We continuously have an issue with our PI Log files on the PI Data Archives growing extremely large.

Every 2-3 weeks, we have an issue with out log files on our PI Data Archive growing extremely large. Working with PI Tech Support a while ago, we were told it was due to have multiple instances of the PI Message Subsystem ( I think ) running at once. This completely bogs down our entire PI system and it takes a complete PI System reboot to get everything back up and running.

 

This happens every 2-3 weeks, always sometime Wednesday into Thursday. Is anyone else having this issue and are there any suggestions on what may be causing this to happen?

 

During our recent PI upgrade, we were told it could have to do with how Microsoft updates are installed and have since changed that process but it's still occurring.

  • There should only be one instance of the PI Message Subsystem on a given Data Archive (there shouldn't be a way to have more than one). What version are you running? The only known issue I have seen with the logs growing rapidly was due to a bug in the Data Archive where (typically) backfilling events would cause the message throttle to become broken that would limit the number of messages written related to writing values at a timestamp that already exist. This was supposedly fixed in 2018 SP6 though.