https://github.com/alexandervanrenen/cab
A benchmark for serverless analytic databases.
https://github.com/alexandervanrenen/cab
Last synced: about 2 months ago
JSON representation
A benchmark for serverless analytic databases.
- Host: GitHub
- URL: https://github.com/alexandervanrenen/cab
- Owner: alexandervanrenen
- License: mit
- Created: 2022-04-13T08:50:29.000Z (about 3 years ago)
- Default Branch: main
- Last Pushed: 2024-09-21T08:41:43.000Z (9 months ago)
- Last Synced: 2025-03-30T16:19:03.420Z (3 months ago)
- Language: R
- Homepage:
- Size: 2.74 MB
- Stars: 20
- Watchers: 2
- Forks: 3
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE.md
Awesome Lists containing this project
README
# Cloud Analytics Benchmark (CAB)
This repository contains the implementatio of the CAB benchmark, as described in [this paper](https://www.vldb.org/pvldb/vol16/p1413-renen.pdf).
A short description of the individual folders and how to use them.
#### ``cab/benchmark-gen``
A c++ program to generate the query streams. Can be configured by changing the variables in the ``main`` function:
```C++
const uint64_t total_size = 4_TB;
const uint64_t total_cpu_hours = 40;
const uint64_t total_duration_in_hours = 1;
const uint64_t database_count = 20;
```
The output is written to ``benchmark-gen/query_streams``.
It can be compiled with:
```bash
clang++ -std=c++17 -Wall -Werror=return-type -Werror=non-virtual-dtor -Werror=sequence-point -Wsign-compare -march=native -O2 -Wfatal-errors benchmark.cpp
```
Note that the distributions in c++ are platform dependent, hence the generated query streams might look different depending on the platform.
However, the overall distribution/pattern is the same.
### ``cab/benchmark-query-streams``
Contains pre generated query streams.
These are the ones we used in the experiments in the paper.
### ``cab/benchmark-results``
Contains the results we obtained by running the experiments as described in the paper.
### ``cab/benchmark-run``
A set of java script programs for running and analyzing the benchmark results.
### ``cab/snowset-analysis``
All R scripts we used to analyze the snowset and to plot the results of the benchmark run.