Yesterday VMware published a paper focused on VMware vMSC (vSphere Metro Storage Cluster), a new configuration within the VMware Hardware Compatibility List intended for environments where disaster/downtime avoidance is a key requirement.
vMSC infrastructures are implemented with the goal of bringing the same benefits of high-availability clusters in local sites, but in a geographically distributed model with two data centers in different locations.
The document, written for a technical audience, describes how vSphere handles specific failure scenarios in environments where the distance between datacenters is limited, often metropolitan or campus environments.
The presented failure scenarios can be summarized as follows:
- Single-Host Failure in Frimley Datacenter
- Single-Host Isolation in Frimley Datacenter
- Storage Partition
- Datacenter Partition
- Disk Shelf Failure in Frimley Datacenter
- Full Storage Failure in Frimley Datacenter
- Permanent Device Loss
- Full Compute Failure in Frimley Datacenter
- Loss of Frimley Datacenter