-
Notifications
You must be signed in to change notification settings - Fork 628
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
Performance hit after upgrade to 3.2.1 #601
Comments
Can you check if this is related to the scheduler or one of the other processes? |
How can I check that? |
I think this may be because it now calls I have 3 options for possible improvements. The pros: The cons: |
Option 4: boolean to enable / disable telemetry? For option 1, if you can give me an example of |
|
Over 50,000 iterations, no significant difference... ➜ time ./test1.sh
./test1.sh 83.36s user 146.00s system 78% cpu 4:53.98 total
➜ time ./test2.sh
./test2.sh 80.47s user 152.67s system 78% cpu 4:56.86 total |
Hmm less then 4% increased performance dosent look promising for option 1. There would also be a reduction in time spent parsing the result, since scaling would only need the number of lines. But I wouldn't expect this to much of a performance issue either. Currently this only appears to cost 0.16% cpu at the tick rate on the given system so perhaps more then just ps is to blame here. |
The original PR that caused the performance hit was reverted. |
Description:
I noticed a performance hit in production after upgrading to Horizon 3.2.1, which is probably in relation to #589. This is on a EC2 t2 instance, with constant workload, the hit is an additional 5-6% of CPU usage.
See graph below, we upgraded to 3.2.1 on May 22 (you will notice also that the CPU credits went down to zero way faster after the upgrade):
And this graph is after reverting to 3.1.2:
Steps To Reproduce:
Upgrade Horizon from 3.1.2 to 3.2.1.
The text was updated successfully, but these errors were encountered: