SMTP Server in Retry Loop

[Rich Loeber <rich@xxxxxxxxx>]

   I have a customer using their i5 server (V5R4) to send direct deposit
   advice notifications to their customers.  Once a month they drop a ton of
   these into the system to be sent out .... probably about 1000 in a single
   day.  They are using the SMTP and POP servers on the i5 to handle the
   outbound email tasks to lighten the load on the house mail server.

   I've noticed in the last few months that the SMTP server appears to go
   into a loop for several days following the monthly process.  Sometimes,
   this looping can take up to 50%-65% of the CPU cycles available.

   I turned SMTP journaling on today for a few minutes and I can see from the
   journal data that the SMTP server is just looping over and over again ....
   constantly .... on two different email addresses.  Why would it do that?

   Is there anywhere I can look to see how to interpret the SMTP journal
   information?

   I checked the SMTP server settings for retry settings and this is what I
   see:

   --------------------------------------------------------------------------

   Retries by minute:
     Number of retries  . . . . . .   3
     Time interval  . . . . . . . .   30
   Retries by day:
     Number of retries  . . . . . .   0
     Time interval  . . . . . . . .   0
   Retries by hour:
     Number of retries  . . . . . .   0
     Time interval  . . . . . . . .   0
   Retry remote name server . . . .   *NO

   --------------------------------------------------------------------------

   If I interpret this correctly, it should retry a failed send three times
   after a 30 minute wait.  But, the last attempt was done on June 26th and
   it was still looping through these today.

   Any ideas or suggestions?

   For starters, based on the journal information, I've advised the customer
   to fix the bad email addresses in their database, but there must be a
   better way.