6.0-beta14b Transport endpoint is not connected

I've been running 6.0-beta14b for a while and every now and then my user shares just dies, disk shares seem fine.

When I telnet to unraid and try to 'cd /mnt/user' I get an error " Transport endpoint is not connected ".

Syslog doesn't show any error messages and stopping/starting the array doesn't fix it. I have to reboot for it to mount the user shares again.

I'd love to know where to start looking. I can't be sure, but I think it's started to happen when I added a few docker images.

I have read that it could be something to do with number of open files, this is just after a reboot, before the problem:

lsof | wc -l

3948

ulimit -n

40960

I have 8 data disks, all XFS filesystem, 1 cache disk on BTRFS.

Running the following docker containers: CouchPotato(needo), Deluge(needo), MariaDB(needo), NzbDrone(needo), NZBGet(gfjardim), pyLoad(gfjardim), XBMCServer(wernerb)

My docker volume mappings all map to user shares, some of which are only on the cache drive - would that make a difference?

I have 9 User shares, one share is restricted to one specific disk, the others all exclude that disk. Some user shares are set not to use cache, others use the cache and 4 shares are set to only use the cache disk (appdata, docker, and two others).

I also have a few scripts that run via cron on unraid that uses rsync to backup other systems (rsync/ssh).

Any idea where I can start looking for the problem? Additional logging I can enable to see what's causing fuse/sshfs to crash?