Archive for the ‘Mail Routing’ Category

Configuring Spamassassin to show score for HAM on cPanel

Lately, I’ve been getting a storm of pesky junk mail on an account which is hosted on a dedicated WHM/cPanel server.

The WHM/cPanel server is a Centos 5 server with Exim mail server and Spamassassin anti-spam software installed and configured.

I went through the normal operation of staring at /var/log/maillog and /var/log/exim_rejectlog, as well as sending a test spam message to the server to make sure that the server was filtering spam properly and it seemed to be filtering messages ok. Read the rest of this entry »

Creating corrective actions with events4.nsf (with screenshots)

For a long time, I’ve been frustrated with Domino on the massive amount of messages that we sometimes receive over the weekend, late at night, or early in the morning indicating that a mail.box or some other database is corrupt.

Here’s a list of the common event generators/notifications we receive and a step by step guide on how I setup corrective actions to fix the problem and prevent the event task from spewing thousands of notification messages.

  • Event/notification: Error compacting mail\somedatabase.nsf: Database is corrupt — Cannot allocate space
  • Corrective Action: New event handler runs “load compact -c

  • Event/notification: Router: Mailbox file mail1.box is corrupt
  • Corrective Action: New event runs “load fixup

  • Event/notification: Unable to update activity document in log database for bookmark.nsf: Database is corrupt — Cannot allocate space
  • Corrective Action: New event handler runs “load fixup log.nsf”

This should significantly reduce the amount of messages that are generated by runaway event monitor notifications.

The following events cannot have event handlers run on them.

  • Event/notification: Unable to store document in MailServer1/MyDomain mail\usermailbox.nsf (NoteID = 1264886) from mail\usermailbox.nsf (NoteID = 1209126): Database is corrupt — Cannot allocate space
  • Reason: the command would have to be generated on remote server. No possible way to pass an argument to remote server on R6.5.6.

  • Event/notification: Unable to replicate MailServer1/MyDomain mail\usermailbox.nsf: Database is corrupt — Cannot allocate space
  • Reason: the command would have to be generated on remote server. No possible way to pass an argument to remote server on R6.5.6.

  • Event/notification: Database is corrupt
  • Reason: This event is generated by the router task based upon messages in the mail routing view in the log. The database that is corrupt is listed on a seperate line in the log than the “Database is corrupt” text, so there is no way to determine by which database to run a command on.

Read the rest of this entry »

Backscatter is killing us, how are you coping?

We use Message Labs for anti-spam prevention service. They are pretty good at stopping most of the spam, but not all.

For those of you who are not familiar, we point our MX records to Message Labs servers that they specify. Then Message Labs SMTP servers filter out spam and viruses using their own algorithms and forward on the good messages to our inbound SMTP gateway.

The problem is that Message Labs can’t block backscatter as they would block valid non delivery failures in the process.

How are you coping??

Pesky debugging and logging notes.ini settings

I’ve been investigating why our primary mail routing hub server is taking so long to route messages recently. The behavior started a few days ago (or maybe it’s been going on for a while and we didn’t notice it).

  • Messages will queue in the mail.boxes for anywhere from 5 to 15 (maybe longer) minutes.
  • They do NOT have the -check in progress- (which sometimes take place when Trend Micro ScanMail or IQSuite, our mail compliancy capturing software, is holding up the router).
  • Then suddenly they all route at once.
  • Then the process will start over again.

Read the rest of this entry »

Enabling message tracking for mail reports and monitoring.

I have just enabled Message Tracking on an R6.5.6 server.

All I had to do was to enable message tracking in the server’s configuration document.
I added LocalDomainServers to “Allowed to track messages” field, but left everything else as is.

Once that was enabled, I issued “load MTC” on the server in question.

I initially got these errors when the task loaded:

MT Collector: Initialization failed: File does not exist
MT Collector: Unable to use directory e:\Lotus\Domino\Data\mtdata: File does not exist

Read the rest of this entry »

Consulting

I'm currently available
for Lotus Notes / Domino consulting engagements.

LinkedIn

Connect with me:

LinkedIn

Advertisement
Advertisement
Categories