spcl / spcl/serverless-benchmarks

New experiments

Open
#239 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

aws azure enhancement gcp
Dominant language
Python
Stars
199
Forks
105
PR merge metrics
No merged PRs in 30d

Description

Currently, we have two experiments fully supported - perf-cost and invocation latency - and two work-in-progress: container evictions and communication benchmarks. We want to extend this with new types of experiments:

  • Invocation latency that would work without requirement on direct connections (no ping-pong).
  • Scalability experiments.
  • Queue and storage invocation latency (see pending PR for support #220 and #201).
  • Multi-threaded performance - many vCPUs on function.
  • Performance stability and straggler frequency.
  • HPL-like experiment to measure max flops per CPU.

Contributor guide

Open the contributing guide

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 reviewing the existing perf-cost and invocation latency experiments, along with the pending support referenced in PRs #220 and #201. Choose and scope one experiment type first; done should include a defined benchmark implementation and a way to measure its results.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
performance
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
20/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.