Leveraging vSAN for Highly Available Management Clusters : Solution Architecture Overview : 2.2 vSAN Conceptual Overview
   
2.2 vSAN Conceptual Overview
vSAN is a hyper-converged, software-defined storage solution offering enterprise class performance, reliability, and availability. vSAN is designed to take advantage of commodity servers with locally-attached storage (traditional spindle-based disks or high-performance flash devices) to create and maintain its storage abstraction layer (datastore), saving and manipulating data in the form of storage objects.
vSAN architecture is natively embedded in the hypervisor storage stack and provides complete integration with the vSphere management interface while supporting its advanced availability features:
VMware vSphere vMotion®
VMware vSphere Fault Tolerance (FT), providing continuous availability for virtual machines up to a limited size of 4 vCPUs and 64 GB RAM
VMware vSphere High Availability (HA), providing automatic restart of virtual machines on remaining hosts in case of failure of one of the cluster nodes
VMware vSphere Data Protection™, providing a combination of backup and restore functionalities for both virtual machines and applications
VMware vSphere Replication™, providing asynchronous virtual machine replication
vSAN supports two main configurations in respect to the underlying storage devices connected to the nodes:
Hybrid – A combination of magnetic disks used for the capacity tier, and flash devices used for the cache tier. Traditional spindle-based disks are supported in several formats and speeds, and are usually selected based on the capacity requirements of the environment.
All-flash optimized – Both capacity and cache are supported by flash-based devices (write-intensive/high-endurance and read-intensive/cost-effective SSD devices, depending on the associated tier they are supporting). vSAN using all-flash clusters is available with version 6.0 and later, and requires a 10-Gbps network between the nodes of the cluster.
The following figure shows a typical hybrid vSAN deployment, where the storage resources of the four hypervisors represented are joined together in the vSAN datastore for that cluster and made available to all the virtual machines running on the nodes.
Figure 2. vSphere and vSAN