We are running a Teleflow applicaiton on 48 ports (handling approx 12,000 calls a day!) <br> <br>Until recently it was handling 8-10K calls a day with no problems, but with the increase in usage, a few periods of 100% processor usage (driven by TEleflow.exe) have emerged. <br> <br>Typically processor usage For Teleflow.exe is less than 1%. The periods of usage hitting 100% do not appear to co-incide with particularly high port usage. <br> <br>Do you have any ideas for identifying/fixing what is causing the processor usage to peak at 100%?

Typically (not necessarily, but often), it is not call receiving / inbound applications that will drive CPU usage higher. For one thing, they spend much of their time simply waiting for a call. Even if they primarily are in active calls, much (depending on the application, of course) of what the application does when a caller is on the line is handled by the telephony hardware, and not the computer itself. <br> <br>When you say that there are "periods" of 100% usage, how long are these? Are they quick 1 or 2 second spikes, or do they last for quite awhile? Does everything on the system slow down during these periods, or did you just happen to notice when watching task manager? <br> <br>Do you have applications doing jobs other than receiving calls? Perhaps applications that loop around, checking for tasks to perform? <br> <br>As you might have guessed from my questions, there are not really any specific things to check, as the causes of high CPU usage can vary a great deal. Sometimes they aren't even REALLY a cause for concern, IF they aren't causing any actual performance issues. (Any active application/exe may spike to 100% from time to time, it is only a concern if it it degrading performance) <br> <br>Have you noticed any particular steps / flowcharts / routines running in TeleFlow (by watching in the TeleFlow Monitor) when these peaks occur? For that matter, are other processes jumping to 100% usage when TFServer isn't?

Thanks for the pointers Chris. <br> <br>We have identified the bottleneck -- interestingly the Teleflow service application CPU usage peaked when the database (running on a separate machine) hit a "bottleneck" and the Teleflow service requested info from the database. <br> <br>We have "restructured" our application to fix the problem!

That is a little odd, but perhaps this is just something to do with how the TFServer/ODBC handling deals with a database bottleneck (and possibly, only on a particular database and/or driver). <br> <br>Thanks for letting us know what the issue was in the end!