Hana Data Center and Zero Downtimes

SAP customers can now use the new version 12 of Suse Linux Enterprise for SAP Applications, which, like the previous versions, already includes the so-called High Availability (HA) Extension including disaster recovery (DR) functionality from the outset.

And SAP-certified (currently together with x86 hardware) for both Any-DB and Hana-DB use. It is obvious that, in view of ever greater Hanaization, customers are also opting for SAP in-memory DB use together with SLES 12 for SAP Applications in the overwhelming majority of use cases.

Three innovations in particular are worth highlighting here, from which customers benefit when using Hana (be it when using Business Suite on Hana, S/4, BW on Hana and others; in fact, all Hana-based SAP solutions up to Business One powered by Hana).

On the one hand, new so-called Hana patterns are made available for further simplification/automation of Hana installations, up to and including complete automation of SAP solution-based systems across entire IT landscapes in accordance with an n-tier approach.

Which ultimately eliminates manual errors and thus minimizes possible downtimes (Zero Downtime support). Also new in Suse SLES 12 for SAP Application: the new technology developed by Suse called kGraft, a live kernel patching technology. It enables online updating of security patches without rebooting or waiting for the next scheduled service window.

Also worth mentioning are the so-called Hana Resource Agents (RA), which are delivered with the system. These can be used to manage, monitor and control Hana database instances and replications. The RAs can also be configured.

Suse supports easy-to-use cluster configurations with Hana Resource Agents. They run on all nodes of a SLES 12 HAE cluster and provide configuration information or system status of Hana systems and Hana replications.

For years, the Suse High Availability Extension (HAE) already included in SLES has provided a proven and leading high availability solution to improve business continuity/HA and disaster recovery capabilities for SAP solutions.

The solution was optimized and further developed especially for Hana use and represents a kind of standard for HA and DR in the Hana environment (previously for scale-up). SAP provides Hana System Replication (SR) mechanisms for HA in particular, which are generally handled and used manually.

Increasing the SLA via automation is possible with the Suse High Availability Solution (HAE) from SLES 12. System replication via memory preload in a cluster (e.g. 2-node cluster) is generally preferred as the central HA solution for Hana.

Automation of the switchover can be conveniently executed using HAE; e.g. in a 2-node cluster with a (second) synchronously running HA system for a Hana on-site system (would then be node one).

For a DR scenario, the Hana SR enables asynchronous replication to another site (e.g., another data center), so that this covers the disaster case.

A cost-optimized approach can be used as an alternative HA in the on- or main-site. Here, there is no memory preload on a second system, but a second node is available as a test and development system. Only in the HA case is the Hana SR performed on the second node.

Compared to an HA solution with existing memory preload, this logically takes longer; however, it also means an advantageous cost minimization in terms of investment/operating costs. In a DR scenario, replication to the secondary site is also asynchronous again.

Conclusion:

Overall, Hana Pattern, kGraft in conjunction with Suse HAE and in combination with Hana System Replication massively reduce downtime - and thus contribute to the realization of a targeted zero downtime in an SAP data center.