With the default reconnect==0 behaviour, we have been experiencing SEGV when Redis server connection attempts where too slow or servers temporarily not reachable.
The patches in this pull request are fixing this.
It could be that relying on redisAsyncFree only be called in DESTROY might introduce leaks for long rung processes with reconnects?
With the default reconnect==0 behaviour, we have been experiencing SEGV when Redis server connection attempts where too slow or servers temporarily not reachable.
The patches in this pull request are fixing this.
It could be that relying on redisAsyncFree only be called in DESTROY might introduce leaks for long rung processes with reconnects?