Hello,
EFA seems to be running fine for the most part. The bulk of emails are processed and delivered normally.
We have 600+ messages building in the Milter Outbound on 2 different EFA servers going back to Nov 2021.
What can I do to determine the cause of the queue building?
How do I release the emails in the Milter Outbound queue?
Thanks!
Todd
Milter Outbound messages
- shawniverson
- Posts: 3783
- Joined: 13 Jan 2014 23:30
- Location: Indianapolis, Indiana USA
- Contact:
Re: Milter Outbound messages
First question, what version of MailScanner are you on?
Code: Select all
rpm -qa | grep MailScanner
Re: Milter Outbound messages
Sorry I should know better. eFa 4.0.4
MailScanner 5.4.2-3.eFa.e17.noarch
MailScanner 5.4.2-3.eFa.e17.noarch
- shawniverson
- Posts: 3783
- Joined: 13 Jan 2014 23:30
- Location: Indianapolis, Indiana USA
- Contact:
Re: Milter Outbound messages
So, this has me concerned. The fact that you mention when it started is vital.
Back in Nov 2021, this change was made:
I am going to check into this. https://github.com/MailScanner/v5/issues/591
In the meantime, please move any orphaned files from /var/spool/MailScanner/milterout to /var/spool/MailScanner/milterin to requeue.
Back in Nov 2021, this change was made:
So, your theory about restarts is plausible. If the child stops or dies before before processing the files it owns, this would be a problem.Limit queue processing to child-owned queue files in MSMail.pm
I am going to check into this. https://github.com/MailScanner/v5/issues/591
In the meantime, please move any orphaned files from /var/spool/MailScanner/milterout to /var/spool/MailScanner/milterin to requeue.
Re: Milter Outbound messages
Thanks for the update Shawn.
I am reasonably sure this has to do with restarting services. I have been performing this restart of the services for many years, since we first started with EFA, and dont recall ever having this problem in the past.
I have 3 EFA boxes and I configured 1x to only restart postfix, the other 2x restart postfix And MailScanner. I also changed the CRON times. The orphaned files in Milter Outbound now match the new CRON times.
I am using the commands
systemctl stop postfix
systemctl stop mailscanner
I WAS running the command to stop MailScanner twice to verify the service shuts down. I have changed that to 1x time in case the 2nd stop command was forcing MailScanner to quit before it finished its business.
The problem with the orphaned files continues on all 3 boxes, albeit less. This leads me to think it has to do with restarting Postfix and not just MailScanner. or the interaction between MailScanner and Postfix may contribute.
Sorry I am not more knowledgeable about this. My Linux skill are rather meager.
I will append to my script a copy command to move files from /milterout to /milterin.
Todd
I am reasonably sure this has to do with restarting services. I have been performing this restart of the services for many years, since we first started with EFA, and dont recall ever having this problem in the past.
I have 3 EFA boxes and I configured 1x to only restart postfix, the other 2x restart postfix And MailScanner. I also changed the CRON times. The orphaned files in Milter Outbound now match the new CRON times.
I am using the commands
systemctl stop postfix
systemctl stop mailscanner
I WAS running the command to stop MailScanner twice to verify the service shuts down. I have changed that to 1x time in case the 2nd stop command was forcing MailScanner to quit before it finished its business.
The problem with the orphaned files continues on all 3 boxes, albeit less. This leads me to think it has to do with restarting Postfix and not just MailScanner. or the interaction between MailScanner and Postfix may contribute.
Sorry I am not more knowledgeable about this. My Linux skill are rather meager.
I will append to my script a copy command to move files from /milterout to /milterin.
Todd