Skip to content
/ UDO Public

Universal Database Optimization using Reinforcement Learning

License

Notifications You must be signed in to change notification settings

jxiw/UDO

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

73 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

UDO Quickstart

The following installation procedure was tested on Ubuntu 20.04 with Python 3. More precisely, we used a t3.medium EC2 instance and the "Ubuntu Server 20.04 LTS (HVM), SSD Volume Type" AMI.

Installation from GitHub

  1. Download UDO package from UDO repository and switch to UDO directory.

    git clone https://[username]@github.com/OVSS/UDO.git
    cd UDO
    
  2. Install DBMS packages. bash ./install.sh

  3. Install UDO packages.

    cd udo
    python3 -m pip install -r requirement.txt`
    
  4. Install UDO Gym environment.

    cd udo-optimization/
    python3 -m pip install -e .
    

Installation via PIP

  1. Install DBMS packages via bash ./install.sh if Postgres and MySQL are not installed.

  2. Use python3 -m pip install UDO-DB to install packages.

Prepare TPC-H Database

The TPC-H schema, dataset, and queries are available at https://drive.google.com/drive/folders/123pwHaoz8C1dakvUef8AjKqci3_JNG47.

  1. To download data from Google Drive, install gdown via python3 -m pip install gdown.

  2. Download TPC-H .zip file using /home/ubuntu/.local/bin/gdown https://drive.google.com/uc?id=1IgzHMOc75Km9h-FLMepV-t9lrQhWGTwt.

  3. Install unzip via sudo apt install unzip and use it to extract files via unzip TPC-H.zip

  4. Create TPC-H database via sudo -u postgres createdb tpch_sf10.

  5. Create TPC-H database schema via sudo -u postgres psql tpch_sf10 < tpch_schema.sql.

  6. Load TPC-H data via sudo -u postgres psql tpch_sf10 < tpch_sf10_data.sql.

Use UDO to Tune Postgres for TPC-H

  1. Optional, using the UDO tool to extract indexes. The output format should be index name;table;columns, which is the index format required by UDO.

    usage: extract_index.py [-h] [-db_schema DB_SCHEMA] [-queries QUERIES]
    
    UDO index candidate generator.
    
    optional arguments:
      -h, --help            show this help message and exit
      -db_schema DB_SCHEMA  the database schmea to optimizes
      -queries QUERIES      queries
    
  2. echo "PATH=$PATH:/home/ubuntu/.local/bin" >> ~/.bashrc

  3. Run agents after installing udo from pip

     usage: python3 -m udo [-h] [-system {mysql,postgres}] [-db DB] [-username USERNAME] [-password PASSWORD] [-queries QUERIES]
               [-indices INDICES] [-sys_params SYS_PARAMS] [-duration DURATION] [-agent {udo,udo-s,ddpg,sarsa}]
               [-horizon HORIZON] [-heavy_horizon HEAVY_HORIZON] [-rl_update {RAVE,MCTS}] [-rl_select {UCB1,UCBV}]
               [-rl_reward {delta,accumulate}] [-rl_delay {UCB,Exp3}] [-rl_max_delay_time RL_MAX_DELAY_TIME]
               [-sample_rate SAMPLE_RATE] [-default_query_time_out DEFAULT_QUERY_TIME_OUT] [-time_out_ratio TIME_OUT_RATIO]
               [--load_json LOAD_JSON]
    
     UDO optimizer.
     
     optional arguments:
       -h, --help            show this help message and exit
       -system {mysql,postgres}
                             Target system driver
       -db DB                the database to optimizes
       -username USERNAME    username
       -password PASSWORD    password
       -queries QUERIES      the input query file
       -indices INDICES      the input query file
       -sys_params SYS_PARAMS
                             the input system params json file
       -duration DURATION    time for tuning in hours
       -agent {udo,udo-s,ddpg,sarsa}
                             reinforcement learning agent
       -horizon HORIZON      the number horizon for reinforcement agent
       -heavy_horizon HEAVY_HORIZON
                             the number horizon for heavy parameters in UDO
       -rl_update {RAVE,MCTS}
                             the update policy of UDO tree search
       -rl_select {UCB1,UCBV}
                             the selection policy of UDO tree search
       -rl_reward {delta,accumulate}
                             the reward of reinforcement learning agent
       -rl_delay {UCB, Exp3}  the delay selection policy
       -rl_max_delay_time RL_MAX_DELAY_TIME
                             the delay selection policy
       -sample_rate SAMPLE_RATE
                             sampled rate from workload
       -default_query_time_out DEFAULT_QUERY_TIME_OUT
                             default timeout in seconds for each query
       -time_out_ratio TIME_OUT_RATIO
                             timeout ratio respect to default time
       --load_json LOAD_JSON
                             Load settings from file in json format. Command line options override values in file.
    

For example

for TPC-H with scaling factor 1

python3 -m udo -system postgres -db tpch_sf1 -username postgres -queries tpch_queries -indices tpch_index.txt -sys_params postgressysparams.json -duration 5 -agent udo -horizon 8 -heavy_horizon 3 -rl_max_delay_time 5 -default_query_time_out 6

for TPC-H with scaling factor 10

python3 -m udo -system postgres -db tpch_sf10 -username postgres -queries tpch_queries -indices tpch_index.txt -sys_params postgressysparams.json -duration 5 -agent udo -horizon 8 -heavy_horizon 3 -rl_max_delay_time 5 -default_query_time_out 18

Citation

If you use our repository, please cite the following papers.

@article{wang2021udo,
  title={UDO: universal database optimization using reinforcement learning},
  author={Wang, Junxiong and Trummer, Immanuel and Basu, Debabrota},
  journal={Proceedings of the VLDB Endowment},
  volume={14},
  number={13},
  pages={3402--3414},
  year={2021},
  publisher={VLDB Endowment}
}

@inproceedings{wang2021demonstrating,
  title={Demonstrating UDO: A Unified Approach for Optimizing Transaction Code, Physical Design, and System Parameters via Reinforcement Learning},
  author={Wang, Junxiong and Trummer, Immanuel and Basu, Debabrota},
  booktitle={Proceedings of the 2021 International Conference on Management of Data},
  pages={2794--2797},
  year={2021}
}

About

Universal Database Optimization using Reinforcement Learning

Resources

License

Stars

Watchers

Forks

Packages

No packages published