Red Hat Bugzilla – Bug 456255
AVC sendmail (system_mail_t) "getattr" to /var/run/sm-client.pid (sendmail_var_run_t).
Last modified: 2008-11-17 17:05:12 EST
Description of problem:
I find that, in the morning, sendmail no longer is sending mail. This has been
happening for, I believe, through the last 3 targeted policy updates. The error
pops up at 4:04:25, right after the cron.daily tasks have run. The 'mailq'
command shows no messages in the queue. However, doing an 'ls' on
/var/spool/clientmqueue shows mail waiting for transport. This results in a
silent failure of mail delivery.
Doing a /etc/init.d/sendmail restart gets the mail moving for another day.
Performing the 'restorecon' operations indicated by the troubleshooter does not
make any changes in the security contexts of the files indicated. The process
15057 indicated in the log is the 'sendmail: Queue runner' process.
I'm guessing that one of the cron.daily jobs may be the real culprit.
Version-Release number of selected component (if applicable):
Steps to Reproduce:
1. Check /var/spool/clientmqueue in the morning.
Mail is piling up.
Mail should be delivered.
Created attachment 312350 [details]
First of 2 avc troubleshooter logs.
Created attachment 312351 [details]
2nd troubleshooter log
The problem seems to be cron does a restart of sendmail and this does not
Fixed in selinux-policy-3.3.1-80.fc9.noarch
Closing all bugs that have been in modified for over a month. Please reopen if the bug is not actually fixed.