Chen Jiaju

Sorry for late. It looks like a good feature, but I have no time to implement it at this time cause of my work.

And I will do it as soon as I finished my work in hand.

Sorry for late, I will check it soon.

> I mean it all depends on @importcjj > I still feel the reaper is locking the pool for too long. I think it is necessary to limit the execution...

Actually, I have long wanted to rebuild this connection pool. I want to remove all the locks and use a single-thread events loop just like Redis. But I doubt how...

If you can bring some things new, it's worth doing.

> o not require a lock? > Unsure what are the next steps here, if we merge this PR it will behave better in general but will lock the mutex...

> Interesting, the challenge will be to ensure the loop is never blocked for long. > SQLx has gone another direction, keeping the multi-thread but using https://github.com/crossbeam-rs/crossbeam to store the...

Hey! Thanks for your feedback. I tried but failed to reproduce your issue, but I can share some thinking with you. At the default situation, when getting a connection from...

A new version of Mobc-Redis was just released. Would you please switch to the `0.5.2` in your production environment and see if it still occurs the same strange issue?