Limitations of NAS

Gridstore Scale-out Storage

Fixed Storage Capacity

The main driver of storage sprawl. Once a NAS is full, another is added resulting in storage silos that multiply management time and single points of failure. Silos also create unbalanced usage with the last NAS bottle necked serving all the hot data while the others idle. Storage silos are high cost, high risk and perform poorly.

Unlimited, Incremental Capacity

Gridstore scale-out NAS allows you to incrementally grow your storage pool capacity one storage block at a time to any size you require – on the fly – no disruption to service. While management effort does not change as you add more blocks of capacity – every additional storage node adds more resiliency and bandwidth making the overall system more powerful.

Fixed Network Bandwidth

Network bandwidth limits the amount of data the system can transfer at one time. Once full, storage requests queue up and overall performance drops. Your only option to address this is to upgrade to higher performing storage servers.

Scalable - Parallel IO

Gridstore eliminates network bottlenecks by using parallel IO to uniformly spread the data across many storage nodes. Because data is evenly striped across storage nodes, the more nodes there are, the less work each one does and the more bandwidth there is to your storage.

Fixed Processing Capacity

All processing is done in the NAS storage controller. As the number of concurrent client requests grows, processing bottlenecks and performance quickly degrades. Processing cannot be increased to match user base grow or temporary spikes in demand. And with the limited processing, higher protection levels such as RAID 6 can perform very poorly.

Scalable - Zero Cost Processing

The Gridstore scale-out architecture seamlessly distributes virtualized storage controllers to every client that accesses the storage pool. We can keep the cost of storage nodes down by offloading storage processing from storage nodes. This also means your storage processing capacity grows at zero cost as your user base grows. Performance bottlenecks are eliminated because performance is always balanced to demand. Best of all, the performance boost is free.

Single Point of Failure

By definition, standalone storage is a single point of failure. Any component in a system can fail – and they do fail everyday. Any disk, network card, power supply or controller failure will result in business disruption and possibly catastrophic data loss.

For most small-medium sized businesses using standalone storage – if your storage stops, your business stops.

Fault Tolerant to Run NonStop.

Component failures are reality. Scale-out storage running on Gridstore technology is designed to handle and automatically recover from multiple concurrent failures (of entire nodes) without data loss or disruption. You simply dial in the number of node failures you want to withstand. This can even be set to levels higher than enterprise RAID 6 (which only covers disk failures – not the other components).

Run your business NonStop on Gridstore.