[DOC] Document whether multiple containers behind a deployment is a valid option

[SeanKilleen]

Change in the documentation

Hi all,

I'm implementing Flagd and while I see it can be deployed as a standalone container, it is less clear to me whether there are any implications of running multiple Flagd instances behind a service. If I deploy 3 Flagd instances to 3 nodes, and each is pulling from the same file (say, a multi-write persistent volume claim), are there any concerns? My thinking is that the 3 instances would synchronize and any http requests to the services would reflect the sync status of whatever running pod they happen to hit (which is fine for our use case).

I'm not in a position at the moment to move to the operator, though certainly considering that for the future.

Can someone evaluate my thinking above? I'm happy to PR the result of the conversation to the docs.

[SeanKilleen]

(PS realizing I don't need a PVC at all; will use a ConfigMap.)

[beeme1mr]

Hey @SeanKilleen, what you're proposing will work fine. Just keep in mind that mounting a ConfigMap as a volume may not update simultaneously.

[SeanKilleen]

Thanks for the confirmation, @beeme1mr! I'll do a PR for the docs and will close this out based on that.