I got a question today from a user wondering why the numbers that Domino reports and the numbers that VitalSigns reports for pending mail don’t always match.
On a healthy Domino server, a given pending message might sit in the mail.box file just a few seconds. New messages are constantly pouring in while the oldest messages are pouring out– in other words the Notes clients are depositing messages into the mail.box at the same time that the router task is taking them out of the mail.box for delivery into a user’s mail or transfer to another server for delivery.
The router operates on LIFO basis– last in, first out. As a result, the actual value of mail.pending is constantly in flux, even second to second. If you open the mail.box file on a server and press F9 every 10 seconds or so and you’ll see what I mean.
The mail statistics reported by Domino (mail.dead, mail.pending, mail.totalpending, mail.held, etc.) are a snapshot in time updated by the router task, approximately every 2 minutes. VitalSigns does not use these statistics, in general. VitalSigns connects to the mailbox or mailboxes, grabs the first document, and then loops through all the messages one by one as fast as possible, categorizing each message as to whether it is pending, held, or dead. We also look at how long the message has been waiting, in case you want an alert on stuck messages. On a busy server, by the time we get to the last message in the queue, most of the messages that we have already looked at have been delivered and removed from the queue.
The numbers we report, and the number that Domino reports, are snapshots in time, and the snapshots are not taken at the exact same moment. But both give a reasonable estimate of what is happening… assuming the router hasn’t hung, in which case our statistics will be better.
Mail Alerts
Our alert algorithm compares the current value of pending mail with the prior value, and only sends an alert if the value is over the threshold and if the current value is higher than the previous value. The pending mail value can be over the alert threshold, but if the value is going down since the last time we checked, VitalSigns won”t send an alert because we figure it was a spike and the router is catching up. In that case we just change the color to yellow and the status to “Pending Mail Warning“.
If the value of pending mail is over the threshold, and the number is still growing since the prior check, then Vitaligns does issue an alert. This condition is a “Pending Mail Alert“.
Hope that helps!
Comments are closed