The code provided in this project is to be regarded as being in a beta state.

The jmasar-service implements the MAchine Save And Restore service as a collection of REST endpoints. These can be used by clients to create and manage configurations and snapshots.

The service depends on the jmasar-model artifact (separate git repository). Java-based clients should also make use of jmasar-model to facilitate marshalling/unmarshalling of data.

Features:

• The service supports a tree structure of objects, much like a file system tree. Tree node objects are either "folders" or configurations.

• Nodes in the tree can be renamed, moved or deleted.

• Snapshots are associated with configurations and are not treated as nodes in the tree.

• The service is built upon Spring Boot and depends on some persistence implementation. In its current version, persistence is implemented against some RDB engine, using Spring's JdbcTemplate to manage SQL queries.

• The service has been verified on Postgres 9.6 and Mysql 8.0, on Mac OS. Database connection parameters are found in src/main/resources. To select a database engine, add -Ddbengine=postgresql or -Ddbengine=mysql to the command line when launching the service.

• Flyway scripts for Postgres and Mysql are provided to set up the database. Flyway is run as part of the application startup, i.e. there is no need to run Flyway scripts manually.

• Unit tests rely on the H2 in-memory database and are hence independent of any external database engine. Note that Flyway scripts for the H2 database are found in src/test/resources/db/migration. Running the unit tests will create the H2 "database file" (h2.db.mv.db) in a folder named db relative to the current directory.

• Swagger endpoint is enabled by default. This may not be desirable as PUT, POST and DELETE operations are not protected.

Missing features:

• Security in terms of authentication and authorization.

• Search for configurations or snapshots. This will be added next.

• JPA as persistence framework, which would make transition to database engines other than Postgresql or Mysql easier, and also hide some of the issues with differences in SQL dialects.

Build and run:

The project was developed on Spring STS (an Eclipse IDE clone).

Being a Spring Boot application, one may build from Maven command line and then launch the jar artifact assembled by the build system. The following parameters (environment variables) must be set on the command line:

• dbengine, must be postgresql or mysql.

• spring.datasource.username, the username for the DB engine connection.

• spring.datasource.password, the password for the DB engine connection.