Could you explain why using loglevels would not work? It feels to me they are meant to solve the problem of selecting the appropriate amount of debug messages… Having a separate flag introduce quite some extra noise in the process.

See D8894 for an alternative implementation.

lgtm

The second part looks good to me. :)

Review the first part.

lgtm

After discussing D8833 and discussing the general API with @olasd, it feels like adding a delete method to swh.journal.writer.interface.JournalWriterInterface is the wrong approach. To try to capture our discussion

Fix the issue by adding a level of indirection in the yaml (replacing the job
by an identical job-template, and instantiating it through a project).
It seems jinja2 templates aren't actually supported in direct job definitions,
only in job templates. Thanks to olasd for finding this out and suggesting a fix.

Rebase

In D8634#227186, @douardda wrote:

with the same idea as the inline comment, it would be better to rename the network like 'swh-mirror' rather than 'swhtest-mirror'.

• Rename swhtest-mirror network to swh-mirror
• Re-add monitoring nodes to the required services to run the tests
• Add missing label to troubleshooting section

Screenshot of resources usage while test runs on my laptop:

I have missed updating tests/README.md which currently mentions testinfra. I should make the change if the rest of these changes are accepted.

Use the template appropriately to run the Docker agent

Here is screenshot of the resource monitor on my system while running the test:

Add two more commits required for predictable runs in Jenkins:

• Remove the uneeded tree_sitter from build system dependencies.
• Remove some leftover breakpoints.

Adjust wait_services_status() for a race condition

Have the log instruction perform the formatting intead of using f-strings

It would do nicely. Thanks! :)

Fix the typos spotted by ardumont