Skip to content

Reproducibility package for "Two Birds With One Stone: Designing a Hybrid Cloud Storage Engine for HTAP"

License

Notifications You must be signed in to change notification settings

umbra-db/colibri-vldb2024

Repository files navigation

Colibri Benchmark Suite

Copyright (c) 2024 TUM. All rights reserved.

Dependencies

The actual benchmark code is encapsulated within multiple Docker containers that are orchestrated using Docker Compose. We host pre-built Docker images for convenience, i.e. there is no need to manually build Docker containers. The docker subdirectory contains the build scripts used to generate the Docker images. To run the benchmarks, the following dependencies are required.

  • bash
  • Docker
  • Docker Compose

Data generation

Before running any benchmarks, the respective data sets have to be generated once. This can be achieved as follows, assuming that the current working directory contains this README file.

./benchmark/00_datagen.sh

The data will be generated into data under the root folder. After successful data generation, this directory should contain the following subdirectories:

data
├─── tatp
|    └─── ...               Multiple .csv files
├─── tpcc
|    └─── ...               Multiple .csv files
└─── tpch
     ├─── delete
     |    └─── ...          Multiple delete.N files
     ├─── update
     |    └─── ...          Multiple lineitem.tbl.uN and orders.tbl.uN files
     └─── ...               Multiple .tbl files

Running benchmarks

Login to the container registry using: docker login gitlab.db.in.tum.de:5005 -u test -p glpat-sgXycvrv8ctxUMko7NAJ

Once data has been generated, the benchmarks can be run as follows. Once again, the listed commands assume that the working directory contains this README file.

  1. ./benchmark/03_benchmark_postgres.sh
  2. ./benchmark/05_benchmark_umbra.sh
  3. ./benchmark/06_benchmark_umbra_column.sh
  4. ./benchmark/07_benchmark_umbra_async.sh
  5. ./benchmark/11_benchmark_umbra_aws_tpcc.sh

All files created by the database will be stored in db under the root folder. It can be linked to an SSD RAID, an EBS device, or any other storage device.

Benchmark results will be placed into the directories output/postgres, output/umbra, or output/singlestore for the respective system.

SingleStore requires a license key: update the benchmark/singlestore/docker-compose.yml file. For running Umbra with Amazon S3 update the benchmark/umbra_cloud/config.json file with the S3 credentials and bucket name.

Plots

The scripts under plots will generate the plots from the paper, the results are under results.

About

Reproducibility package for "Two Birds With One Stone: Designing a Hybrid Cloud Storage Engine for HTAP"

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages