The ins and outs of SimpliVity

If you were thinking of using SimpliVity and weren’t sure what’s what, Lenny’s Likes just came to the rescue. Read the breakdown of what SimpliVity can do for you:

SimpliVity

SimpliVity provides full redundancy with as few as 2 nodes, and we can even offer customers single node configurations for non-mission-critical data applications. It’s what’s inside the node that counts. SimpliVity can package more VMs onto a single node than the competition, often running an entire data centre on just two nodes and 4U of space.

Data Locality

SimpliVity avoids excessive back-and-forth movement of data. Yes, consistent and good performance is essential. We disagree fundamentally with Nutanix about how to achieve it. When Nutanix moves data to follow a VM, they do it automatically, with no thought to existing workload on the system. Often, this leads to excessive back-and-forth movement of data, which saturates the network and slows overall performance. Our approach is to simply access data over the 10G network, until the appropriate time to rebalance nodes.

SimpliVity RAID approach

SimpliVity prioritizes data protection over RAID rebuild time. However, at the same time all VMs are still up, running, and protected as SimpliVity can tolerate double-HDD failures on every node. Nutanixcan only tolerate one disk failure – or node failure – across the entire cluster.

Scalability – Cluster Size + Workloads

SimpliVity scales out from small to large workloads, and requires fewer nodes to manage the same number of VMs. The number of nodes matters, but not in the way that others want you to think. What matters is the total workload capacity and performance of the system. SimpliVity scalability is unmatched in this regard, and achieves the results with vastly fewer nodes than our competitors. This not only reduces the capital cost, but also the management complexity and cost of operations.

SimpliVity scales out from small to large workloads:
SimpliVity’s architecture is built around a scalable building block (“node”) of CPU, RAM, and Disk. Scaling is by unit of node. Customers have the flexibility to start small –smaller than Nutanix–and scale the systems as they grow. Systems can be expanded without disruption of existing operations or downtime.
The smallest node today has a single 8-core socket, 96GB of memory, and 4TB physical of disk. The largest known installation today manages over 2PB of data.

Node Requirements

Customers are frequently surprised at how few nodes are actually required to meet performance and storage requirements because the unique SimpliVity Data Virtualization Platform provides native compression, deduplication and optimization of all data from inception, across all tiers, including flash, HDD.
As a result, customers report typical global resource savings of 50-100:1, so few customers require more than a few nodes to do what other hyperconverged systems require many times the resource to accomplish.

If something was not clear or you have any more questions about SimpliVity, you can contact us here.