Dispatcher Should Flush Its Log File To Disk (Doc ID 2276378.1)

Last updated on JUNE 22, 2023

Applies to:

Goal

Scenario: We were bringing up a Messaging Server on essentially a new system that was turned over to us, where it still had sendmail listening on port 25. This was not noticed since prior to turning it over, sendmail was supposed to be disabled by the sysadmin. So when dispatcher started, Sendmail was already listening on port 25 and the dispatcher logged this error:

10:36:31.28 ( 11): app_listen(SMTP_SUBMIT/587): bind failed with address in use, retrying in 1 second...

This is how it is supposed to work. However, dispatcher did not flush its log file to disk. So although it had logged the above error (several times), we did not see or catch it, particularly since the dispatcher log file appeared to be 0 bytes.

Ex.

With sendmail running, dispatcher log did not show a problem:

mailserver:/gcs/app/mail/msg/log$ ls -alrt disp*
-rw-r----- 1 imsadmin ims 0 May 25 10:36 dispatcher.log-0OQI00400RKFGP00

After getting Sendmail stopped, restarted dispatcher and now  the log:

mailserver:/gcs/app/mail/msg/log$ imsimta restart dispatcher
Stopping dispatcher server 5785 . done
Starting dispatcher server . 9031
mailserver:/gcs/app/mail/msg/log$ ls -alrt disp*
-rw-r----- 1 imsadmin ims 2712 May 25 10:49 dispatcher.log-0OQI00400RKFGP00  <<<<
-rw-r----- 1 imsadmin ims 0 May 25 10:49 dispatcher.log-0OQI00600S77YV00 

It was not until dispatcher was stopped/restarted that the then-previous log file suddenly was larger than 0 bytes and we could see the above log showing the problem.
We would like an Enhancement Request for the dispatcher to flush its log file to disk occasionally so these log messages could be more easily seen.

Solution

My Oracle Support provides customers with access to over a million knowledge articles and a vibrant support community of peers and Oracle experts.