-
Notifications
You must be signed in to change notification settings - Fork 1.4k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
python daemons in bookworm are consuming more memory than in bullseye #19828
Comments
@saiarcot895 - since there is no option to change the speed improvement setting, there is a task to change the threshold, can you take the action item? |
@saiarcot895 Can you please provide ETA to update the test? |
There'll be a determination of what to be done here, but the people involed are out this week. There'll be an update next week. |
Hi @saiarcot895, do we have update on this issue? |
@saiarcot895 Do we have an update on this issue? |
Test case will be updated for this platform. CC @lipxu |
@dgsudharsan @vivekrnv Can you verify that the change done in sonic-net/sonic-mgmt#15948 works? |
HI @saiarcot895, Will check and verify |
Description
python daemons in bookworm is consuming more memory than in bullseye.
Some daemon like xcvrd, thermatctld, featured, hostcfgd, healthd etc has RSS increased by around ~ 20MB. Even smaller daemons has size increased by "container wait", supervisor-proc-exit-listener by 7 MB.
Combined with new daemons such stormon, dhcprelayd the overall mem usage will be higher by atleast ~500 MB. On a system with 8GB RAM this would account to almost 5-6 % more usage. Thus the test_cpu_memory_usage.py test should be updated to handle the increase.
This is happening because of speed optimization made for python 3.11
https://devclass.com/2022/05/31/how-python-3-11-is-gaining-performance-at-the-cost-of-a-bit-more-memory/
and seems this optimization has some memory implications https://peps.python.org/pep-0659/#memory-use
Steps to reproduce the issue:
Run the test_cpu_memory_usage on 202405/master images on devices with RAM <= 8 GB and it will fail
https://github.com/sonic-net/sonic-mgmt/blob/master/tests/platform_tests/test_cpu_memory_usage.py#L54
Or just load the image and check memory usage after some time
Describe the results you received:
On 202311:
On 202405:
Describe the results you expected:
No significant difference ideally, if not test should accomodate the diff
The text was updated successfully, but these errors were encountered: