Recent AWS Customer Success Stories & Videos

More AWS Customer Success Stories...

« New Geo Library for Amazon DynamoDB | Main | Using AWS ElastiCache for Redis With AWS OpsWorks »


Feed You can follow this conversation by subscribing to the comment feed for this post.


Sounds like rds is almost catching up to what an ec2 mysql instance can do. One thing this article did not mention: load balancing. I understand in the past you could not load balance rds instances, has that changed? Say I would like to have a multi-az setup, with 5 read replicas. Would I be able to use an elastic load balancer in front of the read replicas so our app can hit any of those 5 for reads? It would be interesting to know how load balancing works with rds and its capabilities.


What about the case where RDS is a slave of an external database? Early on the replicate-in documentation says:

" The replication should be terminated when the data has been transferred and all applications and services have been cut over to accessing the Amazon RDS instance."

Is this a hard limitation? Our production MySQL database is outside AWS and going to stay that way for the foreseeable future. We'd love to ramp up RDS slaves for backup, reporting, and other purposes, but can't if replication is a one-time thing.


Any word on load balancing rds read replicas? I've been waiting on a response from amazon but nothing, so it makes it hard to even consider rds, unless of course for really small traffic sites that don't need anything but 1 rds instance. It sounds to me, from everything I've read, RDS is in its infancy and shouldn't be considered by anyone with a large traffic site and DBs. Our site gets millions of hits daily and we need to be able to scale the read replicas and be able to load balance the traffic. If the only solution for load balancing rds is by setting up another server running haproxy or mysql proxy, then this definitely isn't a viable solution, especially since now we'd be adding 2 extra servers for haproxy (or mysql proxy) to avoid single point of failure, which would mean 2 more servers to keep track of, monitor, maintain .... you get the picture. Any input on this would be greatly appreciated, although I'm starting to realize this post was more about a sales pitch then really saying what rds can "really" do.

The comments to this entry are closed.

Featured Events

The AWS Report

Brought to You By

Jeff Barr (@jeffbarr):

Jinesh Varia (@jinman):

Email Subscription

Enter your email address:

Delivered by FeedBurner

April 2014

Sun Mon Tue Wed Thu Fri Sat
    1 2 3 4 5
6 7 8 9 10 11 12
13 14 15 16 17 18 19
20 21 22 23 24 25 26
27 28 29 30