Single-Node Scenario Where HA Is Required

Figure 1 and Figure 2 show the single-node scenario where HA is required.

In the single-node scenario where HA is required, active/standby switchovers can be manually performed, or automatically performed using scripts.

Figure 1 Single-node HA deployment within an AZ

Figure 2 Single-node HA deployment across AZs

The concepts involved in this scenario are the same as those involved in Single-Node Scenario Where HA Is Not Required. The differences between the two deployment modes are as follows: Two independent SAP HANA nodes in the same AZ or different AZs are connected to the SAP HANA Studio for management. The two SAP HANA nodes synchronize data and implement HA using System Replication.

Figure 3 shows the deployment flowchart in the single-node scenario where HA is required.

Figure 3 Deployment flowchart in the single-node scenario where HA is required

Compared with the mode in which HA is not required in the single-node scenario, the mode in which HA is required has the following requirements:

• You must create two servers and synchronize time, format disks, and install the SAP HANA software on them.
• The two HANA ECSs must connect to SAP HANA Studio.
• You must manually configure HA on the two SAP HANA nodes using System Replication.
• You must configure scripts for the two SAP HANA nodes for active/standby switchovers.

  This only applies to SAP HANA nodes running the OS SUSE Linux Enterprise Server 12 SP1 for SAP or later.

• In the single-node scenario where SAP HANA nodes are deployed across AZs for HA, three ECSs are required. To support the Split Brain Detection (SBD) function, each ECS is bound to a SCSI disk and iSCSI configuration is required for SBD. For details, see Configuring iSCSI (Cross-AZ HA Deployment).