Re: Redis/Jedis no single point of failure and automated failover

[Maurício Linhares]

Hi Derek,

Not sure how other people have solved this, but I ported the
"redis_failover" gem to Java and it has been working fine for us. If
you have ZooKeeper running it might be worth a try ->
https://github.com/officedrop/jedis_failover
-
Maurício Linhares
http://mauricio.github.io/ - http://twitter.com/#!/mauriciojr


On Wed, May 1, 2013 at 10:53 AM, Derek Organ <derek...@gmail.com> wrote:
> In a simple situation with 3 servers with 1 master and 2 slaves with no
> sharding. Is there a proven solution with java and Jedis that has no single
> point of failure and will automatically deal with a single server going down
> be that master or slave(automated failover). e.g. promoting masters and
> reseting after the failure without any lost data.
>
> It seems to me like it should be a solved problem but I can find any code on
> it just high level descriptions of possible ways to do it.
>
> Who actually has this covered and working in production?
>
> --
> You received this message because you are subscribed to the Google Groups
> "Jedis" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to jedis_redis...@googlegroups.com.
> For more options, visit https://groups.google.com/groups/opt_out.
>
>

[Derek Organ]

Thanks Maurício,

That looks very interesting and something I've been looking for. I'm currently experimenting with using a local proxy but the problem with that is it won't have quorum in deciding if a master is down. One alternative to that was to have the local proxy request master information from the sentinels.  In principle I think it can work but I'd love to use code that is in production instead of having to write it from scratch. 

Can I ask is the data in redis mission critical to the running of the application. We have a event stream with Akka which depends on redis in our design to work. My fear is maybe redis in a distributed sense is not reliable enough for this.

[Maurício Linhares]

Hi Derek,

It is mission critical but it is also reproductible, so for us having
the data itself is not an issue, the main issue is to make sure there
is always a "master" redis instance running so even if we lose some
data it doesn't matter that much as long as one of the servers is
promoted to be the master.

-
Maurício Linhares
http://mauricio.github.io/ - http://twitter.com/#!/mauriciojr