Why change a datastorage architecture?
The traditional datastorage server architecture – as most of you are familiar with – is certainly not the perfect fit and hardly optimized for distributed storage.
Traditional datastorage often results in network traffic congestion and overall low CPU utilization. Using virtual machines or multi-daemons introduces the risk of multiple lost storage devices when a single server fails. A problem besides the extra virtualization software costs and hypervisor system resources overhead. Ambedded Technology’s award winning ARM-based datastorage microservers available in BeNeLux / Europe now, is the answer to many if not all of these problems. Based on a game changing SDS architecture. Designed with distributed storage and optimizing Ceph performance in mind.
The industry exclusive 1-to-1 cluster architecture is pairing every single storage device (either HDD or SSD) to its own dedicated low energy nanoserver. This hardware architecture in combination with the right SDS software, provides great benefits compared to the existing traditional storage obstacles.
Nanoservers? So, performance will be the problem?
No, on the contrary. The Ambedded MARS-series storage solutions will not sacrifice performance once the data storage grows exponentially. (Which is a great solution for well known issues of traditional datastorage server architectures)
In fact the performance will show a linear increase, the more storage devices will be added. A true Scale Out Unified Virtual Storage solution.
OK, but when disaster strikes?
And – as you know – disk disaster will strike. You just never know when. You will be surprised (in a positive way that is). Because only system warnings will tell you something went wrong with a disk in the cluster-system. Not the users complaining about data loss. Not the users suffering from sudden significant performance degradation. And the best news is: the self-healing capacities of the Ambedded CEPH storage solution are much faster than traditional Disk Array technologies. Repair happens without any technician needed to replace a disk on location. The system will continue to operate without data-loss. The replacement of defective disk(s) can be postponed to the next planned maintenance window. A datastorage system lowering operational technicians costs once again.
Meer weten? Dat kan!