Skip to content

Latest commit

 

History

History
78 lines (52 loc) · 3.41 KB

README.md

File metadata and controls

78 lines (52 loc) · 3.41 KB
Raw

Pandora Hydra server

Hydra's Spring Boot backend that is responsible for persisting test results and partitioning tests into equally sized chunks. The Hydra server is stateful so only one instance should be run per testing environment.

Persistence

Hydra can be configured to use two different persistence mechanisms: file based or SQL based.

File Based Persistence

This mechanism simply maintains a flat file on the file system that stores test times. To use file backed storage start Hydra server with the Spring profile file_store enabled (spring.profiles.active=file_store). The persistence directory can be configured by changing the property hydra.repo (it defaults to server-root-folder/.repo)

SQL Based Persistence

SQL is used as the default persistence mechanism. We've tested with PostgreSQL, but other flavors of SQL should also work.

First create the hydra db

createdb hydra

Connect to the database, create a new role, and change DB ownership to the new user. The username and password is set in application.yml and should correspond to the owner of the DB

psql
CREATE ROLE hydra LOGIN;
ALTER DATABASE hydra OWNER TO hydra;
\q

Flyway is used to manage the schemas.

Balancing Strategies

There are three different strategies available creating equally sized test partitions

  • greedy
  • greedy_with_failures
  • affinity

The greedy strategy calculates fresh test partitions for every test run. The most recent test time data is used to calculate the partitions.

The greedy_with_failures strategy calculates fresh test partitions for each test run, but will always put tests that failed on the same host

The affinity strategy tries to keep tests on the same host across test runs. It will rebalance tests if the expected test times across hosts starts to vary too much

If your test cluster consists of a fixed set of servers then affinity is the best strategy. The affinity strategy will eventually lead to the most similarly sized test partitions, because it will eventually account for differences between different nodes on the cluster.

greedy_with_failures can be useful if you have flakey tests that only fail when run in a specific order.

The strategy can be configured with the argument hydra.partition.strategy in application.yml or by passing the argument in as a command line argument, e.g. --hydra.partition.strategy=greedy

Deployment

There are two possible options to deploy the application -- Standalone or Docker container.

Standalone app

The hydra-server is a Spring Boot application, so it does not need a separate container.

  • Build the server with command ./gradlew build (from root folder)
  • Run the server with command java -jar hydra.war (war file is in build/libs, please rename it for convenience)
  • To use different profiles or partition strategies java -jar hydra.war --spring.profiles.active=file_repo --hydra.partition.strategy=greedy

Docker container

Dockerfile and entrypoint.sh scripts are shipped along with the source. War file has version in it's name, please rename it to hydra.war for convenience. Please modify application.yml and set the correct DB host/port/dbname for the docker profile. Then switch to the hydra-server folder and run:

docker build --build-arg VERSION=X.X.X -f ./deployment/docker/Dockerfile .

to create a container, where X.X.X is the current version of the app.