Answer the question
In order to leave comments, you need to log in
What to use for caching?
Hello! I want to ask advice from a respected habra community.
Now in a web project we use redis for the cache, but it has a drawback: the lack of master-master replication, which is critical for us. If at first it is possible to solve the issue with a dedicated cache server, then in the short term this does not seem to be a very good solution, and even a single point of failure is obtained.
I want to know what esteemed colleagues would recommend using a web application for caching data that has a normal master-master replication capability?
PS It is highly desirable not to write something like “take riak, it's cool”, but you still want examples of successful implementations, stricken problems, and so on.
Answer the question
In order to leave comments, you need to log in
Why is memcache not suitable? I don't know about replication, but it scales to a large number of servers.
Have you looked at Redis Sentinel ?
I haven't tried it myself, but they write that if the master falls, a new master is assigned from the slaves. This, in theory, will make the solution more fault-tolerant.
And for scaling, you can easily organize sharding. There is even a special python package for this.
Then you should take a closer look at Riak - www.insight-it.ru/programmirovanie/erlang/obzor-riak/
The master master is the worst thing an admin can come up with.
As already said - sharding; if you need a copy on hard - radish.
As for the memkesh - what kind of cache do you have that does not fit into 128 megabytes?
Couchbase - couchbase.com .
memcached-compatible, scales well, it is possible to make copies of the cache to disk.
Didn't find what you were looking for?
Ask your questionAsk a Question
731 491 924 answers to any question