Add support for jmx prometheus exporter

[fjammes]

• Add python/scala code to store sparkMeasure metrics inside JMX
• Write unit test for jmxexport() function
• Add documentation
• Document how to run unit test
• Add integration tests over k8s

[fjammes]

Hi @LucaCanali , here is the PR related to issue #67 . Thanks in advance for you review. In addition, I've added a job which run e2e tests inside Github action.

[fjammes]

Hi @LucaCanali , I am trying to fix a minor error in e2e test in Github Actions (it runs fine on a regular commit but not on a PR). Could you please approve my 2 workflows?

As you can see it works fine on a regular commit push: https://github.com/k8s-school/sparkMeasure/actions/runs/16571225806/job/46863741443
but the branch name is not managed the same way in git in a PR and this crashes the e2e tests. I'm currently trying to fix this.

[fjammes]

Hi @LucaCanali

Would you mind re-triggering the tests on the PR?
It's a bit tricky for me to test on my side because the bug only seems to happen on the PR itself, not on regular commits.

I even created a fake PR on my own repo to simulate the scenario and everything went fine there:
k8s-school#1
— so I'm hoping it'll also work sparkMeasure official repository.

Thanks!

[fjammes]

Hi @LucaCanali,

All CI tests have passed. Could you please complete your review and then merge?
Let me know if you'd like to have a short meeting so I can walk you through what the code is doing.

Thanks!

Best,
Fabrice

[LucaCanali]

Thank you @fjammes for all the work on this.
A minor comment about Dockerfile, can you please add some header comment there on the general purpose of this Dockerfile and transalte the existing comments into English.

[LucaCanali]

Another point concerns the maintainability of this new component, including its code, Dockerfiles, and dependencies. Could we consider a strategy to make future troubleshooting and dependency updates easier, especially as new versions become available?

[fjammes]

Hi @LucaCanali,

I'm happy to dedicate time to the long-term maintenance of this component.
To simplify usage and onboarding, I've just provided a one-liner script (e2e/run-all.sh) that builds the local Spark Measure container and runs it in Spark on Kubernetes—everything needed to get started on a Linux workstation.

Feel free to ask if you need anything else.

Best regards,
Fabrice

[LucaCanali]

I have merged the change however there seems to be a problem with the job Run spark-measure end-to-end tests