Analyzing an I/O trace log
The Management Server console lets you analyze the I/O trace log that you collected with SmartAssist. The tool helps you to determine the optimal SmartIO cache size for the volume, mount point, disk, or disk group where you collected the I/O trace log. You can also view related latency gain and performance forecast.
If you plan to allocate a custom cache size, SmartAssist also helps you determine the latency gain and performance forecast for that cache size.
To perform this task, your user group must be assigned the Admin role on the host or the Server perspective. The permission on the host may be explicitly assigned or inherited from a parent Organization.
To analyze an I/O trace log
-
In the Management Server console, go to the Server perspective and expand Manage in the left pane.
-
Expand Organization orUncategorized Hosts to locate and select the host.
-
Right-click the host, and select SmartIO>Configuration>SmartAssist.
-
In the SmartAssist tab, right-click the required I/O Trace Log, and click Analyze.
-
The Analyze I/O Trace Log panel, in the lower pane, SmartAssist by default displays the analysis results for the optimal SmartIO cache size, as follows:
-
Total reads: The number of read-hits in the trace log that the optimal cache would serve.
-
Cache size: Optimal cache size is displayed by default. If you run the analysis for a custom cache size, that size is displayed in this column
-
Read-hit ratio: The percentage of reads that the cache would serve.
-
Average latency gain: Response time in milliseconds that the cache would save.
-
Speed of reads: The number of the times by which the cache would increase the speed of reads.
-
Analysis method: Indicates if recommended or custom cache is used for analysis.
-
-
Click View Devices to view the list of data objects for which you created the I/O Trace Log.
-
To view the same parameters for a different cache size, enter the details in the SmartAssist Analysis Options pane, and clickEstimate.
Related information