Elastic Cache Redis Configuration

Question

I'm confused. AWS does that to me. Reams and reams of documentation, and I can't find anything that tells me the simple stuff, in a simple way.

We are looking at using Elasticcache, with Redis, for our app. We need to be ready to scale a couple of things that don't work in Dynamo really well at high volume.

So I set up a Redis Replication Group. It created two Cache Clusters. I can connect to the end point defined for the replication group. I can write, and I can read data. I have figured out how to connect to it from my desktop, (another fine post elsewhere in stackoverflow.)

But I'm not quite sure what I have. I assume that the cache clusters are replicas of each other. I assume that when I write to the main endpoint, both clusters get updated with exactly the same data. I assume when I read from that endpoint, it figures out which one to retrieve from, based on performance, number of connections, queue size, and then makes the call to that replica.

However, I can also create "read replicas." Does that mean that I'm simply adding more cache clusters? I added one, looks exactly the same as the ones that were created already.

Answer 1

So it would seem that the two nodes created, one is a read replica, one is primary. Rebooting the primary causes complete data loss. It's interesting, as I thought that they were supposed to fail over to one of the remaining nodes.

More work to do.

• Bruno: according to this document : docs.aws.amazon.com/AmazonElastiCache/latest/UserGuide/… : "If you have no replicas and the node fails, you will experience total data loss."