Overview

• An Entry must be constructed using the ClientUIDHandler. To create the Entry UID before writing it into the space, the UID should correspond with the primary key – the first defined index field toString() value. To do this, perform one of the following:
  • Implement the __setEntryUid() and __getEntryUid() methods.
  • Extend your class from MetaDataEntry and provide the UID as part of the EntryInfo object, using the MetaDataEntry.__setEntryInfo method.
  • Implement the IMetaDataEntry interface.
• Versioning is not supported for data loaded from the database that was evicted when running in LRU cache policy.
• The Count operation using the data within the space and does not access the database to return the object count (different than Storage Adapter).
• Native serialization should be used.
• The following limitations exist for the Mirror Service:
  • Mirror – only one per cluster.
  • The Mirror service cannot be a clustered space. Deploying it as a Processing unit ensures its high-availability.
  • Memory cannot be recovered when using the Mirror Service. The Mirror does not load data back into the space.
  • The Mirror service cannot be used with a single space or the partitioned cluster schema. It is supported with the sync-replicated, async-replicated, and partitioned-sync2backup cluster schemas.
• engine.initial_load should be configured with a small number, to avoid memory shortage in persistent spaces with large data.
• Partitioned schema should not have the failover policy or the failover policy-type set to fail-to-backup.
• RegExQuery is not supported by the DataProvider or the SQLDataProvider.
• No Lease renewal or less different than Lease.FOREVER. Entries with Lease smaller than Lease.FOREVER will not be evicted. You can use only Lease.FOREVER.
• When using the Map API, key must be Serializable.
• The ExternalDataSource.initial_load() loads data into partitioned spaces by reading all the data into the space and filtering it at the space side. To tune this behavior you should execute relevant query from the database base on the partition ID to fetch relevant result set to load into the space.
• The ExternalDataSource.initial_load() should have the POJO/PONO/POCO RoutingField and SpaceID to be the same field. This limitation will be lifted with future versions.
• Hibernate Lazy load is not supported when using the HibernateDataSource as an External Data Source implementation.