Job repository
A job repository tracks the job status and execution information, and includes checkpoints and application persistent data that can help start a failed job. We can scale your batch environment by configuring multiple servers to use a single repository.
The job repository is a set of tables in a database used by the Liberty batch code to track the status of jobs. It tracks the number of jobs that are submitted, the location where the jobs run, the current state of the jobs, and the results of each job.
To define the scope of a Liberty batch environment, we can configure multiple servers to use the same job repository database. This method enables us to scale a batch environment to leverage multiple servers. The Liberty batch implementation provides the flexibility to determine whether and when multiple servers that use batch can share the same job repository.
Many of the batch functions are scoped to a single repository and apply to an environment that consists of all the servers that use the same job repository.
See Java batch persistence configuration.
We can also use an in-memory job repository that does not provide persistence, and is scoped only to a single server.