High-performance and future-proof
Scalability pays off
Your benefits thanks to scalability
Powerful and reliable
pirobase guarantees continuous performance – even under increasing loads. It is therefore able to run without loss of performance even as data volumes and user numbers increase. The workload that can be managed by our highly scalable system runs linear to the volume of resources required. pirobase can therefore remain consistently stable while responding to twice the number of requests per time unit with twice the resources, for example.
Efficient and cost-effective
A linearly scalable system not only offers consistently stable performance under extremely high access and user numbers but also has considerable economic advantages. All cost factors can be accurately planned and optimised from the hardware to the software and the maintenance and service costs. This ensures that the investment framework is lean and concise from the beginning.
Calculable and reliable
pirobase makes the linear scalability of capacity and performance possible. This creates a system with unlimited scalability which is rendered suitable for large-scale, complex tasks by its program structure and is designed to be constantly expandable. This allows our system to offer unlimited options to meet even increased future demands.
Why scaling?
A definition
The term 'scalability' describes how well a system can process increasing work loads by adding hardware resources. In an ideal case, the workload managed by a scalable system will increase linearly with the volume of resources available. Twice the resources therefore means twice the performance.
Vertical versus horizontal scalability
Vertical scaling is an expansion of the hardware of a system to increase performance by adding components such as memory or additional CPUs. Horizontal scalability means that new nodes are added to the existing nodes to increase performance. A node is a linking point for transfer routes within a network. The individual nodes are clustered to form a single unit. Each node works independently, but remains in constant communication with the other nodes.
Performance under load
To prove the scalability of pirobase, we looked for a strong partner. Our system was subject to a load test lasting several weeks at the IBM Innovation Center. During the course of this process, we verified scalability in various test series. Our system was subjected to extremely high loads by constantly increasing the number of website users and authors every second - the results speak for themselves.
Hardware for the load test
The database (Oracle) functions on an LPar with 64 GB main memory and four cores on Power7 running AIX. An LPar with 64 GB permanently allocated main memory was available for the WebSphere application server on the same Power7 under AIS in which the pirobase CMS® was deployed. During vertical scaling, the number of allocated cores varied from one to 16 cores. Blades that were each equipped with an Intel Dual QuadCore processor with activated hyperthreading were used for horizontal scaling. The pirobase® CMS instances installed under Linux were configured as clusters addressed via a software load balancer.



