guery

Guery v0.1

Guery is a pure-go implementation of distributed SQL query engine for big data (like Presto). It is composed of one master and many agents and supports to query vast amounts of data using distributed queries.

Status

This project is still a work in progress.

Todo list

• Support more SQL statement
• Improve performance
• Improve error control
• Add ut and regression test
• Add Wiki
• ...

NEED YOUR HELP !!!

This project is in progress and a lot of work need to do. If you are interested in it, please give me your hands and anything is welcome! I also list some issues. If you want to improve something, please open an issue to let others know :)

• slack
• google-group

Supported

• Datasource: hive
• FileSystem: local/hdfs/s3
• FileType: csv/parquet/orc
• DataType: bool/int32/int64/float32/float64/string/date/timestamp

Web UI

Web UI is the web interface of a Guery cluster to monitor and inspect the task executions in a web browser. It provides following modules:

Task information/Agent information & management/Execute plan

Deploy

• build guery cd $GOPATH/src/github.com/xitongsys/guery; make build
• run master ./guery master --address 127.0.0.1:1111 --config ./config.json >> m.log &
• run agent ./guery agent --master 127.0.0.1:1111 --config ./config.json >> e.log &
• web ui http://127.0.0.1:1111

Demo

cd $GOPATH/src/github.com/xitongsys/guery
make run
curl -XPOST -d"sql=select * from test.test.csv where var1=1" 127.0.0.1:1111/query
open web browser: http://127.0.0.1:1111

Benchmark

I did a simple and not strict performance test:

env: 1 master and 2 workers(8 CPU, 16G RAM)

table: hive, parquet