I have a server that's been jumping around 100-700
%DiskTime for the past few days.
I understand this indicates an I/O bottleneck -- but what do I do from here? What do I need to check or change in regards to the disk?
I have a server that's been jumping around 100-700
%DiskTime for the past few days.
I understand this indicates an I/O bottleneck -- but what do I do from here? What do I need to check or change in regards to the disk?
Use process explorer (procexp) to track down the hog. You'll need to add additional columns in the view pane specifically for disk reads and disk writes. Once you have the culprit you can go from there (engage the vendor or developer).
You'll likely want to see what that process is doing and you can use process monitor (procmon) to do that. When you launch procmon, you'll likely want to filter out everything but the process you're investigating.
Those two tools come with their own help documentation (and it's pretty good). To deep dive even further, I'd recommend the windows internals books from Mark Russinovich.
Process Explorer
Process Monitor
Sysinternals Process Utilities
Windows Internals Books