In this last lab, we will take a look at the monitoring capabilities of the ZFS Appliance.
5.1.1. View some basic statistics
If not already done, start your zfsapp1 and login to the webinterface (“https//192.168.56.100:215“) using the username root and the password oracle.
Go to Analytics
Here you can view several statistics regarding the performance and workload of your ZFS App.
Let’s start by adding a simple statistic. Click on add statistics.
In the menu, select CPU -> Percent Utilisation -> Broken down by CPU mode
This creates a graph, that lets you see the current CPU load of your system. On my system, most of that load was caused by kernel processes. If you click on one of the items on the left, it will be highlighted in the graph on the right.
Note, that the virtual appliance only has one CPU, while the real one has a lot more, so don’t be alarmed, if the graph is spiking that high without any actual load.
Lets add another graph. Select Disk -> I/O bytes -> Broken down by type of operation
As you can see in my example, there is not much going on at the moment. But you get the basic idea.
5.1.1. Navigating the statistics
You will notice, that both graphs share the same timeline. So you can very easily analyse the Appliance as you can see at first glance, for example how much disk I/O was causing how much CPU load at the time. You can also zoom in or out the graph by clicking the magnifying glasses above the graph.
If you want to see a long term statistics, you can click on the three calender icons, showing you a day, week or month of data.
The red areas represent the timeframes, where no data was collected. In my example those were times, my ZFS Appliance was not running. My ZFS Appliance was not running a lot…
On the next page, we will investigate some workload on the machine.