Chaining multiple experiments
Open
Nobody has claimed this yet.
feature
- Dominant language
- Python
- Stars
- 78
- Forks
- 52
- Avg merge
- 4d 19m
- Merged PRs (30d)
- 10
Description
We often need to run multiple jobs:
- Pre- and post-job system health checks
- Simple workflows with dependencies
- Throughput workflows (e.g., multiple jobs on the node)
We could use Chained experiments in Ramble
ramble workspace info will print out the ordering of tests in a chain::
└─▪ ramble -D . workspace info --expansions
Workspace: chained-exp-example
...
Experiments:
Application: intel-mpi-benchmarks
Workload: collective
Experiment 2: intel-mpi-benchmarks.collective.Alltoall_08
...
Experiment Chain:
- intel-mpi-benchmarks.collective.Alltoall_08.chain.0.intel-mpi-benchmarks.collective.bcast_template
- intel-mpi-benchmarks.collective.Alltoall_08
- intel-mpi-benchmarks.collective.Alltoall_08.chain.1.intel-mpi-benchmarks.collective.bcast_template
Also, https://github.com/GoogleCloudPlatform/ramble/pull/907 fixes chained experiments.
Contributor guide
No contributing guide indexed for this repository
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
Start with the linked Ramble experiment-chain documentation and run ramble workspace info --expansions on the example workspace to understand the expected ordering. Determine which Benchpark entry points and tests would need to support health checks, dependent jobs, and throughput workflows, then verify the requested chained experiments work end to end.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- hpc
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100