The ultimate Database Performance Booster



Scale on lower costs

The operational and licensing costs for high available Oracle database systems as well as the acquisition and maintenance costs of the powerful hardware running it are quite considerable. During the evaluation of more cost-effective alternatives, we developed an engineered system based on an open architecture that could outperform the classic Big Player Systems( e.g. Power8 and IBM® Storage, Oracle® Exadata X5) with significantly lower licensing and hardware costs.



Thanks to our partnership with Violin Memory® , our Performance Booster engineered by [bytesource] can convince with impressive performance figures. The use of especially  powerful Intel CPU cores and the optimal utilization of the available processing power, the system can perform the same or higher amount of tasks with significantly fewer CPU cores. This leads to a noticeable reduction of the Oracle® license costs.

               Comparing pays off!

Performance Booster Oracle® Exadata
Always the fastes Intel CPUs, low number of cores (e.g. 33% more power per core) Fast Intel CPUs, many Cores
All Flash Array with 1 mio. IOPS Flash caches, low cost but slow SATA Disks or cost intesive SSDs, additional complexity
by introducing database software on the storage nodes
4k Redo Logs 512 Redo Logs
Bandwidth up to 4,6GB/s write/read/mixed pro Array Bandwidth appr. 3GB(1/8) EF, appr. 300MB/s HC (when hitting the disks)
Latency < 0,5ms (end-to-end sustainable) Latency < 0,5ms EF, 10ms HC
Hardware compression (LZO 2-4 fach) for all tables HCC (only static tables, up to 30 time)
Hardware failure: low system complexity due to redundant standard components, low maintenace efforts. Hardware failure:: Higher system complexity and higher maintenace efforts.
Backup/Recovery: 10TB/hour. Backup/Recovery: 4TB/hour.
Scaling performance by just adding additional database nodes, more powerful CPUs by exchanging database nodes only and additional storage by just adding an additional Violin box. Reusability of standard components. Limited flexibility through gradual expansion of computing nodes and storage cells, higher upgrade costs for more powerful CPUs and limited reusability of propriatery hardware components.