Resque -> Sidekiq

[Matthew Hensrud]

Hey guys quick clarifying question on this.. we're starting the transition process in our monolith to sidekiq - but want to do it slowly across our jobs so we can carefully ensure our critical tasks (charging, etc) move over smoothly.

But reading here: https://github.com/mperham/sidekiq/wiki/Resque-Compatibility leads me to believe that as soon as we have a Sidekiq runner up in production, it will just start pulling Resque jobs out of redis and run them.  Am i reading that right?  

If so, we'll want to do something different to ensure only jobs we transition are picked up by Sidekiq - like a new namespace i guess.  Any other options, or am i looking at this wrong?

Cheers!

Matt

[Mike Perham]

You should use a different redis instance or database. Sidekiq is not designed to run in the same redis database as Resque so yes there are opportunities for unexpected behavior if they share the same key space. 

--
You received this message because you are subscribed to the Google Groups "Sidekiq" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sidekiq+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/sidekiq/5bd5d981-6542-4ccc-be82-5bc635dbc64bn%40googlegroups.com.

--

Mike Perham, CEO — Contributed Systems

[Matthew Hensrud]

Ah damn, ok that complicated the transition a bit but I understand.

[Matthew Hensrud]

Actually Mike while I have you (hi! we met at RailsCon a few years ago, i was really tall and still am)

If we also take the opportunity to redo our queues (which are currently a mess) and Sidekiq uses the new queue scheme, will that keep the jobs properly separated, or also risky from your POV?