's often more complexity than it's worth for a simple cache.

Aight, I wasn't yet aware of https://redis.io/, so thanks for the feedback 😄