rndc fails when machine swaps

[Will Yardley]

i've been noticing recently that when our nameserver is swapping (it
doesn't have that much ram in it), rndc will become unresponsive, and
bind will no longer respond to a SIGTERM.

decreasing the cache-size in named.conf is a workaround; however i
noticed this problem again yesterday when another application started
taking up a lot of memory / CPU, so i'm wondering if this is a bug.

runing bind 9.2.0 (non-threaded) in i386 linux (debian 2.2 / potato) as
unprivileged user.

has anyone else had similar problems?

-- 
No copies, please.
To reply privately, simply reply; don't remove anything.