Erasure coding vs. Replication
Pivot3’s use of erasure coding in its appliances offers substantial benefits in data protection, cost savings, storage efficiency and scalability compared to competing technologies using replication schemes
By Deni Connor, founding analyst
Earl Follis, senior analyst
When considering the purchase of hyper-converged infrastructure appliances, careful consideration of the accompanying storage technology is critical to making an informed decision that will withstand both the test of time and your next data loss event. The formula for evaluating the best data protection technology for your hyper-converged appliances — and the virtual machines residing on them — can be a complex mix of hardware and software costs, as well as the storage efficiency offered by erasure coding versus replication-based data protection.
Many emerging technology companies are choosing replication-based data protection because that technology is widely understood in the storage industry. Although, it’s relatively easy to incorporate in a hyper-converged appliance, it is relatively expensive to implement in hardware and software. Be sure to consider the true cost of your hyper-converged storage requirements, not just the initial costs, before choosing a vendor to provide your hyper-converged infrastructure. Considerations including the storage efficiency (how much of your total storage is actually available to the compute node), how many disks or nodes can simultaneously fail without losing any data, and the desired time to recover from a data loss event or drive failure.