Setup a remote cache server like redis

[en0ma]

[anjor]

Why do we want it to be a remote server? Shouldn't the cache just be in memory?

[en0ma]

Why do we want it to be a remote server? Shouldn't the cache just be in memory?

3 possible reasons;

• so we can scale it separately and it won't eat the dependent service memory
• in HAAPI would, we would want it distributed
• other services will use it too

[anjor]

I see your point, but I wonder if this is overengineering it a bit; especially if we think of the use case where people might be running their own estuary nodes.

I am of the opinion we should try and do the simple thing first. What do you think?