Fallback for Resque,Sidekiq

[Abhi]

We have a Redis master-slave setup but not in cluster arrangement at the moment. When master goes down, we promote a slave to master manually. As we work on HA for redis, we currently are looking for a fallback solution(when redis goes down) for Sidekiq, Resque which are mission critical for us. As with Resque, Sidekiq having different queues ensures FIFO at least in that queue. We looked at having an SQS queue where we push jobs when Redis goes down. As the order of execution is important for us, some jobs remain in resque while queued jobs in SQS get executed and when Redis comes back it starts executing Resque and Sidekiq jobs while new jobs get enqueued. We haven't found an elegant redis cluster solution for now and having a fallback for Sidekiq and Resque is very essential. I would like to know if there is a way to control the execution order using SQS as a fallback or any solution which ensures the above. Any help would be appreciated. TIA

[Michel Martens]

On Fri, Mar 18, 2016 at 7:42 AM, Abhi <natraj....@gmail.com> wrote:
> We haven't found an elegant redis cluster solution for now and having a
> fallback for Sidekiq and Resque is very essential.

Would Disque work for your use case?

https://github.com/antirez/disque

[Abhi]

Disque still not being production ready, we can't use it for our current usecase.