Skip to content
FunctionBench
Python
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
aws update output latency Sep 21, 2019
azure refactoring linpack code Apr 24, 2019
dataset adding logistic regression model Apr 29, 2019
docs/images adding function_bench_title png Apr 25, 2019
google adding ml_lr_prediction requirements.txt Jan 30, 2019
README.md Update README.md Sep 21, 2019

README.md

function_bench_title

PUBLICATION

Jeongchul Kim and Kyungyong Lee, 'Function Bench : A Suite of Workloads for Serverless Cloud Function Service', IEEE International Conference on Cloud Computing 2019, 07/2019 pdf

The FunctionBench is composed of micro benchmark and application workload; the micro-benchmark uses simple system calls to measure performance of resources exclusively, and the application-benchmark represents realistic data-oriented applications that generally utilize various resources together. With the introduction of FunctionBench, we believe that researchers can easily deploy function applications to evaluate their proposed systems and algorithms justly. To the best of our knowledge, the proposed FunctionBench is the first publicly available realistic FaaS workload suites that are ready to be deployed on public cloud services.

FunctionBench workloads

1. CPU & Memory

2. Disk

3. Network

Required Cloud Service

AWS

Google Cloud

Azure

You can’t perform that action at this time.