Add `git_branch_name` and `github_job_run_link` to `environment.csv` for benchmarks

[tt-rkim]

Requested by @skhorasganiTT

[tt-rkim]

@FrancesssZ @skhorasganiTT We have difficulty getting github_job_id within the actual GitHub running job. It sucks. I can explain more if you'd like.

Because of this, @skhorasganiTT are you ok if we just link to the workflow (whole pipeline)? Rather than the specific job? I propose we change github_job_run_link to github_pipeline_run_link.

[skhorasganiTT]

Discussed this with @tt-rkim , the pipeline link is sufficient for our request.

[FrancesssZ]

@tt-rkim @skhorasganiTT I'll name it as github_pipeline_link to align with another column that we already have in the env file called github_pipeline_id. Please let me know if you're both good with this.

[tt-rkim]

We're both good

[FrancesssZ]

Changes on data team have been deployed @tt-rkim @skhorasganiTT

[uaydonat]

@tt-rkim this issue is a blocker for us for to be able to use the dashboard in our daily analysis. Do you have an ETA? If not soon, then we can take a look?

[tt-rkim]

@skhorasganiTT , for these two fields, edit function infra/data_collection/github/utils.py:create_csv_for_github_benchmark_environment:

• Add the two fields to BENCHMARK_ENVIRONMENT_CSV_FIELDS at top
• git_branch_name: Get the value from GITHUB_REF_NAME environment variable
• github_job_run_link: You can generate at the presentation layer by using github_pipeline_id and git_repo_name: the value should be https://github.com/<git_repo_name>/actions/runs/<github_pipeline_id>. Test this first though to make sure. No code changes needed, as this is a field that can be generated from the current data. We don't need to produce.

[skhorasganiTT]

Thanks @tt-rkim

[tt-rkim]

@skhorasganiTT pls close if done