Python actor throuput, New Commit #830
Open
Conversation
This file contains hidden or 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
Summary: Measures how long it takes to cast and reply a 1Kb msg to N hosts with 8 actors via local transport 100ms of processing time (included in benchmark below) was added to the tested endpoint | Benchmark Name | Python p50 Latency (ms) | Rust p50 Latency (ms) | % Diff | % Diff w/o proc time| |-------------------------|-------------------------------|-------------------------|---------------------------------|-----------------------| | actor_count_1_median_ms | 102 | 101 | .01% | 516% | | actor_count_10_median_ms | 108 | 101 | .69% | 674% | | actor_count_100_median_ms| 145 | 102 | 40.81% | 1510% | * Crashes on 1k "host" set up Differential Revision: D79769681
Summary: For an unknown reason, small payloads consistently perform better in Python compared to Rust (D79577855), eventhough the messaging mechanism is powered by port handles and receivers on both implementations. Once payload size increases, Rust outperforms Python (as expected) Additionally, Python cannot reliable cast 1Gb of data during the bechmakr without throwing a bunch of errors, so it is excluded from benchmark for now. | Benchmark | Time [Min, Median, Max] (ms) | Python Throughput [MiB/s] | Rust Throughput [MiB/s] | Throughput Change % | |---------------------|------------------------------|--------------------------|---------------------------|-------------------------------| | hosts/1/size/10kb | [0, 0, 4] | 17 | 8.4220 | +101.8% | | hosts/1/size/100kb | [0, 0, 4] | 140 | 79.577 | +75.9% | | hosts/1/size/1mb | [1, 2, 9] | 433 | 535.84 | -19.2% | | hosts/1/size/10mb | [18, 20, 33] | 466 | 494.50 | -5.8% | | hosts/1/size/100mb | [202, 223, 318] | 443 | 518.56 | -14.6% | | hosts/10/size/10kb | [1, 1, 59] | 67 | 75.318 | -11.1% | | hosts/10/size/100kb | [1, 1, 59] | 537 | 678.38 | -20.9% | | hosts/10/size/1mb | [4, 6, 70] | 1451 | 2843.14 | -49.0% | | hosts/10/size/10mb | [52, 61, 145] | 1575 | 2243.11 | -29.8% | | hosts/10/size/100mb | [677, 720, 905] | 1353 | 2152.97 | -37.1% | Differential Revision: D80100828
meta-cla
bot
added
the
CLA Signed
|
This pull request was exported from Phabricator. Differential Revision: D80100828 |
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.
Summary:
For an unknown reason, small payloads consistently perform better in Python compared to Rust (D79577855), eventhough the messaging mechanism is powered by port handles and receivers on both implementations.
Once payload size increases, Rust outperforms Python (as expected)
Additionally, Python cannot reliable cast 1Gb of data during the bechmakr without throwing a bunch of errors, so it is excluded from benchmark for now.
Differential Revision: D80100828