spamd on server.xxxxx.xxx has failed

[keddie]

At around midnight last night, spamd failed, unleashing rolex watches and products for extending the usual parts of the anatomy on myself and my clients.

I now get an email roughly every 10 minutes with the subject:

"spamd on server.xxxxx.xxx has failed"

When I try to start spamd manually via SSH, I get the following output:

Errno architecture (i686-linux-2.6.9-023stab037.3-entnosplit) does not match executable architecture (i686-linux-2.6.18-ovz028stab053.4-enterprise) at /us r/lib/perl5/site_perl/5.8.8/Errno.pm line 11.
Compilation failed in require at /usr/lib/perl5/5.8.8/i686-linux/IO/Socket.pm line 17.
BEGIN failed--compilation aborted at /usr/lib/perl5/5.8.8/i686-linux/IO/Socket.pm line 17.
Compilation failed in require at /usr/lib/perl5/5.8.8/i686-linux/IO/Socket/INET.pm line 11.
BEGIN failed--compilation aborted at /usr/lib/perl5/5.8.8/i686-linux/IO/Socket/INET.pm line 11.
Compilation failed in require at ./spamd line 46.
BEGIN failed--compilation aborted at ./spamd line 81.

I am running the following:

WHM 11.23.0 cPanel 11.23.1-R24893
CENTOS Enterprise 4.6 i686 on virtuozzo - WHM X v3.1.0

Any help would be gratefully received!

Ked

[dwayrynen]

Same here - about midnight updates were applied to a system that has "never update" turned on.

Can someone from cpanel explain how a system that has never update can update itself and break things?

Thanks!

Darin

[cPanelDavidG]

I recommend anyone experiencing these issues (either of the issues mentioned), please let our technical analysts look into this for you: http://tickets.cpanel.net/submit

[keddie]

Ok,

Compilation failed in require at /usr/lib/perl5/5.8.8/i686-linux/IO/Socket.pm line 17.
BEGIN failed--compilation aborted at /usr/lib/perl5/5.8.8/i686-linux/IO/Socket.pm line 17.
Compilation failed in require at /usr/lib/perl5/5.8.8/i686-linux/IO/Socket/INET.pm line 11.
BEGIN failed--compilation aborted at /usr/lib/perl5/5.8.8/i686-linux/IO/Socket/INET.pm line 11.

the lines point to the code:

use errno

Something in errno.pm seems to be broken. After backing this file up and deleting it, spamd now runs.

The timing of this fits around the last automated cpanel update. This leads me to think that errno has been created / replaced and thrown a spanner in the works somehow.

[ddent]

We had this problem show up today as well. Fix was the following:

/scripts/perlinstaller Errno

[darren.nolan]

Same problem last night at 3am (all the way until 4am when I finally woke to the constant beeping of my mobile).

Using current - as there were a lot of changes in the version recently, I'm not totally surprised at this.

Reinstallig SA fixed my error.

I have yet to do another /scripts/upcp --force to see if SA breaks again (as it has since VPS Optimised, on my dedicated >_<), which has required a reinstall of SA each update - removing previous SA directories to ensure a clean install.

[cpanelnick]

The latest RELEASE build has this change:



Added /scripts/fix_sa_compile script to autocorrect Errno.pm problems
and invoke it before running sa-compile in update_sa_rules

[GarnetR]

I tried running all of these:

/scripts/perlinstaller Errno
/scripts/fix_sa_compile
update_sa_rules

But I am still getting the errors. Do I also need to reinstall SA? What are the exact steps to do that? - Thanks....

I also tried running /scripts/fixspamassassinfailedupdate, but even after running it, service status shows spamd failed, and running it again, starts with the same messages:

cp4:~ # /scripts/fixspamassassinfailedupdate
Mail::SpamAssassin::Plugin::Rule2XSBody test passed: enabled
Config File Count: 53
Spamassassin Update Failed, reinstalling!

----------
At the bash prompt, I tried just running

spamd &

I now see lots of messages to my console about Spamd getting connections, etc. The /etc/rc.d/spamd script doesn't seem to do anything, but just calling spamd directly seems to have started it... A few minutes later the service status is showing green.

[cpanelnick]

Please open a ticket @ https://tickets.cpanel.net/submit/

[rockster]

we have a server with the planet that was put back online and we updated cpanel. Since the update, we cannot get spamd running. upgraded from stable to to current. The Planet is still working on restoring to line 3000 servers.. anyone have an idea or can point us...

[cPanelTodd]

we have a server with the planet that was put back online and we updated cpanel. Since the update, we cannot get spamd running. upgraded from stable to to current. The Planet is still working on restoring to line 3000 servers.. anyone have an idea or can point us...

Could you open a ticket at https://tickets.cpanel.net/submit/ and send me the ticket ID and we will help get this resolved for you.

[rockster]

Thanks cPanelTodd.. I have tried everything for two days. I sent a dm with number.

[darren.nolan]

To reinstall I do pretty much the following (ensuring that I delete any left-over SA stuff before I start).

Please check these out for yourself - before using them.

killall spamd
cd /var/lib/
rm -rf spamassassin
cd /usr/share/spamassassin
rm -rf *
/scripts/perlinstaller --force Mail::SpamAssassin
/etc/init.d/exim restart
tail -f /var/log/maillog|grep spamd
/scripts/restartsrv_spamd --status

Basically - Stop any running spam services, remove SA directories, install SA, restart Exim, check that spamd is behaving in the mail log and ensure that the status of SA reports back proper settings.

[nettigritty]

changed to Stable build from WHM and then ran

/scripts/upcp --force

from SSH. Spamassassin working fine after that.

[cPanelTodd]

changed to Stable build from WHM and then ran

/scripts/upcp --force

from SSH. Spamassassin working fine after that.

I'd like to ask you to submit a ticket at https://tickets.cpanel.net/submit/in...eqtype=tickets if you continue to experience this issue on RELEASE or CURRENT builds. We would like to take a look at the issue further to be sure it doesn't re-occur.