redisson
ability to mark redis sentinel cluster as down and stop trying to reconnect to it on every request
hello, we faced the following issue on production, we have a redis sentinel cluster of 3 servers ( so 3 redis sentinels and 3 redis services)
we noticed the following behavior, if the redis cluster was down and the application started then we don't create a connection and consider redis as down for the lifetime of the application until a restart (this is done programmatically by us).
however if both the service and redis cluster were up and connected, then if redis cluster goes down, the service is slowed down considerably.
by investigating the logs i figured its trying to reconnect to redis on each redis related command in the operation (with its timeouts and retry attempts)
so say we had a rest api that throughout the call will call redis 5 times, and redis cluster was down. redisson will try to write the command and get an error 5 times if the retry attempts was set to 0 and more if it was set to more.
this caused immense slowness on our end and we would rather in such cases rare as they maybe, to just mark redis as down and continue with calling the database rather than attempt to connect with redis again.
it would be nice if there was a way to say check if redis is back up every 100th call to redis and then attempt to reconnect. this way its not constantly trying to reconnect to redis.
i tried to play around with the configurations and nothing gave me the desired behavior, i am not sure if such behavior is supported by the library or not. I would apricate any help regarding this question
to just mark redis as down and continue with calling the database rather than attempt to connect with redis again.
Which Redisson objects do you use?
