Viewing NUMA Refined Analysis Results

Prerequisites

A NUMA refined 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 general report of the Summary tab page is displayed by default, as shown in Figure 1. Table 1 describes the parameters.

   

   • The Summary tab page displays tuning suggestions and scores the NUMA status. The NUMA score is used to measure the NUMA memory access status of the entire system. The score ranges from 0 to 1. If the score is 1, all memory access operations are local. A score closer to 0 indicates more cross-NUMA remote access.
   • Click to locate the latest report. When all reports are generated, this icon becomes unavailable.
   • Click a time point on the timeline to view the data of a report interval. You can also click General Report to view all the data collected.
   • Move the mouse pointer to to view the tuning suggestions for the current process.
   • In the Top 10 Memory NUMA Access Processes area, click a process ID to view the process details. Click the button before the process ID to view the thread information, and click a thread ID to view the thread details.
   • In the system memory NUMA access matrix and process memory NUMA access matrix, you can move the mouse pointer over different color blocks to view detailed data of each access mode, including the memory access type, node access, traffic, and the percentage of the traffic to total system access traffic.
   Figure 1 Summary
   

   Table 1 Parameters on the Summary tab page

   Parameter	Description
   System Memory NUMA Access Matrix
   NUMA Node0	Memory access traffic from NUMA Node0 to each node and the percentage in the total access traffic.
   NUMA Node1	Memory access traffic from NUMA Node1 to each node and the percentage in the total access traffic.
   NUMA Node2	Memory access traffic from NUMA Node2 to each node and the percentage in the total access traffic.
   NUMA Node3	Memory access traffic from NUMA Node3 to each node and the percentage in the total access traffic.
   System NUMA Configuration
   NUMA Node	NUMA node name.
   %CPU	Number of occupied CPU cores. For example, 600% indicates that 6 CPU cores are occupied.
   %MEM	Memory usage.
   CPU core distribution	CPU core distribution of the NUMA node.
   Mem size	Memory size of the NUMA node.
   Mem free	Available memory size of the NUMA node.
   Node0 Distance	Distance that measures the costs to the memory of NUMA Node0.
   Node1 Distance	Distance that measures the costs to the memory of NUMA Node1.
   Node2 Distance	Distance that measures the costs to the memory of NUMA Node2.
   Node3 Distance	Distance that measures the costs to the memory of NUMA Node3.
   Top 10 Memory NUMA Access Processes
   Top by	Select a parameter to sort the top 10 information. The default value is RMA-CHIP. The options are RMA, RMA-CHIP, RMA-DIE, LMA, SCORE, MIGRATE, and CPU%.
   PID/TID	Process ID and thread ID.
   Command	Command line of a process/thread.
   RMA	Remote memory traffic (RMA-CHIP + RMA-DIE).
   RMA-CHIP	Cross-chip memory traffic.
   RMA-DIE	Cross-die memory traffic.
   LMA	Local memory traffic.
   SCORE	NUMA score of a process/thread.
   MIGRATE	Number of times that threads are migrated between NUMA nodes and number of threads in a process.
   CPU%	CPU usage.

   Figure 2 Viewing the details about a process
   

   Table 2 Parameters in the process details

   Parameter	Description
   Process Memory NUMA Access Matrix
   NUMA Node0	Memory access traffic of a process from NUMA Node0 to each node and the percentage in the total access traffic.
   NUMA Node1	Memory access traffic of a process from NUMA Node1 to each node and the percentage in the total access traffic.
   NUMA Node2	Memory access traffic of a process from NUMA Node2 to each node and the percentage in the total access traffic.
   NUMA Node3	Memory access traffic of a process from NUMA Node3 to each node and the percentage in the total access traffic.
   Process Function Memory Access
   Function	Currently called function.
   Module	Currently used driver.
   RMA	Remote memory traffic (RMA-CHIP + RMA-DIE).
   RMA-CHIP	Cross-chip memory traffic.
   RMA-DIE	Cross-die memory traffic.
   LMA	Local memory traffic.

   Figure 3 Viewing the details about a thread
   

   Table 3 Parameters in the thread details

   Parameter	Description
   Thread Memory NUMA Access Matrix
   NUMA Node0	Memory access traffic of a thread from NUMA Node0 to each node and the percentage in the total access traffic.
   NUMA Node1	Memory access traffic of a thread from NUMA Node1 to each node and the percentage in the total access traffic.
   NUMA Node2	Memory access traffic of a thread from NUMA Node2 to each node and the percentage in the total access traffic.
   NUMA Node3	Memory access traffic of a thread from NUMA Node3 to each node and the percentage in the total access traffic.
   Thread Function Memory Access
   Function	Currently called function.
   Module	Currently used driver.
   RMA	Remote memory traffic (RMA-CHIP + RMA-DIE).
   RMA-CHIP	Cross-chip memory traffic.
   RMA-DIE	Cross-die memory traffic.
   LMA	Local memory traffic.