Description

This PR adds a new package autoware_component_monitor under the system. It aims to attach existing component containers and monitor their CPU and memory usages.

Related links

Issue:

• Measuring performance of component containers #7092

Corresponding message type PR:

• feat(system): create system usage message autoware_internal_msgs#12

Tests performed

Logging simulator was used with the sample data on Autoware tutorials. The composable node was attached to pointcloud_container and published the system usage of process. The launch file inside the package was used to attach. Here is a sample topic output:

---
header:
  stamp:
    sec: 1718094543
    nanosec: 106255881
  frame_id: component_monitor
pid: 225916
cpu_usage_percentage: 53.29999923706055
total_memory_bytes: 67104985088
free_memory_bytes: 7425822720
used_memory_bytes: 759532
memory_usage_percentage: 1.2000000476837158

Here is a sample visualization of pointcloud_container with the sample bag on Foxglove:

Notes for reviewers

The CPU monitor inside the system_monitor package uses boost::process to run mpstat command, gets the stdout and parses it to get necessary information.

The same mechanism was also used in this package. The top package was preferred to get CPU and memory usage.
top -b -d 0.1 -n 1 -p PID outputs the memory usage as below:

top - 15:48:04 up  4:50,  1 user,  load average: 6,59, 4,29, 3,46
Tasks:   1 total,   0 running,   1 sleeping,   0 stopped,   0 zombie
%Cpu(s):  3,8 us,  0,9 sy,  0,0 ni, 94,5 id,  0,4 wa,  0,0 hi,  0,4 si,  0,0 st
MiB Mem :  63996,4 total,   5108,0 free,  40659,5 used,  18228,9 buff/cache
MiB Swap:   2048,0 total,    906,5 free,   1141,5 used.  21457,1 avail Mem

    PID USER      PR  NI    VIRT    RES    SHR S  %CPU  %MEM     TIME+ COMMAND
 573147 meb       20   0   16,6g   2,0g 549812 S   6,7   3,2   0:05.55 component_conta

• %MEM is calculated with the RES field, which is the physical memory usage. These fields are parsed inside the node for memory usage.

• %CPU field is parsed to get CPU usage rate.

• The LC_NUMERIC environment variable determines the output format of floating numbers. In my computer it is LC_NUMERIC=tr_TR.UTF-8 and because of that top uses , as the separator for floating numbers. To ensure that it is always ., this environment variable is overridden.

env["LC_NUMERIC"] = "en_US.UTF-8"; // To make sure that decimal separator is a dot.

• A while (rclcpp::ok()) loop was used with a rclcpp::Rate to monitor periodically. The loop is inside a try-catch block so that if the node dies, it won't crash the whole component.

Interface changes

ROS Topic Changes

A message type was created as follows (autowarefoundation/autoware_internal_msgs#12):

std_msgs/Header header

# Process identifier
uint32 pid

# CPU usage metrics
float32 cpu_usage_percentage

# Memory usage metrics for the whole system
uint64 total_memory_bytes
uint64 free_memory_bytes

# Memory usage metrics for the process
uint64 used_memory_bytes
float32 memory_usage_percentage

It is currently planned to have this message inside autoware_internal_msgs.

ROS Parameter Changes

Currently, there aren't any parameters.

Effects on system behavior

It doesn't affect the system. The node is attached to an existing container and publishes a message if there is a subscriber.

Pre-review checklist for the PR author

The PR author must check the checkboxes below when creating the PR.

• I've confirmed the contribution guidelines.
• The PR follows the pull request guidelines.

In-review checklist for the PR reviewers

The PR reviewers must check the checkboxes below before approval.

• The PR follows the pull request guidelines.
• The PR has been properly tested.
• The PR has been reviewed by the code owners.

Post-review checklist for the PR author

The PR author must check the checkboxes below before merging.

• There are no open discussions, or they are tracked via tickets.
• The PR is ready for merge.

After all checkboxes are checked, anyone who has write access can merge the PR.