mars-project / mars-project/mars
Optimize preparation of huge graphs
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 2.7k
- Forks
- 325
- PR merge metrics
- No merged PRs in 30d
Description
Is your feature request related to a problem? Please describe.
When running huge graphs (~80k nodes) in Mars, runners have to wait for a long time (~20min) before the graph is tiled into chunks and actually start running. Therefore optimization is needed.
Describe the solution you'd like
All time-consuming steps and possible optimizations:
- graph tiling (~10min -> 1.5min by applying #107)
- initial placement (~6min)
- creating and distributing operands (~2min) -> eliminated by upgrading to v0.7
These steps can be optimized one by one.
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Begin by measuring the remaining initial-placement step in huge-graph runs, after reviewing the graph-tiling optimization referenced in #107 and the eliminated operand-distribution work. Done means a concrete optimization reduces the preparation delay and is validated on an approximately 80k-node graph.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- performance
- Issue type
- Feature
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 35/100