Viewing Analysis Results

Prerequisites

A process/thread 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 the following tables.
     

     • The Tuning Suggestions area is displayed only when the system detects metrics to be tuned. You can expand the area to view detailed suggestions and solutions. Tuning suggestions are provided for certain configuration items and marked with . Hover your mouse pointer to view details about a tuning suggestion.
     • If Analysis Object is set to Application and System Call Tracing is enabled during the creation of a process/thread analysis task, the analysis results contain the System Call area. For details about the parameters, see Table 5.
     • When a process/thread analysis task is created, all GPU collection types are selected by default. If no related operation is performed in the program, no data is displayed in the related area on the report page.
     Figure 1 Summary
     

     Table 1 Parameters in the CPU area

     Parameter	Description
     PID/TID	Process or thread ID.
     %user	CPU usage of a task in the user space.
     %system	CPU usage of a task in the kernel space.
     %wait	CPU usage of a task in the I/O waiting state.
     %CPU	CPU usage of a task.
     Command	Command corresponding to the current task.

     Table 2 Parameters in the Memory area

     Parameter	Description
     PID/TID	Process or thread ID.
     minflt/s	Number of secondary page faults per second, that is, number of page faults generated when a virtual memory address is mapped to a physical memory address. No page needs to be loaded from the drive.
     majflt/s	Number of major page faults per second. When a virtual memory address is mapped to a physical memory address, the corresponding page is in the swap memory. Such page faults are major page faults, which are generated when the memory is insufficient. Pages need to be loaded from the drive.
     VSZ	Size of the virtual memory used by a task, in KB.
     RSS	Resident set size (RSS), that is, the size of the physical memory used by a task, in KB.
     %MEM	Memory usage of a task.
     Command	Command corresponding to the current task.

     Table 3 Parameters in the Storage I/O area

     Parameter	Description
     PID/TID	Process or thread ID.
     kB_rd/s	Volume of data (in KB) read by a task from a drive per second.
     kB_wr/s	Volume of data (in KB) written by a task to a drive per second.
     iodelay	I/O delay (in clock cycles), including the I/O waiting time and I/O insertion end time.
     Command	Command corresponding to the current task.

     Table 4 Parameters in the Context Switch area

     Parameter	Description
     PID/TID	Process or thread ID.
     cswch/s	Number of proactive task context switchovers per second. Generally, a proactive task context switchover occurs because the task cannot obtain required resources. For example, when system resources such as I/O and memory resources are insufficient, a proactive task context switchover occurs.
     nvcswch/s	Number of passive task context switchovers per second. Generally, a task is forcibly scheduled by the system because the time slice is due or the process is preempted by a process with a higher priority. As a result, a passive task context switchover occurs. For example, when a large number of processes are contending for CPU, a passive task context switchover is likely to occur.
     Command	Command corresponding to the current task.

     Table 5 Parameters in the System Call area

     Parameter	Description
     PID/TID	Process or thread ID.
     %time	Percentage of the system CPU time.
     seconds/s	Total system CPU time, in seconds.
     usecs/call (ms)	Average system CPU time for each call, in milliseconds.
     calls	Number of system calls in the sampling process.
     errors	Number of system calling failures in the sampling process.
     syscall	System call name.

   • Click the CPU tab page. The CPU analysis result of the process or thread is displayed, as shown in Figure 2. For details about the parameters, see Table 6.
     

     • The CPU tab page displays the collected time series data of CPU performance metrics in line charts.
     • By default, the analysis results of certain processes or threads are displayed in a line chart. When you hover the mouse pointer over a process or thread, detailed parameter information is displayed. You can click the process or thread to determine whether to display it in the current view. If there are too many processes or threads, the page is automatically collapsed.
     • If the lines overlap, you can click the color icon corresponding to the process or thread to filter the lines. You can hover the mouse pointer over a process or thread ID to view detailed parameter information.
     • Click to go to the process filtering page. By default, processes and threads are sorted by attribute %CPU. The first 10 processes and threads are selected by default. You can click a process in the filter box to view the threads of the process.
     Figure 2 CPU tab page
     

     Table 6 Parameters on the CPU tab page

     Parameter	Description
     %user	CPU usage of a process in the user space.
     %system	CPU usage of a process in the kernel space.
     %wait	CPU usage of a process in the I/O waiting state.
     %CPU	CPU usage of a process.
     CPU ID	ID of the CPU core that handles a process.

   • Click the Memory tab page. The memory analysis result of the process or thread is displayed, as shown in Figure 3. For details about the parameters, see Table 2.
     

     • The Memory tab page displays the collected time series data of memory performance metrics in line charts.
     • By default, the analysis results of certain processes or threads are displayed in a line chart. When you hover the mouse pointer over a process or thread, detailed parameter information is displayed. You can click the process or thread to determine whether to display it in the current view. If there are too many processes or threads, the page is automatically collapsed.
     • You can click and select a process or thread to view the line chart of the corresponding analysis result. By default, the processes and threads are sorted by attribute %MEM in descending order. By default, the first 10 processes and threads are selected. You can click a process in the filter box to view the threads of the process.
     • If the lines overlap, you can click the color icon corresponding to the process or thread to filter the lines. You can hover the mouse pointer over a process or thread ID to view detailed parameter information.
     Figure 3 Memory tab page
     
   • Click the Storage I/O tab page. The storage I/O analysis result of the process or thread is displayed, as shown in Figure 4. For details about the parameters, see Table 3.
     

     • The Storage I/O tab page displays the collected time series data of storage I/O performance metrics in line charts.
     • By default, the analysis results of certain processes or threads are displayed in a line chart. When you hover the mouse pointer over a process or thread, detailed parameter information is displayed. You can click the process or thread to determine whether to display it in the current view. If there are too many processes or threads, the page is automatically collapsed.
     • You can click and select a process or thread to view the line chart of the corresponding analysis result. By default, the processes and threads are sorted by attribute kB_wr/s in descending order. By default, the first 10 processes and threads are selected. You can click a process in the filter box to view the threads of the process.
     • If the lines overlap, you can click the color icon corresponding to the process or thread to filter the lines. You can hover the mouse pointer over a process or thread ID to view detailed parameter information.
     Figure 4 Storage I/O tab page
     
   • Click the Context Switch tab page. The context switch analysis result of the process or thread is displayed, as shown in Figure 5. For details about the parameters, see Table 4.
     

     • The Context Switch tab page displays the collected time series data of context switching performance metrics in line charts.
     • By default, the analysis results of certain processes or threads are displayed in a line chart. When you hover the mouse pointer over a process or thread, detailed parameter information is displayed. You can click the process or thread to determine whether to display it in the current view. If there are too many processes or threads, the page is automatically collapsed.
     • You can click and select a process or thread to view the line chart of the corresponding analysis result. By default, the processes and threads are sorted by attribute cswch/s in descending order. By default, the first 10 processes and threads are selected. You can click in front of a process in the filter box to view the threads of the process.
     • If the lines overlap, you can click the color icon corresponding to the process or thread to filter the lines. You can hover the mouse pointer over a process or thread ID to view detailed parameter information.
     Figure 5 Context Switch tab page
     
   • 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."