Rudy Gevaert wrote:
/var/log/named/named.log {
[...]
create 644 dnsadm arcs
postrotate
if [ -f /var/run/bind/run/named.pid ]; then
/etc/init.d/bind9 restart > /dev/null
fi
endscript
}
[...]
/etc/cron.daily/logrotate:
error: bind9:8 unknown group 'dnsadm'
run-parts: /etc/cron.daily/logrotate exited with return code 1
Paul Martin wrote:
The odd thing is that logrotate's saying "unknown group" rather than
"unknown user".
Are you sure it is logrotate producing that error?
If it was, then where is it getting the "bind9:8" string from? And why
would the error only appear after the OP stopped throwing away error
messages from the restart script.
It seems more likely that the error is a result of security checks
performed by bind, and that bind is responsible for the message. It also
means that this is unrelated to logwatch, unless it is determined that
logwatch isn't executing the 'create' command in the expected way, but
that could be verified by looking at the actual file ownership and mode.
(If this problem still persists, I'd recommend posting a long directory
listing of /var/log/named/ right after a failed restart.)
After manually stopping bind and rotating the log, the user should be
able to do:
% touch /var/log/named/named.log
% chmod 644 /var/log/named/named.log
% chown dnsadm arcs /var/log/named/named.log
% /etc/init.d/bind9 restart
and see the same error external to logrotate, if logrotate isn't at fault.
-Tom
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
