Metrics and Reports Confusion

[MBaldwin]

I am kinda confused when it comes to interpreting some of the reports and correlating them to metrics.

In the trend report it looks like on March 16^th the VM CPU Effective demand jumped way up close to 80% of the allowable CPU util.   It looks like on Mar 2 and 9^th the VM CPU effective demand was down to 2 on average.

I am fine with that however if I throw in a screenshot of USAGE % over the last 30 days it does not look like this server is over trending CPU at all.  In fact it looks like what it is running today is about average.

Am I reading the report wrong in comparison to the USAGE % ?

I've attached screenshots of what I am looking at.

[mark_j]

The screenshot of the report you show is a little bit obscure. I always recommend using the table view to actually see values, as the 20 varieties of colors on the chart can become difficult to separate.

On your metric graph, go ahead and graph both usage % and demand % above each other. You'll see the difference. Also, tighten up the Y-axis and turn on the trend line on the metric graph, as what you show there is zoomed out too far and you can't actually see the trends.

[MBaldwin]

Ok I'll give it a shot.  But either way, the two charts/graphs are looking basically at two different things.  One is trend and the other is actual??

[mark_j]

That's right, one is trend and the other is actual. However, if you look at the trend-lines for demand in the metric graph, it'll roughly line up to what the CapIQ engine is showing/forecasting. Obviously there is some smoothing of the data when the cap planning is shown... the smoothing/trend criteria being controlled via your configuration policies.