Peer Fusion Hybrid/Private Cloud Solutions

Peer Fusion clusters are designed for the massive scaling of capacity and performance with simple administration:

• Multi-dimensional scaling.

A Peer Fusion cluster scales in several ways:

• By adding peers to a cluster

You are not just adding disk drives, your are also adding network bandwidth, computational power, memory and cache to your cluster. As the capacity of a Peer Fusion cluster grows so does its performance because data is striped across every peer so that I/O is parallel at all times. The more peers in your cluster the more parallel the I/O can be which results in sustained higher IOPS as well as raw throughput. The more peers in your cluster the more faults it can withstand so the better protected your data.

• By upgrading peers to gateways to sub-clusters.

Each peer in a cluster can become a gateway to another cluster. Such sub

• By aggregating clusters under the aegis of a supra-cluster.

Independent clusters can be grouped by the thousands into a supra-cluster.



• Simplicity retained.

Peer Fusion clusters retain their simple administration as they scale. The peers are designed to work efficiently whether there are ten or hundreds. Once a peer performs network discovery and joins a cluster it becomes part of the quorum and operates like all the other peers. Your administrative tasks that are applied to the cluster will apply to all the peers no matter how many. The Peer Fusion cluster is a single namespace and a single virtualized repository.



• Cost-efficiency scales.

Peer Fusion clusters benefit from a resiliency virtuous cycle as they scale. The count of recoverable peer failures need not grow linearly with the count of peers added. For example if your target was to continue operating your ten-peer cluster after three peer failures, it may not be necessary to configure your scaled-up twenty-peer cluster to operate after six peer failures. The fault-tolerance requirements depend upon your applications and only you can weight the possibility that seven peers will fail simultaneously.



• Repurposing Previous Generation Servers.

For extra cost savings consider repurposing previous generation servers as you select the commodity hardware for your Peer Fusion cluster. This is an excellent solution when acquiring the greatest hardware performance is not a critical requirement. Repurposing servers is good for the environment and good for your budget!.