About three months ago, we started noticing a slow down in AC response time. Two months ago it became critical. AC tech said to add more memory. I maxed out the server at 16Gb with 6GB in a virtual machine dedicated to AC. (We had run for over 2 years just fine on a 10GB server with 4GB in the VM.) Still having the same problem.

One of the techs noticed that amazingcharts.service.winservice.exe was consuming much more memory that it should. Restarting the service brought memory usage down from 755MB to just under 10MB. And the system response time improved almost immediately. Even with the file at max memory, the total memory used in the VM never got over 3GB. By experimenting, response time slows drastically when memory usage for that file approaches 400 MB. Watching the task manager performance tab, opening a chart would peg the cpu utilization at 100% for several seconds (10 ish secs, about 10 to 15 times longer than normal.)

When the service is restarted, CPU utilization while pulling a chart would never go over 50%, and is normally around 30. And the response time is normal.

I've watched AC tech check all of the markers. Other than that one file's ever increasing use of memory, everything else seems fine. The only suggestion so far from AC tech is add more memory. But, according to task manager, that is not the problem, as existing memory is not anywhere near fully utilized.

Nobody has addressed why memory consumption continues to go up for a simple scheduling program, nor the correlation between it's memory size, and severely degraded response times.

Anybody else having similar issues or have any idea what to do to troubleshoot the problem?

Thanks,
Henry