It is not recommended to fix this for the client if it can be avoided because it creates a 3-4hour performance hit on the server. An hour of this is 100% proc usage.
The reason it breaks is because the server was rebooted during the stats update, between 3AM – 5AM GMT.
*TS to Plesk2
*From the quick launch area (or start menu) click “Scheduled Tasks”
*Find the task marked “Plesk Scheduler Task #ad08f1e5-f9e2-4be7-b4f7-fc8248acf5e8” (Runs at 3:00AM Daily)
*Run this task
*Log off TS
This will repopulate the stats with yesterdays info.
__NOTE:__ This can only be run the day following the failure. That’s before midnight the following day. This is because it imports from yesterdays log file ONLY. If you run this task after midnight, it will not import the data you are looking for. Instead it will do the next day.