Disk usage of all the logical disks available
Hi,
I am using a windows performance monitor to capture basic performance counter of some of the servers that dont have agents on them. I need to fetch the disk utilization % of all the instances of the logical disks on each server. By default the 'instanceName' under the Key Properties is set to '_Total', which will give the overall usage and I am able to pull usage of a specific disk for ex. using instanceName as 'C:' for C: drive.
But when i use the instanceName as '<All instances>' (which is exactly the same instance name in perfmon to get usage of all the logical disks), it returns me null, probably as it is not coded to get more than a value . Can some one please suggest a way to get disk usage of all the logical drives?
Thanks,
Sandeep.
Hi
If you add the counter in PerfMon it seems it actually uses the "_Total" instanceName. At least when I test this on my machine. I guess the intention of <All Instances> is that the perfmon tool itself is dynamically evaluating the avialable disks and then querying the values for each disk. Our Windows Performance Monitor doesnt have that capability built-in to first evaluate the dynamic list and then querying the data for each instance. What you should look at is our Host Monitoring Agent. Simply install a Host Monitor Agent on those machines where you dont have a Java, .NET, Web Server ... agent. This agent will then automatically capture these values for your disks, NICs, CPU and Memory
Andi
Thanks Andi !
I have almost a 100 servers that I need this information and we may not be able to get host agents on all of them. As a work around I have created measures dives C: to I: (which i thought might be the farthest we go in drives) and running the monitor. If the drive isn't available on the server, it just returns me a null value, which will anyways be ignored while generating alerts.
Sandeep.
Keep me posted on how this works for you. also - if you can do me a favor: can you let me know which measures you are interested in? Our team is extending our host monitoring capabilities and is always looking for measures that we should capture out-of-the-box
