Skip to content

Build all the Space ROS world packages for CI tests - #58

Open
robwoolley wants to merge 6 commits into
elisa-tech:mainfrom
robwoolley:spaceros-world
Open

robwoolley wants to merge 6 commits into
elisa-tech:mainfrom
robwoolley:spaceros-world

Conversation

@robwoolley

Copy link
Copy Markdown
Contributor

We temporarily reduce the tests to build the core packages to allow the CI to progress.

This restores it to build all the packages.

This reverts commit 0e190df.

Signed-off-by: Rob Woolley <rob.woolley@windriver.com>
Signed-off-by: Rob Woolley <rob.woolley@windriver.com>
Signed-off-by: Rob Woolley <rob.woolley@windriver.com>
The output from bitbake can be very verbose. This can make it difficult
for CI systems that get overwhelmed by the sheer volume of output.

Signed-off-by: Rob Woolley <rob.woolley@windriver.com>
Signed-off-by: Rob Woolley <rob.woolley@windriver.com>
@robwoolley

robwoolley commented Jun 19, 2026

Copy link
Copy Markdown
Contributor Author

I made a change to put bitbake in quiet mode to supress the information messages. Warnings and errors should still appear. I also added the console-latest.log as a build artifact. It may be found in the artifact ZIP file with the information messages intact.

Still no clue as to why the world builds are failing. They now terminate after 15 minutes instead of an hour. It is also not the build output filling up the CI log.

Could we perhaps be filling up the local storage for the runner or consuming so much RAM that the OOM killer is being triggered?

@robwoolley
robwoolley force-pushed the spaceros-world branch 2 times, most recently from f39d655 to f374462 Compare June 26, 2026 14:12
The intent behind this split is to confirm that each packagegroup
builds individually for Space ROS.

This should help diagnose problems in the CI system when building
the Space ROS world packagegroup.

Signed-off-by: Rob Woolley <rob.woolley@windriver.com>
@robwoolley

Copy link
Copy Markdown
Contributor Author

I made a change to add each packagegroup for Space ROS on BeagleV-Fire as an individual CI test. Everything passes which seems to indicate that it is an infrastructure problem. I think there may be problems with having so many test cases and running them all on all commits.

We may want to consider how to scale out the CI infrastructure to make it more efficient for people that aren't making changes that need testing with Space ROS.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant