>>>I suspect that this is what causes the problem. I will ask the customer to check the CPU if/when the problem occurs again. Thank you.
>>
>>We had this happen at a couple of customer sites and couldn't determine what was causing it until we setup a 2nd timer to do nothing except periodically check the status of the 1st timer and enable it when necessary. A global switch turns the 2nd timer on or off. Whenever the 2nd timer found that the 1st timer wasn't enabled or firing, we logged everything including all running processes on the system. It helped to determine what was causing the problem.
>
>So, what did you find in the end? Unless it's a secret.

It was caused by a method down the line (levels deep) that would occasionally hang at one customer's location. The timer was doing some temporary file writes and the customer could set the location and the location they chose didn't have a reliable connection. We implemented a means of reconnecting when that happens and pausing until the connection was re-established.