- Newest
- Most votes
- Most comments
What instance size are you using? If you are using Instances which have more than 4 cores - this comes with a feature called Enhanced IO enabled by default - which improves throughput by upto 83% and upto 47% reduction in latency with no application code changes required.
Additional details can be found in this Blog - https://aws.amazon.com/blogs/database/boosting-application-performance-and-reducing-costs-with-amazon-elasticache-for-redis/
There are a few other changes that can improve overall latencies but that could involve looking at the metrics.
answered 4 years ago
Unfortunately, there's no beating the speed of light. As you've seen, using Global Accelerator can make a difference but it isn't going to reduce latency enough for some users - because they are just further away.
Adding a read-only replica in a region closer to the users is a good idea; but that does mean that writes have to go back to the read/write replica (assuming you want synchronous writes) or you could have some process in the region local to the users that accept asynchronous write requests. Either way means a change in your application architecture.
Relevant content
asked 3 years ago

At the moment it's a r6a.xlarge. You do have a point. I did see that in the docs. I will spin up a new server and do some testing and return with my findings.