ElastiCache for Redis quota that I did not expect
Today's entry will be very short, but I decided to share it on a blog as I did not find any other information about it online.
Recently, I've been working on upgrading ElastiCache for Redis cluster on a production environment and to make sure the operation will be safe, I was running multiple tests in various scenarios. After testing out node and Redis version upgrades on cluster with heavy load (sidenote: don't do this unless you're okay with losing some writes), I decided to aditionally test failovers. ElastiCache makes it really easy to test via UI Console. However, after 5 failover tests, I got an error saying that I cannot do anymore failovers that day, because I've already met failover quota. I quickly checked docs and to my surprise there is not a single mention of that quota. After trying it out again on a different Redis cluster, I confirmed that the quota is per account, not per cluster. Takeaway: If you want to do some extensive failover testing, make sure to plan it over a few days or have a few AWS accounts ready for testing. Thanks for reading!