I have a VM GPL install under openvz on Ubuntu 10.04.2. Periodically I’d like to be able to send myself a logwatch report. When I do this is the result:
# logwatch --mailto=geek@uniserve.com
perl: warning: Setting locale failed.
perl: warning: Please check that your locale settings:
LANGUAGE = (unset),
LC_ALL = (unset),
LANG = "en_CA.UTF-8"
are supported and installed on your system.
perl: warning: Falling back to the standard locale ("C").
Killed
Killed
Killed
system 'cat '/var/log/messages' | /usr/bin/perl /usr/share/logwatch/scripts/shared/expandrepeats ''| /usr/bin/perl /usr/share/logwatch/scripts/shared/removeservice 'talkd,telnetd,inetd,nfsd,/sbin/mingetty,netscreen,netscreen'| /usr/bin/perl /usr/share/logwatch/scripts/shared/applystddate ''>/tmp/logwatch.85OCq98X/messages' failed: 35072 at /usr/sbin/logwatch line 870.
I know that this is not a perl support site, but perl was installed as a part of installation of VM. I have not tinkered with the installation and would much appreciate advice or pointers.
in the interval from my last post the failcount on that line has increased to 13 for unknown reasons; running logwatch again does not alter the value.
running apt-get install locale tells me there is no locale available to install. But locale IS already installed and your suggestion to try installing english language support got lots of action and seems to have worked out ok. Likewise with setting a default locale to GB worked fine after I read the manual about having to re-login for changes to take effect.
So far so good and I have no more locale errors.
It is always tempting when the computer misbehaves to think that there is a problem with the computer. But often what is actually the case is that there are several problems. And that seems to be the case here.
The logwatch command still fails, though not with locale errors and still at line 870 as above.
It develops that postfix is not running although dovecot is. The VM UI status page shows this and restarting in either the UI or the commandline, using service postfix start both fail, although the command line version returns [OK]
syslog shows this:
Nov 4 17:01:15 cl28810 postfix/master[11337]: fatal: bind: private/lmtp: No such file or directory
Nov 4 17:01:36 cl28810 postfix/postfix-script[11515]: warning: not owned by postfix: /var/spool/postfix/private/anvil
Nov 4 17:01:36 cl28810 postfix/postfix-script[11516]: warning: not owned by postfix: /var/spool/postfix/private/lmtp
Nov 4 17:01:36 cl28810 postfix/postfix-script[11517]: warning: not owned by postfix: /var/spool/postfix/private/scache
Nov 4 17:01:36 cl28810 postfix/postfix-script[11518]: warning: not owned by postfix: /var/spool/postfix/private/maildrop
Nov 4 17:01:37 cl28810 postfix/postfix-script[11549]: starting the Postfix mail system
Nov 4 17:01:37 cl28810 postfix/master[11550]: fatal: bind: private/lmtp: No such file or directory
Looks like a permission issue, but I don’t see how this came about. logwatch used to work ok and it isn’t like I’d intentionally do anything to alter that.
So if you have the patience I’d like to get postfix back up, too.
Thanks for the reference to the ubuntu docs, quite worthwhile; I’ll be reading what it has to say about postfix, too.
in the interval from my last post the failcount on that line has increased to 13 for unknown reasons; running logwatch again does not alter the value.
Well, it’s good news that it’s not related to logwatch… but it sounds like other applications you’re running are bumping into a memory barrier and ultimately failing.
Looks like a permission issue, but I don’t see how this came about.
Hmm, Postfix permissions should all by okay by default – those are some unusual errors! What do these two commands show:
ls -l /var/spool/postfix/private/
rpm -qa | grep postfix
It may just be a simple matter of changing the permissions of the files in /var/spool/postfix/private/, but it’d be good to see how those are currently set before changing them.