borzaya - Fotolia
VMware Virtual SAN (VSAN) first arrived in vSphere 5.5. It allowed storage resources to be pooled and then provisioned according to established policies. The storage policies work to ensure constant storage performance regardless of changes in VM computing demands or availability changes in hardware and network components. VSANs are often clustered to further improve availability.
VSAN functionality was also improved with the release of vSphere 6. Some of the noteworthy improvements include scalability features. For example, vSphere 6 administrators can scale the VSAN cluster to 64 nodes in hybrid -- flash and magnetic disk -- and all-flash disk configurations. VSAN now supports virtual machine disk files to 64 TB in size -- up from 2 TB -- allowing VSAN 6.0 to support VMs for the largest enterprise applications. The number of VMs per host has doubled from 100 to 200 VMs per host, and the total number of VMs supported by the VSAN cluster moves from 3,200 to 6,400 total VMs. This provides a large margin for VM migration and failover within ESXi host clusters. A move to the new Virsto file system (VirstoFS) allows up to 9,000 components per host to support more complex deployments.
VSAN performance is also bolstered in vSphere 6. Hybrid disk VSAN performance can achieve up to 40,000 IOPS per host, while all-flash VSAN deployments can reach 90,000 IOPS per host. This boost is critical for VSANs with supporting greater scalability in the number of VMs, components and so on. The new VirstoFS disk format supports faster VM snapshots and cloning. Evacuations due to maintenance or faults could unbalance host workloads, but vSphere 6 adds workload balancing mechanisms designed to maintain performance in the face of workload evacuations. In addition, evacuations can be tailored to individual disks, allowing troubled disks to be checked and replaced rather than evacuating entire disk groups.
Dig Deeper on Selecting storage and hardware for VMware environments
Related Q&A from Stephen J. Bigelow
WET code leaves apps bogged down. Learn how to reduce the challenges brought on by code redundancy by programming based on the DRY principle. Continue Reading
A virtualization layer in an embedded system provides better efficiency for tasks such as network virtualization. Some examples of embedded ... Continue Reading
An embedded hypervisor offers several benefits, such as VM security, system reliability and improved hardware use, and is ideal for admins looking to... Continue Reading
Have a question for an expert?
Please add a title for your question
Get answers from a TechTarget expert on whatever's puzzling you.