1 Reply Latest reply on Jan 12, 2011 9:08 AM by bshopp

    Questions for developers of 1.1 version of OpsMgr management pack

    dvanorder

      Hi,

      I opened a support incident to report what may be a problem and also ask questions about the MP, specifically the feasibility of disabling certain features. Support closed the case and asked I post it here, so here goes...

      My primary function is working with OpsMgr and we'd like to use the 1.1 MP in our enterprise. We are a large OpsMgr environment and our Orion NPM deployment is also large. The Orion connector service is installed and having tested with the sealed MP, then examining the contents of the MP's, I think we can use it if we can turn off a few things that go beyond using the unsealed XML and changing enabled=true to false. Here are the details:

      1. Orion's web console is fantastic and network engineers and admins like it. These folks do not use OpsMgr, but we want to use the node/interface up/down monitor alerts and perhaps later on alerts from the Orion alert engine. I see the up/down monitor uses the value (0/1/2) of the Status perf counter to trigger. And this perf data is brought over from NPM to OpsMgr via the Orion connector service using the OpsMgr SDK. In looking at the modules in the MP, it appears there is no way to disable collecting all perf data except Status. Or is there? There is no point collecting data that will never be used for reporting.

      As a temporary workaround during testing, to keep data collection minimal I have removed the DW writer action for the collect perf data rule (just a comment). I was also wondering why the initial state of the node down monitor is Warning (1) when new nodes are brought in. They change to Healthy on the next polling cycle, but it's an unnecessary state change, that's all.

      I am not sure if this is a problem, but there is an event written to the OpsMgr event log that I wanted to mention. When the Orion service brings data over, it writes 3 events, this is the first one:

      Could not acquire the mutex for read access to the performance data file.

      Thanks---Drew