Viewing I/O Analysis Results

Prerequisites

An I/O analysis task has been created and the analysis is complete.

Procedure

1. In the System Profiler area on the left, click the name of the target analysis task.

   The node list is displayed.

2. Click the name of the target node to view the analysis results.
   • Click the node name. The Summary tab page is displayed by default, as shown in Figure 1. For details about the parameters, see Table 1.
     

     • The Summary tab page displays details about block I/O trace statistics, drive I/O operations, and I/O API calls in the entire sampling period.
     • You can click to filter drives and view corresponding information.
     • In the Block I/O Trace Statistics area, click to view data drive details. You can switch the display mode in the upper right corner of the details page. The statistics can be displayed in charts or tables.
     • You can move the mouse pointer to the graph to view the data flow details.
     Figure 1 Summary
     

     Table 1 Parameters on the Summary tab page

     Parameter	Description
     Block I/O Trace Statistics
     DEV	Block device name.
     Q Requests	Sum of the number of requests detected by the block_bio_queue and block_split tracepoints.
     D Requests	Number of requests detected by the block_rq_issue tracepoint.
     X Requests	Number of requests detected by the block_split tracepoint.
     M Requests	Number of requests detected by the block_bio_backmerge and block_bio_frontmerge tracepoints, indicating the number of requests for which merge occurs.
     Merge ratio	Ratio of the number of Q requests to the number of D requests.
     BLKmin	Minimum requested sector size detected by the block_rq_complete tracepoint.
     BLKavg	Average requested sector size detected by the block_rq_complete tracepoint.
     BLKmax	Maximum requested sector size detected by the block_rq_complete tracepoint.
     BLKtotal	Sector size of all requests detected by the block_rq_complete tracepoint.
     Q2Q Discontinuous Requests	Two consecutive requests collected by the block_bio_queue tracepoint. If the sectors are discontinuous, the count is accumulated.
     Q2Q Request Average Distance	Q2Q request average distance = Distance between all requests collected by the block_bio_queue tracepoint/Number of Q2Q discontinuous requests
     D2D Discontinuous Requests	Two consecutive requests collected by the block_rq_issue tracepoint. If the sectors are discontinuous, the count is accumulated.
     D2D Request Average Distance	D2D request average distance = Distance between all requests collected by the block_rq_issue tracepoint/Number of D2D discontinuous requests
     Live Period Count	Count of active I/O periods. If requests from D to C exist in a time period, the time period may be used as an active time segment.
     AvgAct	Average active period of I/O activities, that is, the accumulated active time divided by the number of active time segments.
     Avg!Act	Average inactive period, that is, (Total time – Accumulated active time)/Number of inactive time segments.
     Live Percent	Percentage of active time, that is, Accumulated active time/Total time.
     Drive I/O
     DEV	Block device name.
     Number of Read Operations	Number of read operations on the drive.
     Number of Write Operations	Number of write operations on the drive.
     Read IOPS	Number of read operations per second on the drive.
     Write IOPS	Number of write operations per second on the drive.
     Read Data Size (MiB)	Size of data read from the drive.
     Write Data Size (MiB)	Size of data written to the drive.
     Read Throughput (MiB/s)	Read throughput of the drive I/O.
     Write Throughput (MiB/s)	Write throughput of the drive I/O.
     Operation	Click View Details to view drive I/O details.
     I/O APIs
     PID	Process ID.
     Process	Process name.
     Function Name	Function name.
     Call Times	Number of times that the I/O API is called.
     Average Execution Time (ms)	Average call time of the I/O API.
     Total Execution Time (ms)	Total call time of the I/O API.
     Execution Time Ratio (%)	Percentage of the total call time of an API in the total call time of all APIs in a process.
     Operation	Click View Details to view I/O API calling details.

   • Click the I/O APIs tab page to view I/O API calling information, as shown in Figure 2. For details about the parameters, see Table 2.
     

     • Click and select a time period to view the detailed information within the period.
     • You can click on the right to filter processes or functions and view the corresponding line chart.
     Figure 2 I/O APIs tab page
     

     Table 2 Parameters on the I/O APIs tab page

     Parameter	Description
     Number of Calls	Number of times that the I/O API is called.
     Average Execution Time (s)	Average call time of the I/O API.
     Total Execution Time (s)	Total call time of the I/O API.
     Time (s)	Sampling duration.
     Subprocess ID/PID	Subprocess or process ID.
     Process	Process name.
     Parameter List	List of parameters.
     Return Value	Return value of the parameter.
     Execution Time (ms)	Call time of the I/O API.

   • Click the Drive I/O tab page to view drive I/O information. See Figure 3.
     

     • Click and select a time period to view the detailed information within the period.
     • You can click on the right to filter performance metrics to view the corresponding line charts.
     Figure 3 Drive I/O tab page
     

     Table 3 Parameters on the Drive I/O tab page

     Parameter	Description
     Time (s)	Time taken by the drive I/O operation.
     DEV	Name of the current device.
     Operation Type	Name of the current operation.
     I2D Delay	How long it takes from the time a request is inserted into the device's queue to the time it is actually issued to the device.
     D2C delay	How long it takes from the time a request is issued to the time it is actually completed.
     Start Block No.	Start block number of the drive I/O operation.
     End Block No.	End block number of the drive I/O operation.
     Data Size (MiB)	Size of drive I/O data.
     Throughput (MiB/s)	Drive I/O throughput.
     Events	Event name.
     Number of blocks	Number of blocks involved in the current operation.
     Core	Currently used CPU core.
     PID	ID of the current process.
     Process Name	Name of the current process.
     Callstack	Call stack involved in the current operation.

   • Click the Task Information tab to view the detailed configuration and sampling information about the task on the current node.
     

     If the task fails to be executed, the failure cause is displayed on the Task Information tab page.

     If part of data fails to be collected but the overall task execution is not affected, you can view the exception message in Exception Information.

     Collection End Cause displays the reason why the data collection of the current task ends, for example, "Task collection times up" or "File size reaches the collection limit."