sshd with zombie process on FreeBSD 10.0-STABLE

[Karl Pielorz]

--On 11 March 2014 10:45:18 +0200 Alexander Yerenkow <yerenkow at gmail.com> 
wrote:

> Could you try to change shell to simple "sh" and see if this helps?
> This could be due to non-atomic work with history file when multiple
> clients log off, causing csh produce large broken history files.
> This is not 100% your case, but you could at least try :)

Hi,

I also have a system that has this problem:

FreeBSD 10.0-STABLE FreeBSD 10.0-STABLE #0 r261289M: Thu Jan 30 13:33:35 
UTC 2014     xxxxxxxxx:/usr/src/sys/amd64/compile/GENERIC  amd64

The sshd processes lock in 'urdlck' - I set this box up a while ago as 
10.0-R - installed subversion, and updated to 10.0-S (r261289). It's been 
left 40+ days since then.

Very little usage - but I have over 120 'stuck' sshd daemons.

I initially posted to the freebsd-xen list (this box is running under Xen) 
as we've a number of other FBSD10 boxes that don't have the issue (but 
these get more usage, and frequent restarts). It now looks like it's not a 
Xen issue - but a FBSD10 issue...

I've left the machine 'as-is' with all these zombies / stuck sshd's - if 
anyone wants me to run anything against / on it?

I can't see that this has been fixed in a newer -stable?

Cheers,

-Karl