The announcement by SAP that it will only support its own database from 2025 onwards means that enterprises with databases from other manufacturers will have to migrate to S/4HANA. In addition to the new infrastructure a new concept for high availability will have to be put in place. If for example a major trade chain is no longer able to issue delivery notes and is therefore unable to load its lorries, the financial losses within a short space of time will be considerable.
Just pause to think about your business
What would it mean if the critical SAP services were no longer available? Many business processes are affected and large financial losses are incurred. At the very least a standstill would damage your business’s image.
Necessary migration to a SAP database
The ERP solution SAP HANA is the critical platform for any business since it handles all relevant processes such as merchandise management, CRM or accounting.
SAP has announced that from 2025 onwards it will only support its own database S/4HANA. Support for all other solutions will be terminated. This means:
- If you are a SAP customer and use a database from another provider such as Oracle or MS/SQL you should begin thinking about migration as soon as possible.
- The second step you must undertake is the development of a new concept ensuring the high availability of the infrastructure – here you will find out why this is important.
… requires a new concept for high availability
S/4HANA is a Linux-based database. HANA employs system replication, which means that data is not mirrored in the storage site but directly within the HANA database or application, irrespective of the underlying infrastructure.
For this to work, cluster technology is required. In SAP the technology used is a Linux Cluster, which SAP customers must operate for this solution. The cluster services of a Linux Cluster for data replication have certain drawbacks: laborious setup, complex management resulting in higher error rates and opaque functions.
In order to continue to be able to guarantee the high availability of business-critical data and applications you must implement an automatic switchover. To do this you need a third site and suitable connection lines – all of which make your infrastructure more expensive and more complex.
Excursus cluster technology: the challenge posed by power outages and split-brain syndrome
In all cluster systems there are situations in which the central system malfunctions and business-critical IT services no longer work. Each Linux Cluster set up across two sites must be able to cope with ‘split-brain behaviour‘. This challenge posed by architecture affects each two-site cluster in equal measure and hence also Linux-based clusters.
Let us explain the technological reasons for this:
Cluster systems often work with synchronous data mirroring over two sites, so as to guarantee maximum availability. If hardware components malfunction, any applications running on them can continue to be operated at a second site.
In the event of a total outage, e.g. if there is a power failure the storage cluster is unable to determine whether only the interconnects between the sites have been interrupted or whether a total outage affecting the cluster partner has taken place. In order to exclude the risk of cluster inconsistency (split brain) no services from any other sites are taken over.
A split-brain syndrome must be avoided at all costs. In order to ensure that transactions are not processed more than once for different volumes which can no longer be consolidated later one and resulting in data losses, no automatic cluster takeover takes place when for example there is a power outage. The result: your business-critical SAP applications malfunction.
Possible relief in a SAP Linux environment could be provided by a third site, responsible for monitoring the other two sites. You can choose from two different solutions:
- Pacemaker by RedHat (open source)
- Service Guard by HPEnterprise (for which licences must be purchased)
The solution: ClusterLion for HANA guarantees high availability in Linux environments with just two sites
ClusterLion for SAP HANA was developed to implement automatic switchover and hence the highest degree of availability for SAP HANA environments.
ClusterLion for HANA is the only solution in the market offering the very highest degree of availability and ensuring data consistency at all times even in a split-brain scenario. You do not have to operate a third site or the connection lines associated therewith – this task is performed by ClusterLion for HANA with the provision of a Cloud Quorum service.
You only require Linux Single Server systems without cluster services. This reduces the complexity of your infrastructure.
With ClusterLion for HANA all storage services remain online, even if one of your sites is affected by a total outage.
- High availability for your business-critical SAP applications
- Permanent data consistency is guaranteed and a split-brain scenario is excluded
- The Cloud Quorum solution does not require a third site, thus saving you money
- The use of ClusterLion for HANA dispenses with the need for complex Linux clusters
- Single systems simplify management thereby reducing errors
- Simple monitoring on the status page including Notification Manager with push warnings to a defined mail distributor
- You receive transparent event logs, incorporation into your own superordinate system monitoring tool is performed easily via a standard API
- ClusterLion for Hana supports you both in emergencies and during regular software upgrades: the controlled switching between computing centres renders the task of regular updates simple and secure
- If you operate larger HANA environments, the solution supports you with multi-client enabled user management. For example, you can stipulate whether users should only be able to use CRM or only business warehouse functions
Once it has been set up ClusterLion for HANA permanently monitors all critical HANA services.
If services are suddenly affected by an outage, the solution immediately recognises this and reacts accordingly.
ClusterLion for HANA recognises which cluster nodes are still working and creates a consistent state in the cluster. In addition, there is no possibility of an unintentional restart of the cluster node.
The cluster services are then taken over on the functioning site by ClusterLion for HANA.
All services are now running securely and properly as before on the functioning site – with ClusterLion for HANA you ensure the high availability of your critical SAP applications.
ClusterLion for HANA is a software-only solution which can be implemented during operations at any time. The solution is up and running in around an hour. It is started up simply and with a minimum of effort. A graphical interface guides users step by step through the configuration process. The SAP HANA environment is automatically recognised and displayed in the portal.