i have searched everywhere online but have been unable to work out how to delete a worker or process from sidekiq/redis. I see there is a 'cleanup' function (Sidekiq::ProcessSet.cleanup) which should check all processes heartbeats and delete if it is no longer in contact. this function works but returns 0 when i KNOW there are processes no longer exist yet still present in console.
So basically, we have kubernetes which deploys sidekiq pods as workers that connect to our sidekiq to process jobs. this way we can scale pods rather than threads. If i delete a pod, i can see it is removed from sidekiq portal, however, somehow we have about 5 pods which have been deleted that have not been removed from the portal (they exist in redis still). i need to remove these they are causing issues and i cannot figure out how.
as a test i backed up redis and deployed to a test redis. the dead workers were still present there. i flushdb in redis and it actually removed them, so i know they exist in redis somewhere.help please!