Isilon InsightIQ User Guide

You can also break out this information by node, protocol, client, or operation class. The Protocol Operations Average Latency report can be a helpful report, especially when you break out the details. For example, view the latency by protocol, then at the operation class for each protocol, to check for specific operations that might be causing excessive latency.

For the SMB protocol, the Change Notify operation can be the source of high latency. You might also see that namespace operations introduce significant latency, which can be alleviated with SSDs. HTTP transactions also exhibit high latency and skew the results. Filter out HTTP traffic, unless you are specifically interested in that data.