In SAP HANA 2.0 SPS 00, the dynamic tiering option supports HANA system replication to protect your production system against unexpected downtime due to disasters.
What is system replication and how does it work? System replication is a mirroring configuration where you set up a secondary system that is an exact copy of the main, or primary, system.
The primary system sends redo log buffers to the secondary system, where they are persisted and replayed to build a shadow database.
The secondary system is passive until the primary system becomes unavailable and you perform a takeover. At this point, the secondary system becomes a standalone server using the data from the redo log buffers.
And the benefit? System replication supports rapid failover for planned downtime or to address local faults or storage corruption.
For a thorough background in HANA system replication, see the resources in blog post HANA System Replication – Take-over process by Frank Bannert.
You’ll need to be aware that there are some behavior differences when your landscape includes dynamic tiering. For example:
- You can only enable 2-tier synchronous mode replication,. In 2-tier replication, there is only one secondary system per primary system. Synchronous mode means that the primary system waits for confirmation that the log is persisted in the secondary before committing a transaction.
- Failback always requires full initialization (full data shipping) for the dynamic tiering service. The landscape is not available for takeover (not replication safe) until initialization is complete for all services.
- Some downtime is required for software upgrade.
See SAP Note 2356851 for a complete description of behavior differences between DT and HANA and a list of supported configuration parameters.
Want to learn more? Check this blog in March for a link to a video tutorial for setting up system replication and performing takeover in a landscape with dynamic tiering.