ROS Index
|
buildfarm_perf_tests package from buildfarm_perf_tests repobuildfarm_perf_tests |
ROS Distro
|
Package Summary
| Tags | No category tags. |
| Version | 0.0.0 |
| License | Apache License 2.0 |
| Build type | AMENT_CMAKE |
| Use | RECOMMENDED |
Repository Summary
| Description | Performance tests which run regularly on the buildfarm |
| Checkout URI | https://github.com/ros2/buildfarm_perf_tests.git |
| VCS Type | git |
| VCS Version | master |
| Last Updated | 2024-02-16 |
| Dev Status | UNKNOWN |
| Released | UNRELEASED |
| Tags | No category tags. |
| Contributing |
Help Wanted (-)
Good First Issues (-) Pull Requests to Review (-) |
Package Description
Additional Links
Maintainers
- Scott K Logan
Authors
buildfarm perf tests
Purpose
This package defines some tests.
On one hand it invokes perf_test from Apex.AI’s performance_test package.
This allows you to test performance and latency of several ROS 2 RMW implementations.
On the other hand we are evaluating the additional overhead caused by a single pub/sub topic or one process spinning and detect potential leaks related to theses activities.
-
There is a test for each RMW:
- CycloneDDS (standalone test)
- FastRTPS (standalone test)
- rmw_connext_cpp
- rmw_cyclonedds_cpp
- rmw_fastrtps_cpp
- rmw_fastrtps_dynamic_cpp
- rmw_opensplice_cpp
Test 1 - Performance Test (Apex.AI)
In this test we are running the Performance Test provided by Apex.AI. Right now we have our own fork because there are some pending pull requests in the official gitlab repository.
In this test we are measurement:
- Average latency
- CPU usage (provided by Apex.AI tool)
- Sent/Received messages per second
- Total lost messages
- Max resident set size
We are generating two plots per measurement
The test only measures the latency between the same RMW implementation
| Publisher/Subscriber | rmw_fastrtps_cpp | rmw_opensplice_cpp | rmw_cyclonedds_cpp | rmw_fastrtps_dynamic_cpp | rmw_connext_cpp |
|---|---|---|---|---|---|
| rmw_fastrtps_cpp | :heavy_check_mark: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: |
| rmw_opensplice_cpp | :heavy_multiplication_x: | :heavy_check_mark: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: |
| rmw_cyclonedds_cpp | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_check_mark: | :heavy_multiplication_x: | :heavy_multiplication_x: |
| rmw_fastrtps_dynamic_cpp | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_check_mark: | :heavy_multiplication_x: |
| rmw_connext_cpp | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_check_mark: |
Test 2 - Simple pub/sub
In this case we are testing one publisher and one subscriber in different processes sending a 1kArray at 5Hz. This will allow us to evaluate additional overhead caused by a single pub/sub topic and detect leaks related to this activity.
We measure for both publisher and subscriber:
- Average round trip
- CPU usage ( readed from the filesystem )
- Total lost messages
- Received/Sent messages per second
- Physical memory
- Resident anonymous memory
- Virtual memory
Again we plot measurement:
| Publisher/Subscriber | rmw_fastrtps_cpp | rmw_opensplice_cpp | rmw_cyclonedds_cpp | rmw_fastrtps_dynamic_cpp | rmw_connext_cpp |
|---|---|---|---|---|---|
| rmw_fastrtps_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| rmw_opensplice_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| rmw_cyclonedds_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| rmw_fastrtps_dynamic_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| rmw_connext_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
Test 3 - Node spinning
This test creates one process and spins for 1 minute to evaluate ROS 2 overhead and detect obvious leaks.
We measure:
- CPU usage ( readed from file system )
- Physical memory
- Resident anonymous memory
- Virtual memory
Again we plot measurement:
| DDS | Process 1 |
|---|---|
| rmw_fastrtps_cpp | :heavy_check_mark: |
| rmw_opensplice_cpp | :heavy_check_mark: |
| rmw_cyclonedds_cpp | :heavy_check_mark: |
| rmw_fastrtps_dynamic_cpp | :heavy_check_mark: |
| rmw_connext_cpp | :heavy_check_mark: |
Build
- Install ROS 2 (https://index.ros.org/doc/ros2/Installation/).
File truncated at 100 lines see the full file
Package Dependencies
| Deps | Name |
|---|---|
| ament_cmake | |
| ament_cmake_python | |
| ament_index_python | |
| launch | |
| launch_ros | |
| ament_lint_auto | |
| ament_lint_common | |
| launch_testing_ament_cmake | |
| performance_test | |
| rmw_implementation_cmake | |
| rclcpp |
System Dependencies
Dependant Packages
Launch files
Messages
Services
Plugins
Recent questions tagged buildfarm_perf_tests at Robotics Stack Exchange
|
|
buildfarm_perf_tests package from buildfarm_perf_tests repobuildfarm_perf_tests |
ROS Distro
|
Package Summary
| Tags | No category tags. |
| Version | 0.0.0 |
| License | Apache License 2.0 |
| Build type | AMENT_CMAKE |
| Use | RECOMMENDED |
Repository Summary
| Description | Performance tests which run regularly on the buildfarm |
| Checkout URI | https://github.com/ros2/buildfarm_perf_tests.git |
| VCS Type | git |
| VCS Version | master |
| Last Updated | 2024-02-16 |
| Dev Status | UNKNOWN |
| Released | UNRELEASED |
| Tags | No category tags. |
| Contributing |
Help Wanted (-)
Good First Issues (-) Pull Requests to Review (-) |
Package Description
Additional Links
Maintainers
- Scott K Logan
Authors
buildfarm perf tests
Purpose
This package defines some tests.
On one hand it invokes perf_test from Apex.AI’s performance_test package.
This allows you to test performance and latency of several ROS 2 RMW implementations.
On the other hand we are evaluating the additional overhead caused by a single pub/sub topic or one process spinning and detect potential leaks related to theses activities.
-
There is a test for each RMW:
- CycloneDDS (standalone test)
- FastRTPS (standalone test)
- rmw_connext_cpp
- rmw_cyclonedds_cpp
- rmw_fastrtps_cpp
- rmw_fastrtps_dynamic_cpp
- rmw_opensplice_cpp
Test 1 - Performance Test (Apex.AI)
In this test we are running the Performance Test provided by Apex.AI. Right now we have our own fork because there are some pending pull requests in the official gitlab repository.
In this test we are measurement:
- Average latency
- CPU usage (provided by Apex.AI tool)
- Sent/Received messages per second
- Total lost messages
- Max resident set size
We are generating two plots per measurement
The test only measures the latency between the same RMW implementation
| Publisher/Subscriber | rmw_fastrtps_cpp | rmw_opensplice_cpp | rmw_cyclonedds_cpp | rmw_fastrtps_dynamic_cpp | rmw_connext_cpp |
|---|---|---|---|---|---|
| rmw_fastrtps_cpp | :heavy_check_mark: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: |
| rmw_opensplice_cpp | :heavy_multiplication_x: | :heavy_check_mark: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: |
| rmw_cyclonedds_cpp | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_check_mark: | :heavy_multiplication_x: | :heavy_multiplication_x: |
| rmw_fastrtps_dynamic_cpp | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_check_mark: | :heavy_multiplication_x: |
| rmw_connext_cpp | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_check_mark: |
Test 2 - Simple pub/sub
In this case we are testing one publisher and one subscriber in different processes sending a 1kArray at 5Hz. This will allow us to evaluate additional overhead caused by a single pub/sub topic and detect leaks related to this activity.
We measure for both publisher and subscriber:
- Average round trip
- CPU usage ( readed from the filesystem )
- Total lost messages
- Received/Sent messages per second
- Physical memory
- Resident anonymous memory
- Virtual memory
Again we plot measurement:
| Publisher/Subscriber | rmw_fastrtps_cpp | rmw_opensplice_cpp | rmw_cyclonedds_cpp | rmw_fastrtps_dynamic_cpp | rmw_connext_cpp |
|---|---|---|---|---|---|
| rmw_fastrtps_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| rmw_opensplice_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| rmw_cyclonedds_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| rmw_fastrtps_dynamic_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| rmw_connext_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
Test 3 - Node spinning
This test creates one process and spins for 1 minute to evaluate ROS 2 overhead and detect obvious leaks.
We measure:
- CPU usage ( readed from file system )
- Physical memory
- Resident anonymous memory
- Virtual memory
Again we plot measurement:
| DDS | Process 1 |
|---|---|
| rmw_fastrtps_cpp | :heavy_check_mark: |
| rmw_opensplice_cpp | :heavy_check_mark: |
| rmw_cyclonedds_cpp | :heavy_check_mark: |
| rmw_fastrtps_dynamic_cpp | :heavy_check_mark: |
| rmw_connext_cpp | :heavy_check_mark: |
Build
- Install ROS 2 (https://index.ros.org/doc/ros2/Installation/).
File truncated at 100 lines see the full file
Package Dependencies
| Deps | Name |
|---|---|
| ament_cmake | |
| ament_cmake_python | |
| ament_index_python | |
| launch | |
| launch_ros | |
| ament_lint_auto | |
| ament_lint_common | |
| launch_testing_ament_cmake | |
| performance_test | |
| rmw_implementation_cmake | |
| rclcpp |
System Dependencies
Dependant Packages
Launch files
Messages
Services
Plugins
Recent questions tagged buildfarm_perf_tests at Robotics Stack Exchange
|
|
buildfarm_perf_tests package from buildfarm_perf_tests repobuildfarm_perf_tests |
ROS Distro
|
Package Summary
| Tags | No category tags. |
| Version | 0.0.0 |
| License | Apache License 2.0 |
| Build type | AMENT_CMAKE |
| Use | RECOMMENDED |
Repository Summary
| Description | Performance tests which run regularly on the buildfarm |
| Checkout URI | https://github.com/ros2/buildfarm_perf_tests.git |
| VCS Type | git |
| VCS Version | master |
| Last Updated | 2024-02-16 |
| Dev Status | UNKNOWN |
| Released | UNRELEASED |
| Tags | No category tags. |
| Contributing |
Help Wanted (-)
Good First Issues (-) Pull Requests to Review (-) |
Package Description
Additional Links
Maintainers
- Scott K Logan
Authors
buildfarm perf tests
Purpose
This package defines some tests.
On one hand it invokes perf_test from Apex.AI’s performance_test package.
This allows you to test performance and latency of several ROS 2 RMW implementations.
On the other hand we are evaluating the additional overhead caused by a single pub/sub topic or one process spinning and detect potential leaks related to theses activities.
-
There is a test for each RMW:
- CycloneDDS (standalone test)
- FastRTPS (standalone test)
- rmw_connext_cpp
- rmw_cyclonedds_cpp
- rmw_fastrtps_cpp
- rmw_fastrtps_dynamic_cpp
- rmw_opensplice_cpp
Test 1 - Performance Test (Apex.AI)
In this test we are running the Performance Test provided by Apex.AI. Right now we have our own fork because there are some pending pull requests in the official gitlab repository.
In this test we are measurement:
- Average latency
- CPU usage (provided by Apex.AI tool)
- Sent/Received messages per second
- Total lost messages
- Max resident set size
We are generating two plots per measurement
The test only measures the latency between the same RMW implementation
| Publisher/Subscriber | rmw_fastrtps_cpp | rmw_opensplice_cpp | rmw_cyclonedds_cpp | rmw_fastrtps_dynamic_cpp | rmw_connext_cpp |
|---|---|---|---|---|---|
| rmw_fastrtps_cpp | :heavy_check_mark: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: |
| rmw_opensplice_cpp | :heavy_multiplication_x: | :heavy_check_mark: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: |
| rmw_cyclonedds_cpp | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_check_mark: | :heavy_multiplication_x: | :heavy_multiplication_x: |
| rmw_fastrtps_dynamic_cpp | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_check_mark: | :heavy_multiplication_x: |
| rmw_connext_cpp | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_check_mark: |
Test 2 - Simple pub/sub
In this case we are testing one publisher and one subscriber in different processes sending a 1kArray at 5Hz. This will allow us to evaluate additional overhead caused by a single pub/sub topic and detect leaks related to this activity.
We measure for both publisher and subscriber:
- Average round trip
- CPU usage ( readed from the filesystem )
- Total lost messages
- Received/Sent messages per second
- Physical memory
- Resident anonymous memory
- Virtual memory
Again we plot measurement:
| Publisher/Subscriber | rmw_fastrtps_cpp | rmw_opensplice_cpp | rmw_cyclonedds_cpp | rmw_fastrtps_dynamic_cpp | rmw_connext_cpp |
|---|---|---|---|---|---|
| rmw_fastrtps_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| rmw_opensplice_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| rmw_cyclonedds_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| rmw_fastrtps_dynamic_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| rmw_connext_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
Test 3 - Node spinning
This test creates one process and spins for 1 minute to evaluate ROS 2 overhead and detect obvious leaks.
We measure:
- CPU usage ( readed from file system )
- Physical memory
- Resident anonymous memory
- Virtual memory
Again we plot measurement:
| DDS | Process 1 |
|---|---|
| rmw_fastrtps_cpp | :heavy_check_mark: |
| rmw_opensplice_cpp | :heavy_check_mark: |
| rmw_cyclonedds_cpp | :heavy_check_mark: |
| rmw_fastrtps_dynamic_cpp | :heavy_check_mark: |
| rmw_connext_cpp | :heavy_check_mark: |
Build
- Install ROS 2 (https://index.ros.org/doc/ros2/Installation/).
File truncated at 100 lines see the full file
Package Dependencies
| Deps | Name |
|---|---|
| ament_cmake | |
| ament_cmake_python | |
| ament_index_python | |
| launch | |
| launch_ros | |
| ament_lint_auto | |
| ament_lint_common | |
| launch_testing_ament_cmake | |
| performance_test | |
| rmw_implementation_cmake | |
| rclcpp |
System Dependencies
Dependant Packages
Launch files
Messages
Services
Plugins
Recent questions tagged buildfarm_perf_tests at Robotics Stack Exchange
|
|
buildfarm_perf_tests package from buildfarm_perf_tests repobuildfarm_perf_tests |
ROS Distro
|
Package Summary
| Tags | No category tags. |
| Version | 0.0.0 |
| License | Apache License 2.0 |
| Build type | AMENT_CMAKE |
| Use | RECOMMENDED |
Repository Summary
| Description | Performance tests which run regularly on the buildfarm |
| Checkout URI | https://github.com/ros2/buildfarm_perf_tests.git |
| VCS Type | git |
| VCS Version | master |
| Last Updated | 2024-02-16 |
| Dev Status | UNKNOWN |
| Released | UNRELEASED |
| Tags | No category tags. |
| Contributing |
Help Wanted (-)
Good First Issues (-) Pull Requests to Review (-) |
Package Description
Additional Links
Maintainers
- Scott K Logan
Authors
buildfarm perf tests
Purpose
This package defines some tests.
On one hand it invokes perf_test from Apex.AI’s performance_test package.
This allows you to test performance and latency of several ROS 2 RMW implementations.
On the other hand we are evaluating the additional overhead caused by a single pub/sub topic or one process spinning and detect potential leaks related to theses activities.
-
There is a test for each RMW:
- CycloneDDS (standalone test)
- FastRTPS (standalone test)
- rmw_connext_cpp
- rmw_cyclonedds_cpp
- rmw_fastrtps_cpp
- rmw_fastrtps_dynamic_cpp
- rmw_opensplice_cpp
Test 1 - Performance Test (Apex.AI)
In this test we are running the Performance Test provided by Apex.AI. Right now we have our own fork because there are some pending pull requests in the official gitlab repository.
In this test we are measurement:
- Average latency
- CPU usage (provided by Apex.AI tool)
- Sent/Received messages per second
- Total lost messages
- Max resident set size
We are generating two plots per measurement
The test only measures the latency between the same RMW implementation
| Publisher/Subscriber | rmw_fastrtps_cpp | rmw_opensplice_cpp | rmw_cyclonedds_cpp | rmw_fastrtps_dynamic_cpp | rmw_connext_cpp |
|---|---|---|---|---|---|
| rmw_fastrtps_cpp | :heavy_check_mark: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: |
| rmw_opensplice_cpp | :heavy_multiplication_x: | :heavy_check_mark: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: |
| rmw_cyclonedds_cpp | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_check_mark: | :heavy_multiplication_x: | :heavy_multiplication_x: |
| rmw_fastrtps_dynamic_cpp | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_check_mark: | :heavy_multiplication_x: |
| rmw_connext_cpp | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_check_mark: |
Test 2 - Simple pub/sub
In this case we are testing one publisher and one subscriber in different processes sending a 1kArray at 5Hz. This will allow us to evaluate additional overhead caused by a single pub/sub topic and detect leaks related to this activity.
We measure for both publisher and subscriber:
- Average round trip
- CPU usage ( readed from the filesystem )
- Total lost messages
- Received/Sent messages per second
- Physical memory
- Resident anonymous memory
- Virtual memory
Again we plot measurement:
| Publisher/Subscriber | rmw_fastrtps_cpp | rmw_opensplice_cpp | rmw_cyclonedds_cpp | rmw_fastrtps_dynamic_cpp | rmw_connext_cpp |
|---|---|---|---|---|---|
| rmw_fastrtps_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| rmw_opensplice_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| rmw_cyclonedds_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| rmw_fastrtps_dynamic_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| rmw_connext_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
Test 3 - Node spinning
This test creates one process and spins for 1 minute to evaluate ROS 2 overhead and detect obvious leaks.
We measure:
- CPU usage ( readed from file system )
- Physical memory
- Resident anonymous memory
- Virtual memory
Again we plot measurement:
| DDS | Process 1 |
|---|---|
| rmw_fastrtps_cpp | :heavy_check_mark: |
| rmw_opensplice_cpp | :heavy_check_mark: |
| rmw_cyclonedds_cpp | :heavy_check_mark: |
| rmw_fastrtps_dynamic_cpp | :heavy_check_mark: |
| rmw_connext_cpp | :heavy_check_mark: |
Build
- Install ROS 2 (https://index.ros.org/doc/ros2/Installation/).
File truncated at 100 lines see the full file
Package Dependencies
| Deps | Name |
|---|---|
| ament_cmake | |
| ament_cmake_python | |
| ament_index_python | |
| launch | |
| launch_ros | |
| ament_lint_auto | |
| ament_lint_common | |
| launch_testing_ament_cmake | |
| performance_test | |
| rmw_implementation_cmake | |
| rclcpp |
System Dependencies
Dependant Packages
Launch files
Messages
Services
Plugins
Recent questions tagged buildfarm_perf_tests at Robotics Stack Exchange
|
|
buildfarm_perf_tests package from buildfarm_perf_tests repobuildfarm_perf_tests |
ROS Distro
|
Package Summary
| Tags | No category tags. |
| Version | 0.0.0 |
| License | Apache License 2.0 |
| Build type | AMENT_CMAKE |
| Use | RECOMMENDED |
Repository Summary
| Description | Performance tests which run regularly on the buildfarm |
| Checkout URI | https://github.com/ros2/buildfarm_perf_tests.git |
| VCS Type | git |
| VCS Version | master |
| Last Updated | 2024-02-16 |
| Dev Status | UNKNOWN |
| Released | UNRELEASED |
| Tags | No category tags. |
| Contributing |
Help Wanted (-)
Good First Issues (-) Pull Requests to Review (-) |
Package Description
Additional Links
Maintainers
- Scott K Logan
Authors
buildfarm perf tests
Purpose
This package defines some tests.
On one hand it invokes perf_test from Apex.AI’s performance_test package.
This allows you to test performance and latency of several ROS 2 RMW implementations.
On the other hand we are evaluating the additional overhead caused by a single pub/sub topic or one process spinning and detect potential leaks related to theses activities.
-
There is a test for each RMW:
- CycloneDDS (standalone test)
- FastRTPS (standalone test)
- rmw_connext_cpp
- rmw_cyclonedds_cpp
- rmw_fastrtps_cpp
- rmw_fastrtps_dynamic_cpp
- rmw_opensplice_cpp
Test 1 - Performance Test (Apex.AI)
In this test we are running the Performance Test provided by Apex.AI. Right now we have our own fork because there are some pending pull requests in the official gitlab repository.
In this test we are measurement:
- Average latency
- CPU usage (provided by Apex.AI tool)
- Sent/Received messages per second
- Total lost messages
- Max resident set size
We are generating two plots per measurement
The test only measures the latency between the same RMW implementation
| Publisher/Subscriber | rmw_fastrtps_cpp | rmw_opensplice_cpp | rmw_cyclonedds_cpp | rmw_fastrtps_dynamic_cpp | rmw_connext_cpp |
|---|---|---|---|---|---|
| rmw_fastrtps_cpp | :heavy_check_mark: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: |
| rmw_opensplice_cpp | :heavy_multiplication_x: | :heavy_check_mark: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: |
| rmw_cyclonedds_cpp | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_check_mark: | :heavy_multiplication_x: | :heavy_multiplication_x: |
| rmw_fastrtps_dynamic_cpp | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_check_mark: | :heavy_multiplication_x: |
| rmw_connext_cpp | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_check_mark: |
Test 2 - Simple pub/sub
In this case we are testing one publisher and one subscriber in different processes sending a 1kArray at 5Hz. This will allow us to evaluate additional overhead caused by a single pub/sub topic and detect leaks related to this activity.
We measure for both publisher and subscriber:
- Average round trip
- CPU usage ( readed from the filesystem )
- Total lost messages
- Received/Sent messages per second
- Physical memory
- Resident anonymous memory
- Virtual memory
Again we plot measurement:
| Publisher/Subscriber | rmw_fastrtps_cpp | rmw_opensplice_cpp | rmw_cyclonedds_cpp | rmw_fastrtps_dynamic_cpp | rmw_connext_cpp |
|---|---|---|---|---|---|
| rmw_fastrtps_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| rmw_opensplice_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| rmw_cyclonedds_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| rmw_fastrtps_dynamic_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| rmw_connext_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
Test 3 - Node spinning
This test creates one process and spins for 1 minute to evaluate ROS 2 overhead and detect obvious leaks.
We measure:
- CPU usage ( readed from file system )
- Physical memory
- Resident anonymous memory
- Virtual memory
Again we plot measurement:
| DDS | Process 1 |
|---|---|
| rmw_fastrtps_cpp | :heavy_check_mark: |
| rmw_opensplice_cpp | :heavy_check_mark: |
| rmw_cyclonedds_cpp | :heavy_check_mark: |
| rmw_fastrtps_dynamic_cpp | :heavy_check_mark: |
| rmw_connext_cpp | :heavy_check_mark: |
Build
- Install ROS 2 (https://index.ros.org/doc/ros2/Installation/).
File truncated at 100 lines see the full file
Package Dependencies
| Deps | Name |
|---|---|
| ament_cmake | |
| ament_cmake_python | |
| ament_index_python | |
| launch | |
| launch_ros | |
| ament_lint_auto | |
| ament_lint_common | |
| launch_testing_ament_cmake | |
| performance_test | |
| rmw_implementation_cmake | |
| rclcpp |
System Dependencies
Dependant Packages
Launch files
Messages
Services
Plugins
Recent questions tagged buildfarm_perf_tests at Robotics Stack Exchange
|
|
buildfarm_perf_tests package from buildfarm_perf_tests repobuildfarm_perf_tests |
ROS Distro
|
Package Summary
| Tags | No category tags. |
| Version | 0.0.0 |
| License | Apache License 2.0 |
| Build type | AMENT_CMAKE |
| Use | RECOMMENDED |
Repository Summary
| Description | Performance tests which run regularly on the buildfarm |
| Checkout URI | https://github.com/ros2/buildfarm_perf_tests.git |
| VCS Type | git |
| VCS Version | master |
| Last Updated | 2024-02-16 |
| Dev Status | UNKNOWN |
| Released | UNRELEASED |
| Tags | No category tags. |
| Contributing |
Help Wanted (-)
Good First Issues (-) Pull Requests to Review (-) |
Package Description
Additional Links
Maintainers
- Scott K Logan
Authors
buildfarm perf tests
Purpose
This package defines some tests.
On one hand it invokes perf_test from Apex.AI’s performance_test package.
This allows you to test performance and latency of several ROS 2 RMW implementations.
On the other hand we are evaluating the additional overhead caused by a single pub/sub topic or one process spinning and detect potential leaks related to theses activities.
-
There is a test for each RMW:
- CycloneDDS (standalone test)
- FastRTPS (standalone test)
- rmw_connext_cpp
- rmw_cyclonedds_cpp
- rmw_fastrtps_cpp
- rmw_fastrtps_dynamic_cpp
- rmw_opensplice_cpp
Test 1 - Performance Test (Apex.AI)
In this test we are running the Performance Test provided by Apex.AI. Right now we have our own fork because there are some pending pull requests in the official gitlab repository.
In this test we are measurement:
- Average latency
- CPU usage (provided by Apex.AI tool)
- Sent/Received messages per second
- Total lost messages
- Max resident set size
We are generating two plots per measurement
The test only measures the latency between the same RMW implementation
| Publisher/Subscriber | rmw_fastrtps_cpp | rmw_opensplice_cpp | rmw_cyclonedds_cpp | rmw_fastrtps_dynamic_cpp | rmw_connext_cpp |
|---|---|---|---|---|---|
| rmw_fastrtps_cpp | :heavy_check_mark: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: |
| rmw_opensplice_cpp | :heavy_multiplication_x: | :heavy_check_mark: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: |
| rmw_cyclonedds_cpp | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_check_mark: | :heavy_multiplication_x: | :heavy_multiplication_x: |
| rmw_fastrtps_dynamic_cpp | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_check_mark: | :heavy_multiplication_x: |
| rmw_connext_cpp | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_check_mark: |
Test 2 - Simple pub/sub
In this case we are testing one publisher and one subscriber in different processes sending a 1kArray at 5Hz. This will allow us to evaluate additional overhead caused by a single pub/sub topic and detect leaks related to this activity.
We measure for both publisher and subscriber:
- Average round trip
- CPU usage ( readed from the filesystem )
- Total lost messages
- Received/Sent messages per second
- Physical memory
- Resident anonymous memory
- Virtual memory
Again we plot measurement:
| Publisher/Subscriber | rmw_fastrtps_cpp | rmw_opensplice_cpp | rmw_cyclonedds_cpp | rmw_fastrtps_dynamic_cpp | rmw_connext_cpp |
|---|---|---|---|---|---|
| rmw_fastrtps_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| rmw_opensplice_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| rmw_cyclonedds_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| rmw_fastrtps_dynamic_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| rmw_connext_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
Test 3 - Node spinning
This test creates one process and spins for 1 minute to evaluate ROS 2 overhead and detect obvious leaks.
We measure:
- CPU usage ( readed from file system )
- Physical memory
- Resident anonymous memory
- Virtual memory
Again we plot measurement:
| DDS | Process 1 |
|---|---|
| rmw_fastrtps_cpp | :heavy_check_mark: |
| rmw_opensplice_cpp | :heavy_check_mark: |
| rmw_cyclonedds_cpp | :heavy_check_mark: |
| rmw_fastrtps_dynamic_cpp | :heavy_check_mark: |
| rmw_connext_cpp | :heavy_check_mark: |
Build
- Install ROS 2 (https://index.ros.org/doc/ros2/Installation/).
File truncated at 100 lines see the full file
Package Dependencies
| Deps | Name |
|---|---|
| ament_cmake | |
| ament_cmake_python | |
| ament_index_python | |
| launch | |
| launch_ros | |
| ament_lint_auto | |
| ament_lint_common | |
| launch_testing_ament_cmake | |
| performance_test | |
| rmw_implementation_cmake | |
| rclcpp |
System Dependencies
Dependant Packages
Launch files
Messages
Services
Plugins
Recent questions tagged buildfarm_perf_tests at Robotics Stack Exchange
|
|
buildfarm_perf_tests package from buildfarm_perf_tests repobuildfarm_perf_tests |
ROS Distro
|
Package Summary
| Tags | No category tags. |
| Version | 0.0.0 |
| License | Apache License 2.0 |
| Build type | AMENT_CMAKE |
| Use | RECOMMENDED |
Repository Summary
| Description | Performance tests which run regularly on the buildfarm |
| Checkout URI | https://github.com/ros2/buildfarm_perf_tests.git |
| VCS Type | git |
| VCS Version | master |
| Last Updated | 2024-02-16 |
| Dev Status | UNKNOWN |
| Released | UNRELEASED |
| Tags | No category tags. |
| Contributing |
Help Wanted (-)
Good First Issues (-) Pull Requests to Review (-) |
Package Description
Additional Links
Maintainers
- Scott K Logan
Authors
buildfarm perf tests
Purpose
This package defines some tests.
On one hand it invokes perf_test from Apex.AI’s performance_test package.
This allows you to test performance and latency of several ROS 2 RMW implementations.
On the other hand we are evaluating the additional overhead caused by a single pub/sub topic or one process spinning and detect potential leaks related to theses activities.
-
There is a test for each RMW:
- CycloneDDS (standalone test)
- FastRTPS (standalone test)
- rmw_connext_cpp
- rmw_cyclonedds_cpp
- rmw_fastrtps_cpp
- rmw_fastrtps_dynamic_cpp
- rmw_opensplice_cpp
Test 1 - Performance Test (Apex.AI)
In this test we are running the Performance Test provided by Apex.AI. Right now we have our own fork because there are some pending pull requests in the official gitlab repository.
In this test we are measurement:
- Average latency
- CPU usage (provided by Apex.AI tool)
- Sent/Received messages per second
- Total lost messages
- Max resident set size
We are generating two plots per measurement
The test only measures the latency between the same RMW implementation
| Publisher/Subscriber | rmw_fastrtps_cpp | rmw_opensplice_cpp | rmw_cyclonedds_cpp | rmw_fastrtps_dynamic_cpp | rmw_connext_cpp |
|---|---|---|---|---|---|
| rmw_fastrtps_cpp | :heavy_check_mark: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: |
| rmw_opensplice_cpp | :heavy_multiplication_x: | :heavy_check_mark: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: |
| rmw_cyclonedds_cpp | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_check_mark: | :heavy_multiplication_x: | :heavy_multiplication_x: |
| rmw_fastrtps_dynamic_cpp | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_check_mark: | :heavy_multiplication_x: |
| rmw_connext_cpp | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_check_mark: |
Test 2 - Simple pub/sub
In this case we are testing one publisher and one subscriber in different processes sending a 1kArray at 5Hz. This will allow us to evaluate additional overhead caused by a single pub/sub topic and detect leaks related to this activity.
We measure for both publisher and subscriber:
- Average round trip
- CPU usage ( readed from the filesystem )
- Total lost messages
- Received/Sent messages per second
- Physical memory
- Resident anonymous memory
- Virtual memory
Again we plot measurement:
| Publisher/Subscriber | rmw_fastrtps_cpp | rmw_opensplice_cpp | rmw_cyclonedds_cpp | rmw_fastrtps_dynamic_cpp | rmw_connext_cpp |
|---|---|---|---|---|---|
| rmw_fastrtps_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| rmw_opensplice_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| rmw_cyclonedds_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| rmw_fastrtps_dynamic_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| rmw_connext_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
Test 3 - Node spinning
This test creates one process and spins for 1 minute to evaluate ROS 2 overhead and detect obvious leaks.
We measure:
- CPU usage ( readed from file system )
- Physical memory
- Resident anonymous memory
- Virtual memory
Again we plot measurement:
| DDS | Process 1 |
|---|---|
| rmw_fastrtps_cpp | :heavy_check_mark: |
| rmw_opensplice_cpp | :heavy_check_mark: |
| rmw_cyclonedds_cpp | :heavy_check_mark: |
| rmw_fastrtps_dynamic_cpp | :heavy_check_mark: |
| rmw_connext_cpp | :heavy_check_mark: |
Build
- Install ROS 2 (https://index.ros.org/doc/ros2/Installation/).
File truncated at 100 lines see the full file
Package Dependencies
| Deps | Name |
|---|---|
| ament_cmake | |
| ament_cmake_python | |
| ament_index_python | |
| launch | |
| launch_ros | |
| ament_lint_auto | |
| ament_lint_common | |
| launch_testing_ament_cmake | |
| performance_test | |
| rmw_implementation_cmake | |
| rclcpp |
System Dependencies
Dependant Packages
Launch files
Messages
Services
Plugins
Recent questions tagged buildfarm_perf_tests at Robotics Stack Exchange
|
|
buildfarm_perf_tests package from buildfarm_perf_tests repobuildfarm_perf_tests |
ROS Distro
|
Package Summary
| Tags | No category tags. |
| Version | 0.0.0 |
| License | Apache License 2.0 |
| Build type | AMENT_CMAKE |
| Use | RECOMMENDED |
Repository Summary
| Description | Performance tests which run regularly on the buildfarm |
| Checkout URI | https://github.com/ros2/buildfarm_perf_tests.git |
| VCS Type | git |
| VCS Version | master |
| Last Updated | 2024-02-16 |
| Dev Status | UNKNOWN |
| Released | UNRELEASED |
| Tags | No category tags. |
| Contributing |
Help Wanted (-)
Good First Issues (-) Pull Requests to Review (-) |
Package Description
Additional Links
Maintainers
- Scott K Logan
Authors
buildfarm perf tests
Purpose
This package defines some tests.
On one hand it invokes perf_test from Apex.AI’s performance_test package.
This allows you to test performance and latency of several ROS 2 RMW implementations.
On the other hand we are evaluating the additional overhead caused by a single pub/sub topic or one process spinning and detect potential leaks related to theses activities.
-
There is a test for each RMW:
- CycloneDDS (standalone test)
- FastRTPS (standalone test)
- rmw_connext_cpp
- rmw_cyclonedds_cpp
- rmw_fastrtps_cpp
- rmw_fastrtps_dynamic_cpp
- rmw_opensplice_cpp
Test 1 - Performance Test (Apex.AI)
In this test we are running the Performance Test provided by Apex.AI. Right now we have our own fork because there are some pending pull requests in the official gitlab repository.
In this test we are measurement:
- Average latency
- CPU usage (provided by Apex.AI tool)
- Sent/Received messages per second
- Total lost messages
- Max resident set size
We are generating two plots per measurement
The test only measures the latency between the same RMW implementation
| Publisher/Subscriber | rmw_fastrtps_cpp | rmw_opensplice_cpp | rmw_cyclonedds_cpp | rmw_fastrtps_dynamic_cpp | rmw_connext_cpp |
|---|---|---|---|---|---|
| rmw_fastrtps_cpp | :heavy_check_mark: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: |
| rmw_opensplice_cpp | :heavy_multiplication_x: | :heavy_check_mark: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: |
| rmw_cyclonedds_cpp | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_check_mark: | :heavy_multiplication_x: | :heavy_multiplication_x: |
| rmw_fastrtps_dynamic_cpp | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_check_mark: | :heavy_multiplication_x: |
| rmw_connext_cpp | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_check_mark: |
Test 2 - Simple pub/sub
In this case we are testing one publisher and one subscriber in different processes sending a 1kArray at 5Hz. This will allow us to evaluate additional overhead caused by a single pub/sub topic and detect leaks related to this activity.
We measure for both publisher and subscriber:
- Average round trip
- CPU usage ( readed from the filesystem )
- Total lost messages
- Received/Sent messages per second
- Physical memory
- Resident anonymous memory
- Virtual memory
Again we plot measurement:
| Publisher/Subscriber | rmw_fastrtps_cpp | rmw_opensplice_cpp | rmw_cyclonedds_cpp | rmw_fastrtps_dynamic_cpp | rmw_connext_cpp |
|---|---|---|---|---|---|
| rmw_fastrtps_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| rmw_opensplice_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| rmw_cyclonedds_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| rmw_fastrtps_dynamic_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| rmw_connext_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
Test 3 - Node spinning
This test creates one process and spins for 1 minute to evaluate ROS 2 overhead and detect obvious leaks.
We measure:
- CPU usage ( readed from file system )
- Physical memory
- Resident anonymous memory
- Virtual memory
Again we plot measurement:
| DDS | Process 1 |
|---|---|
| rmw_fastrtps_cpp | :heavy_check_mark: |
| rmw_opensplice_cpp | :heavy_check_mark: |
| rmw_cyclonedds_cpp | :heavy_check_mark: |
| rmw_fastrtps_dynamic_cpp | :heavy_check_mark: |
| rmw_connext_cpp | :heavy_check_mark: |
Build
- Install ROS 2 (https://index.ros.org/doc/ros2/Installation/).
File truncated at 100 lines see the full file
Package Dependencies
| Deps | Name |
|---|---|
| ament_cmake | |
| ament_cmake_python | |
| ament_index_python | |
| launch | |
| launch_ros | |
| ament_lint_auto | |
| ament_lint_common | |
| launch_testing_ament_cmake | |
| performance_test | |
| rmw_implementation_cmake | |
| rclcpp |
System Dependencies
Dependant Packages
Launch files
Messages
Services
Plugins
Recent questions tagged buildfarm_perf_tests at Robotics Stack Exchange
|
|
buildfarm_perf_tests package from buildfarm_perf_tests repobuildfarm_perf_tests |
ROS Distro
|
Package Summary
| Tags | No category tags. |
| Version | 0.0.0 |
| License | Apache License 2.0 |
| Build type | AMENT_CMAKE |
| Use | RECOMMENDED |
Repository Summary
| Description | Performance tests which run regularly on the buildfarm |
| Checkout URI | https://github.com/ros2/buildfarm_perf_tests.git |
| VCS Type | git |
| VCS Version | master |
| Last Updated | 2024-02-16 |
| Dev Status | UNKNOWN |
| Released | UNRELEASED |
| Tags | No category tags. |
| Contributing |
Help Wanted (-)
Good First Issues (-) Pull Requests to Review (-) |
Package Description
Additional Links
Maintainers
- Scott K Logan
Authors
buildfarm perf tests
Purpose
This package defines some tests.
On one hand it invokes perf_test from Apex.AI’s performance_test package.
This allows you to test performance and latency of several ROS 2 RMW implementations.
On the other hand we are evaluating the additional overhead caused by a single pub/sub topic or one process spinning and detect potential leaks related to theses activities.
-
There is a test for each RMW:
- CycloneDDS (standalone test)
- FastRTPS (standalone test)
- rmw_connext_cpp
- rmw_cyclonedds_cpp
- rmw_fastrtps_cpp
- rmw_fastrtps_dynamic_cpp
- rmw_opensplice_cpp
Test 1 - Performance Test (Apex.AI)
In this test we are running the Performance Test provided by Apex.AI. Right now we have our own fork because there are some pending pull requests in the official gitlab repository.
In this test we are measurement:
- Average latency
- CPU usage (provided by Apex.AI tool)
- Sent/Received messages per second
- Total lost messages
- Max resident set size
We are generating two plots per measurement
The test only measures the latency between the same RMW implementation
| Publisher/Subscriber | rmw_fastrtps_cpp | rmw_opensplice_cpp | rmw_cyclonedds_cpp | rmw_fastrtps_dynamic_cpp | rmw_connext_cpp |
|---|---|---|---|---|---|
| rmw_fastrtps_cpp | :heavy_check_mark: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: |
| rmw_opensplice_cpp | :heavy_multiplication_x: | :heavy_check_mark: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: |
| rmw_cyclonedds_cpp | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_check_mark: | :heavy_multiplication_x: | :heavy_multiplication_x: |
| rmw_fastrtps_dynamic_cpp | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_check_mark: | :heavy_multiplication_x: |
| rmw_connext_cpp | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_multiplication_x: | :heavy_check_mark: |
Test 2 - Simple pub/sub
In this case we are testing one publisher and one subscriber in different processes sending a 1kArray at 5Hz. This will allow us to evaluate additional overhead caused by a single pub/sub topic and detect leaks related to this activity.
We measure for both publisher and subscriber:
- Average round trip
- CPU usage ( readed from the filesystem )
- Total lost messages
- Received/Sent messages per second
- Physical memory
- Resident anonymous memory
- Virtual memory
Again we plot measurement:
| Publisher/Subscriber | rmw_fastrtps_cpp | rmw_opensplice_cpp | rmw_cyclonedds_cpp | rmw_fastrtps_dynamic_cpp | rmw_connext_cpp |
|---|---|---|---|---|---|
| rmw_fastrtps_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| rmw_opensplice_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| rmw_cyclonedds_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| rmw_fastrtps_dynamic_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| rmw_connext_cpp | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
Test 3 - Node spinning
This test creates one process and spins for 1 minute to evaluate ROS 2 overhead and detect obvious leaks.
We measure:
- CPU usage ( readed from file system )
- Physical memory
- Resident anonymous memory
- Virtual memory
Again we plot measurement:
| DDS | Process 1 |
|---|---|
| rmw_fastrtps_cpp | :heavy_check_mark: |
| rmw_opensplice_cpp | :heavy_check_mark: |
| rmw_cyclonedds_cpp | :heavy_check_mark: |
| rmw_fastrtps_dynamic_cpp | :heavy_check_mark: |
| rmw_connext_cpp | :heavy_check_mark: |
Build
- Install ROS 2 (https://index.ros.org/doc/ros2/Installation/).
File truncated at 100 lines see the full file
Package Dependencies
| Deps | Name |
|---|---|
| ament_cmake | |
| ament_cmake_python | |
| ament_index_python | |
| launch | |
| launch_ros | |
| ament_lint_auto | |
| ament_lint_common | |
| launch_testing_ament_cmake | |
| performance_test | |
| rmw_implementation_cmake | |
| rclcpp |