Hello!
I am working with Windows storage clusters, and I am wanting to monitor disk I/O for the cluster object. The issue I'm running into is that if monitoring using WMI Logical Disk I/O BETA or Windows Physical Disk I/O BETA, I run into the issue of the disk disappearing on the cluster object and alerting down. I'd like to be able to keep a solid monitor on the disk that the CSV lives on regardless of which node the cluster service is currently running on. Is there a performance monitor that I am missing that can accomplish this?
Article Comments
Stephan,
Disks sensors will usually create after the initial scan and stay green for some time, but will alert down randomly. Sometimes they will come back up, sometimes they won't. They will do this even if the cluster service hasn't switched nodes. I have had physical disk monitors both on the host node and on the cluster object before. Both will be green and show the same data, but then the sensor on the cluster object will alert down, while the sensor on the host node will remain green. I need to be able to have a continuous monitor of the clustered virtual disk and volume I/O regardless of whether or not the service switches host nodes, since sometimes the virtual disk will exist on one node and sometimes another depending on the status of the cluster object.
Aug, 2016 - Permalink
Do the disks appear in the metascan of the sensor and turn red after being created? What exact error do you get there?
Aug, 2016 - Permalink