Shawn Gibbs on Application Management - All Comments

re: Application Dependencies and Chained Events

Shawn Gibbs [MSFT] — Wed, 07 Nov 2012 14:43:08 GMT

Keep in mind chained events are based on several things. First is that you capture events from all systems in the chain. Second that they have specific calling functions and answering functions, for example a client service model invoke talks to a wcf service with a server service model invoke. So for failures the exceptions needs to be in the caller and in the callee. Timings are also matched but I suspect the problem is one is handled and rethrown which places the exception in a part of the stack that doesn't appear to be a client or server.

re: Application Dependencies and Chained Events

deniz — Wed, 07 Nov 2012 14:22:16 GMT

Hello Shawn,

I just found your blog post while searching for a specific trouble that we are experiencing.

Any opinions why I can never have chained "Exception" events, while having many chained "Performance" events?

I mean when I click the "Distributed Chains" tab for an Exception Event (when the chained exception events are grossly obvious and observable by inspection) APM cannot seem to relate & chain them .

I suspect it is some incorrect setting or missing configuration on our account.

re: System Center .NET Application Alerts vs Events

Daniele Muscetta — Sun, 15 Jul 2012 18:57:52 GMT

I had another post on a similar topic (more on the "behind the scenes" of SCOM workflows in this regard than on the process, to be honest) which probably complements this one of yours - blogs.technet.com/.../custom-apm-rules-for-granular-alerting.aspx