Spring Data Cassandra
- 4. Rationale • Reduces ramp up time and code required to work with Cassandra • Clear, annotation based mappings for java objects to Cassandra tables • Good fit if already using spring data / spring boot • Minimal property based configuration
- 5. Features • Uses common spring data repository pattern • Synchronous and Asynchronous requests via the DataStax Java Driver • XML based CQL schema creation • JavaConfig / XML support via properties • QueryBuilder interface for complex query creation.
- 6. Annotated POJOs • Allows easy mapping of annotated POJOs to Cassandra tables. @Table("users") public class User { @PrimaryKey("user_id") private String userId; @Column("first_name") private String firstName; @Column("last_name") private String lastName; private String password;
- 7. Spring Repository Support • Basic Support of Spring Data Repository pattern. • Provides CRUD operations without code, findOne, findAll, save (upsert), delete. • Easily extensible with custom queries. public interface UserRepository extends CassandraRepository<User> { @Query("select * from users where user_id = ?0") User findByUserId(String userId); }
- 8. Primary Keys • Easily define simple, compound or composite primary keys. Allows for narrow or wide rows. • Simple Key example: @PrimaryKey("user_id") private String userId; • Compound Key with Cluster Column Example: @PrimaryKeyColumn(name = "eid", ordinal = 0, type = PrimaryKeyType.PARTITIONED) private Integer eid; @PrimaryKeyColumn(name = "day", ordinal = 1, type = PrimaryKeyType.PARTITIONED) private Integer day; @PrimaryKeyColumn(name = "event_id", ordinal = 2, type = PrimaryKeyType.CLUSTERED) private UUID eventId;
- 9. Query Support • No support for spring data “findBy” style queries at present, e.g. findByNameLikeOrderBy • Use @Query annotation for custom queries @Query("select * from events_by_tag where eid = ?0 and tag = ?1") List<Event> findByEidAndTag(Integer eid, String tag);
- 10. Query Support • Use static BasicMapId.id() to construct primary key based queries. MapId id = id() .with("eid", 1) .with("day", 12345) .with("eventId”, “XXXX-XXXX-XXXX”); return eventsByDayRepository.findOne(id);
- 11. Or use Java Driver QueryBuilder • Use QueryBuilder to create queries including batch and transaction based queries. Select select = QueryBuilder .select() .all() .from(“events_by_day”) .where(eq("eid", 1)) .and(eq("day", 12345)) .and(eq("eventId", “XXXX-XXXX-XXXX”)); return cassandraOperations .select (select, Event.class);
- 12. DEMO