Closed amedeiros closed 5 years ago
Also side note this could be completely because of the multi threading in my app.
this quite hack, i think this should not be the case. Should be some wrong usage, your env really threaded or just concurrent? what pool size?
A hack indeed. Well currently I spawn a few different fibers each one using the the same PooledClient instance with the default size of 5.
Closing this was my fault.
Actually, when you use same PooledClient between fibers, it should not raise this error even if pool small (can only be problem if you use threads). When connection busy, it wait or raise PoolTimeoutError error. So, i only can imagine some crazy combination of multi
or pipeline
with other commands in concurrent env.
When using commands
zrangebyscore
andzremrangebyscore
its possible for the command to return an Int64 from redis. inRedis::CommandExecution::ValueOriented#string_array_command
andRedis::CommandExecution::ValueOriented#array_or_nil_command
.Yes I am in a multithreaded environment and yes I am using the
PooledClient
(saw the other ticket).Making these changes stopped the exceptions. I can make a pull request if this is the appropriate fix thanks.