python / python/pyperformance

"upload" only works with "compiled" benchmarks

Open
#38 2 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Python
Stars
1k
Forks
203
Avg merge
1h 20m
Merged PRs (30d)
2

Description

My workflow for PyPy is

  1. download latest version compiled on buildbot,
  2. run benchmarks,
  3. upload data (using command line option upload, which goes through compile.py, but should it?.

The cmd_upload function in compile.py assumes BenchmarkRevision.update_metadata from compile.py has been run to add commit information (commit_id, commit_branch, commit_date) and possibly patch_file to the json benchmark info. But because I am not running compile this info is lacking. Possible solutions:

  • add a download_prebuilt option to compile.py and the Python class to support this workflow
  • refactor the update_metadata method from BenchmarkRevision to become part of every benchmark run

Any thoughts? Should there be an upload.py separate from compile.py?

Contributor guide

No contributing guide indexed for this repository

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start by reading compile.py, especially cmd_upload, BenchmarkRevision.update_metadata, and the Python class. Trace how downloaded buildbot benchmarks differ from compiled benchmarks and how metadata is added before upload. Done means the described PyPy workflow can upload benchmark data without requiring a local compile step, with the chosen separation or refactoring covered by relevant tests.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
performance, tooling
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
30/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.