If the Redis server goes down, and a :closed is returned, the pool isn't properly destroyed

[hiett]

If the remote Redis server goes down, whereby a :closed is returned for a command, the pool isn't destroyed. It is destroyed after the 15 seconds, but if there are frequent commands to SRH this timout is likely to be reached. This means that a broken pool will exist, with the only way to fix it being to restart SRH itself (or not have any commands for 15 seconds).