Re: Problem with redis on AWS and transition to dynamodb

2016-12-06 Thread GMail
Well, my money are on AWS Elasticache, see if bypassing their proxy will do any good. It's probably a good idea to do so in testing environment with high load targeting elasticache and nothing else. Also it might be a good idea to write to AWS too, maybe there're other people with the same

Re: Problem with redis on AWS and transition to dynamodb

2016-12-06 Thread Andreas Kuhne
The way it is set up there are 2 cache nodes in a cluster - a primary and a secondary. Then I have a seperate address to connect to the cache cluster - so there probably is some sort of proxy there, but I can't manipulate it - it's in the AWS Elasticache offering. I don't want to bypass the proxy,

Re: Problem with redis on AWS and transition to dynamodb

2016-12-06 Thread GMail
You didn't answer about proxy. Also, do you reuse connections to redis.example.com ? Or create a new one for every request? If you aren't reusing them, maybe you should add timeout for connection creation and retry creating after timeout has exceeded. I know this

Re: Problem with redis on AWS and transition to dynamodb

2016-12-06 Thread Andreas Kuhne
Hi, Thanks for you answer - I don't think the problem is with django either, but thought that maybe someone else has run into the same issues on AWS. My cache cluster worked fine for over 2 year as well, but now, not so much :-) 2016-12-06 14:35 GMT+01:00 GMail : > Hi! > >

Re: Problem with redis on AWS and transition to dynamodb

2016-12-06 Thread GMail
Hi! Do you by any chance have any proxy on top of Redis? Seems like Django has nothing to do with it, though. I don't know anything about dynamodb really, but I think implementing Django cache with dynamo as backend should work. I've implemented Redis Cluster cache for my project and it works

Problem with redis on AWS and transition to dynamodb

2016-12-06 Thread Andreas Kuhne
Hi, We are having a strange problem with our redis elasticache instances on AWS. We have our sessions stored in a redis cluster on AWS. Our webservers sometimes get a: * Error connecting to redis.example.com:6379. timed out * Timeout reading from socket (I haven't included our real domain for