chroot bind var/named/dev mounted twice and init script's stop hanging

[fwyKKCkQze2z]

Hi

I am running bind in a chroot'ed environment on a FreeBSD 11 system.

# pkg version | grep bind
bind910-9.10.6

For some unknown reason, I sometimes see /var/named/dev mounted twice. This seems often (always?) to be the case after a reboot. Unfortunately this is a production system, so I cannot reboot it at will to test.

# mount -v
....
devfs on /var/named/dev (devfs, local, multilabel, fsid 04ff007171000000)
devfs on /var/named/dev (devfs, local, multilabel, fsid 05ff007171000000)

Whenever this happens, stopping named via /usr/local/etc/rc.d/named restart fails (the command hangs forever); but I can only find one /usr/local/sbin/named -t /var/named process.

It seems that DNS resolution is working just fine anyway.

Once I kill the one process I find, one of the mounts is gone. After umounting the remaining one and restarting named ( /usr/local/etc/rc.d/named start works just fine then), everything seems to be OK/as expected again.

Here is an excerpt from my /etc/rc.conf:

named_enable="yes"
named_chrootdir="/var/named"

Is someone able to shed some light onto the question why this might be happening?

 

[SirDice]

Is someone able to shed some light onto the question why this might be happening?

The only thing I can think off is a failed restart. For some reason it may have failed to shutdown completely and then started again causing devfs(5) to be mounted twice.

 

[fwyKKCkQze2z]

Thank you for the reply and sorry for the late update. I have upgraded bind to 9.11 by now (because pkg told me that 9.10 will not be given much love any longer) and have not experienced the "double mount" any more since then. I am absolutely not sure if the upgrade really changed anything but I keep my fingers crossed for now.