AnsweredAssumed Answered

Best practice question: synthetic test every front end on every instance?

Question asked by lofgrenl on Mar 18, 2013
Latest reply on Apr 23, 2013 by bwcole
My company has a relatively traditional support structure that means that lower tiers attempt to determine the issue before escalating to upper tiers (which include myself as our APM support of course). Before the use of APM the only option was to synthetically (e.g. Wily Transaction Generator or various other products) each frontend on each service in order to isolate a failing service.

With the use of APM I had generally been trying to reduce the number of synthetic tests we are running because we are now monitoring every transaction. However specific failure of a frontend on a specific instance has become longer and less clear to diagnose.

Does anyone have a similar experience that they could share a best practice on how they decide if a front end on a specific instance should be synthetically tested? I am thinking I want to try and do it more as a matter of course (it will also help ensure our CEM definitions and certificates are remaining correct), but I may use a longer (e.g. 5-10 minute) interval in running a test. Thanks!