Performance Optimization: Difference between revisions

From HPCWIKI
Jump to navigation Jump to search
No edit summary
No edit summary
Line 1: Line 1:
{| class="wikitable"
{| class="wikitable"
|+
|+
!typical symptom
!Typical symptom
!Possible  reason
!Possible  reason
!Solutions
!Solutions
|-
|-
|sharply rising system CPU utilization.
|sharply rising system CPU utilization. In addition, the page fault exception handler is <code>do_huge_pmd_anonymous_page</code>. This means that the current system doesn’t have 2 MB of contiguous physical memory and that triggers the direct memory compaction. The direct memory compaction is time-consuming, so it leads to high system CPU utilization
|THP-caused performance fluctuation is  
|THP-caused performance fluctuation.
The fragmentation index is associated with the <code>/proc/sys/vm/extfrag_threshold</code> parameter. The closer the number is to 1,000, the more the allocation failure is related to memory fragmentation.
 
To verify whether the performance fluctuation is related to memory fragmentation:
|Disable [[THP (Transparent Huge Pages)]]
|Disable [[THP (Transparent Huge Pages)]]
|-
|-

Revision as of 17:07, 30 March 2023

Typical symptom Possible reason Solutions
sharply rising system CPU utilization. In addition, the page fault exception handler is do_huge_pmd_anonymous_page. This means that the current system doesn’t have 2 MB of contiguous physical memory and that triggers the direct memory compaction. The direct memory compaction is time-consuming, so it leads to high system CPU utilization THP-caused performance fluctuation.

The fragmentation index is associated with the /proc/sys/vm/extfrag_threshold parameter. The closer the number is to 1,000, the more the allocation failure is related to memory fragmentation.

To verify whether the performance fluctuation is related to memory fragmentation:

Disable THP (Transparent Huge Pages)

Reference