Speed up --gil by skipping work for non-GIL threads #630
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
0eed74d to
e81b3e4
Compare
When the user has specified --gil we needn't collect stack traces for the
non-GIL threads. This is a significant speedup on heavily threaded Python
code.
My server has 271 threads (most of which are idle).
With the previous code (sampling at 10Hz) I get:
$ ./py-spy record -o /var/opt/stbt/profile.svg --pid $(pgrep -f http_service.service) -r 10 -d 10 --gil
py-spy> Sampling process 10 times a second for 10 seconds. Press Control-C to exit.
py-spy> 1.29s behind in sampling, results may be inaccurate. Try reducing the sampling rate
py-spy> 1.36s behind in sampling, results may be inaccurate. Try reducing the sampling rate
py-spy> 1.32s behind in sampling, results may be inaccurate. Try reducing the sampling rate
py-spy> 1.22s behind in sampling, results may be inaccurate. Try reducing the sampling rate
py-spy> 1.00s behind in sampling, results may be inaccurate. Try reducing the sampling rate
py-spy> Wrote flamegraph data to '/var/opt/stbt/profile.svg'. Samples: 3 Errors: 51
With this commit I can sample at 80Hz, and the error rate is way lower:
$ ./py-spy2 record -o /var/opt/stbt/profile.svg --pid $(pgrep -f http_service.service) -r 80 -d 10 --gil
py-spy> Sampling process 80 times a second for 10 seconds. Press Control-C to exit.
py-spy> Wrote flamegraph data to '/var/opt/stbt/profile.svg'. Samples: 65 Errors: 0
I suspect the difference in error rate is due to the error handling in
`_get_stack_traces` - an error getting information about any thread means
that all the traces are thrown away.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
When the user has specified --gil we needn't collect stack traces for the non-GIL threads. This is a significant speedup on heavily threaded Python code.
My server has 271 threads (most of which are idle).
With the previous code (sampling at 10Hz) I get:
With this commit I can sample at 80Hz, and the error rate is way lower:
I suspect the difference in error rate is due to the error handling in
_get_stack_traces- an error getting information about any thread means that all the traces are thrown away.