A brief overview of how to create a Performance Metric and how to added them to a dashboard.
In this video we're going to cover the basics of the performance metrics report attachment. Now this report attachment has a lot of options available to the user. We're not going to cover everything in great detail but we'll get you comfortable populating your dashboards with performance metrics report attachments. At a very high level, you could simply select any device in the system, select an object, then an indicator such as a availability and hit finish. You then have a graph of the availability of the device in question, because we ship with a number of I guess, default parameters. Time span of today, no baseline etc., now I could make this more complicated and I will.
Device groups always tend to give us context, what I'm going to do is filter by the Cisco device group. Select a device and then select an object from that device, for example, the 2 CPU course, on this particular device I'm going to look at the one minute averages. I'm going to hit next, now you can elect to express items as a percentage. This is most common for interfaces, where max speed value derived either from the ifspeed value on the lateral switch or defined in the object manager as discussed in another video, to help show things in percentages this is unnecessary for the indicators for graphing right now. I'm going to move to the data aggregation tab and you could aggregate across a number of different options the default being average.
You have some trending options, you can elect to show percentile such as the 95th percentile as well as displaying baselines. I'm actually going to put that on and apply a column one standard deviation. You could also graph time over time showing last weeks values on the same graph as this weeks or any time span for that matter. Then you have the option of selecting work hours. I can find them on the fly by selecting custom or I can just use the system default or any of the other work hours I've all ready defined in the system. The time span is pretty straight forward, I'm actually going to do the past 7 days. You can see here we have a number of options we can toggle on and off that are relevant to a line graph. We have other graphing options here such as the bar graph, pie or stacked line diagram.
I'm actually going to select both at the same time and we'll see what that looks like in a moment. These options you can get a little bit more information on by just hovering over them or consult the user manual at any time. I'm going to hit next, I'm brought to a summary page, I can click any portion here and it'll bring me to the appropriate tab at the left here, to edit my choices if I deem that necessary. I'm simply going to hit finish and bring in the photo attachment. Now you will see dark areas that are punctuated by these duller areas here an both graphs. What that represents are the off hours or the hours that are outside of the work hours we defined as the system default.
The good thing about that is, A you can visually focus on the time frames that matter most through your devices. Back up jobs and events that occur outside of working hours are less prevalent, also the averages and peak values will correspond to only the values that occurred during work hours. The averages don't get down weighted by periods of inactivity and the peaks don't get skewed by let's say back up jobs for example. Now on the line graph at the top here, you can see the baseline tracking very closely to the CPU load for both of these cores. The shaded area represents as defined in the options a once standard deviation change from the baseline.
I can see that just barely the load this, I believe it was the past 7 days, was within one standard deviation of what's normal for this device. This is helpful for A, giving a little greater context in terms of what the baseline means as well as helping define the alert conditions that you'd like to apply to your infrastructure. You could toggle this to 2, 3, 4, 5 standard deviations and then ask yourself, is that a sensible alert condition to apply to this device? It's a visual heuristic that helps reinforce your intuition and your knowledge of your own infrastructure, but you can do a lot with these performance metrics. This is just a very quick introduction to populating reports manually through the addition of performance metrics to our report.
In addition to these videos, remember that the Data Appliance NMS User Manual is available through the NMS UI as shown below. Clicking on the question mark at the top right of any page will automatically bring you to the section of the manual that corresponds to the page you are currently on.