SAP HANA system replication

Overview

Usually, you use system replication to support high availability and disaster recovery. SAP HANA® system replication lets you copy and continuously synchronize a SAP HANA database to a secondary location on the same or another data center.

This mechanism ensures the high availability of your SAP HANA system. System replication is the SAP recommended configuration for addressing SAP HANA outage reduction due to planned maintenance, faults, and disasters. It supports a recovery point objective (RPO) of 0 seconds and a recovery time objective (RTO) measured in minutes.

How it works

After you enable SAP HANA system replication, each server process on the secondary system establishes a connection with its primary system counterpart and requests a snapshot of the data. From there, SAP HANA continuously replicates all logged changes in the primary system. At the same time that the system writes logs to the log volumes of each service, persisting the logs, it also sends them to the secondary system.

The primary system commits a transaction after replicating the redo logs. This process depends on the replication mode that you selected when you set up the system replication.

 Image Source

Replication modes and operation modes play a crucial role when configuring HANA system replication. You need to select these modes depending on your requirements.

Replication modes

While registering the secondary system, you need to decide which replication mode to use. These modes generally determine how logs replicate to the secondary system.

SAP HANA offers the following replication modes:

Synchronous in-memory (SYNCMEM)

The primary system commits the transaction only after receiving a reply that the secondary system received the log and stored it in the memory. The transaction delay in the primary system is shorter because it only includes data transmission time.

Data loss can occur in the following situations:

• When the primary and secondary systems fail at the same time while the secondary system is connected, data is not written to either the primary or secondary system disk.
• When a takeover occurs while the secondary system is unavailable. The data that arrived on the secondary is outdated compared to the data on the primary.

This option is vulnerable to data loss, but it provides better performance because it doesn't need to wait for the secondary system disk I/O.

Synchronous on disk (SYNC)

The primary system waits to commit the transaction until it gets a reply that the log is persisted on the secondary system disk. When the connection to the secondary system is lost, the primary system continues to process transactions and writes the changes to only the local disk.

If the secondary system is connected, you won't lose any data. However, data loss can occur when a takeover occurs while the secondary system is disconnected.

Synchronous full sync

Synchronous replication mode can run with a full sync option. The log write succeeds when the system writes the log buffer to the log file of both the primary and secondary systems. You can set the full sync option with the [system_replication]/enable_full_sync parameter.

When the secondary system is disconnected (for example, because of network failure), the primary system suspends the transaction processing until it reestablishes the connection to the secondary system. No data loss occurs in this scenario.

Asynchronous (ASYNC)

The primary system commits the transaction after sending the log without waiting for a response. This process causes no delay because the data transmission is asynchronous to the transaction in the primary system.

Because you don't need to wait for the log I/O on the secondary system, this method offers better performance. The process guarantees database consistency across all services on the secondary system. However, it is more vulnerable to data loss, you might lose data changes during a takeover.

Replication mode behavior

The following table gives a brief idea about each replication mode's behavior when the secondary system is not available: 

Operation Modes 

While registering the secondary system, you need to decide which operation mode to use with SAP HANA system replication. Depending on the configured operation mode, the database sends different types of data packages to the secondary system.

Choose from the following operation modes:

delta_datashipping

This mode performs a delta data shipping occasionally (default: every 10 minutes) in addition to the continuous log shipping. The system does not replay the shipped redo log on the secondary site. During a takeover, the system replays the redo log up to the latest delta data shipment.

The following image shows this traffic on the transportation channel between the primary and secondary systems for the delta_datashipping operation mode:
 

Image Source 

logreplay

In this operation mode, the system does a redo log shipping after you initially configure system replication with one full data shipping. Then, the system continuously replays the redo log on the secondary site immediately after arrival. During a takeover, this replay process makes this step unnecessary. The logreplay mode does not require delta data shipping, which helps to reduce the amount of data transferred to the secondary system.

 logreplay_readacess

You need to use this mode for replication to an Active/Active, read-enabled, secondary system. It is similar to the logreplay operation mode in the following ways:

• It has continuous log shipping.
• It uses the redo log replay on the secondary system.
• It requires an initial full data shipping and a takeover.

This mode reduces the load on the primary system because it allows read-only access from the secondary system.

The following figure shows this traffic on the transportation channel between the primary and the secondary system for the logreplay and logreplay_readaccess operation modes:

Image Source

 Configure SAP HANA replication

You can configure SAP HANA system replication by using the following tools:

- SAP HANA cockpit

- SAP HANA studio

- `Hdbnsutil` (a command-line tool)

Perform the following steps to configure replication by using SAP HANA studio:

 Step one: Enable system replication

1. Start the primary system.

2. Perform an initial data backup or storage snapshot on the primary system.

3. Enable system replication on the primary system by using the following steps
   in SAP HANA studio:

     a. In the Systems view, right-click the primary system and choose
     Configuration and Monitoring -> Configure System Replication.

     b. The Configure System Replication dialog opens
     and selects the Enable System Replication option by default. Click Next

     c. Enter the logical name used to represent the primary system and click Next.

     d. Review the configured information and choose Finish.

     e. To stop the secondary system, right-click on the secondary system and choose Configuration and Monitoring -> Stop System.

4. To prepare the secondary system for authentication, copy the system PKI SSFS    4 and the .dat file from the primary system to the secondary system.
   For more information, see [SAP Note 2369981

5. Use the following steps to register the secondary system with the primary    system in SAP HANA Studio:

     a. To stop the secondary system if it is still running, right-click the
     secondary system and choose Configuration and Monitoring ->Stop System.

     b. In the Systems view, right-click the secondary system and choose
     Configuration and Monitoring -> Configure System Replication.

     c. In the Configure System Replication dialog, choose
     Register Secondary System and click Next.

     d. Enter the system information and the logical name for the secondary system.

     e. Specify the log replication mode and operation mode.

     f. Review the configured information and click Finish.

6. The secondary system requests an initial full data replica from the primary system.

The system replication is enabled and the secondary system is registered with a
primary system now

Step two:  Disable system replication

1. Use the following steps to unregister the secondary system in SAP HANA studio:

     a. In the Systems view, right-click the primary system and choose
     Configuration and Monitoring -> Configure System Replication.

     b. In the Configure System Replication dialog, choose
     Unregister secondary system and click Next.

     c. Enter the required system information and click Next.

     d. Review the configured information and click Finish.

 

2. Use the following steps to disable replication system on the primary system:

     a. In the Systems view, right-click the primary system and choose
     Configuration and Monitoring -> Configure System Replication.

     b. Choose Disable system replication and click Next.

     c. Review the configured information and click Finish.

Step three: Takeover the secondary system

The takeover process switches your active system from the current primary system to the secondary system. If you perform a takeover as part of a planned downtime, you should first make sure that the primary system has completely stopped before
performing a takeover to the secondary system. 

1. Perform a takeover in the secondary system at data center B by using the following steps in SAP HANA studio:

     a. In the Systems view, right-click the secondary system and choose
     Configuration and Monitoring -> Configure System Replication.

     b. Choose Perform Takeover from the actions list.

     c. Enter the required system information and click Next.

     d. Review the information and click Finish.

 

2. At this point, the Secondary system becomes the production system. If the system is running, it leaves the recovery mode and becomes fully operational, replaying the latest transaction logs and accepting queries. If the system is offline, start it, and it will take over production operations.

Step four: Failback to the original primary system

1. To failback, you must attach your former primary system as the new secondary  system to the current primary system by using the following steps in SAP HANA   studio:

     a. In the Systems view, right-click the primary system and choose
     Configuration and Monitoring -> Configure System Replication.

     b. In the Configure System Replication dialog, choose
     Register for Secondary System and click Next

     c. Enter the system information and the logical name for the system.

     d. Specify log replication mode and operation mode as required.

     e. Review the configured information and click Finish.

2. The former primary system is now registered as the secondary system with the current primary system, formerly the secondary system. Because the system cannot use the data in the former primary system, it carries out a complete
   initialization. A full data replication runs until the previous primary system is entirely in sync.

3. Verify that the secondary system replication status in All Services is active and in sync. You can see this status in the Administration editor on the Overview tab.
 

Conclusion

By using system replication, you can ensure high availability and enable disaster recovery for your SAP HANA system. Achieve this by continuously replicating the data from a primary to a secondary system, including in-memory loading. System replication facilitates rapid failover in the event of a disaster. The main advantage of SAP HANA replication is a zero-second RPO and an RTO of mere minutes.