Jim Dory wrote:
> I came to work today and discovered our server not connecting for email,
> although the webserver part was running fine. I tried stopping and
> restarting dovecot and sendmail but that didn't work so a reboot fixed
> things. (I have since found that you need to stop a few things and maybe
> do a dbrecover) since there were some authentication errors in the log
> (maillog if I remember correctly).
This procedure will usually clear it:
# service dovecot stop
# service xinetd stop
# service dbrecover stop
# service dbrecover start
# service xinetd start
# service dovecot start
> Then after lunch email had stopped again. I tried a "service dovecot
> stop" but it failed because apparently it wasn't running. Doing a start
> worked - email started working again. Anyway, something is going wrong.
> Not sure where to look
We saw a spate of this occur late last week on between 30-40 individual
boxes. All last Friday. The above procedure worked on them.
> In the maillog I found this but not sure what it means:
>
>
> Jan 28 14:55:54 srv1 spamd[3077]: prefork: child states: II
> Jan 28 14:59:23 srv1 spamd[7651]: prefork: sysread(8) failed after 300
> secs at
> /usr/lib/perl5/vendor_perl/5.8.5/Mail/SpamAssassin/SpamdForkScaling.pm
> line 575.
> Jan 28 14:59:23 srv1 spamd[3077]: prefork: child states: II
> Jan 28 14:59:23 srv1 spamd[3077]: spamd: handled cleanup of child pid
> 7651 due to SIGCHLD
> Jan 28 14:59:23 srv1 spamd[3077]: spamd: server successfully spawned
> child process, pid 7717
> Jan 28 14:59:23 srv1 spamd[3077]: prefork: child states: II
>
> I've not updated in a month or more so just went ahead and did that and
> don't know if that will help.
This looks normal for a box running Mailscanner. At any rate, it does
not reflect the issue that you are having.
--
Chris Gebhardt
VIRTBIZ Internet Services
Access, Web Hosting, Colocation, Dedicated
www.virtbiz.com | toll-free (866) 4 VIRTBIZ