Administrator Guide

VMware vSphere and Live Volume
54 Dell EMC SC Series: Synchronous Replication and Live Volume | CML1064
In this configuration, the MPIO policy for the primary Live Volume can be configured as either Round Robin
(preferred) or Fixed. Enabling Live Volume automatic role swap would be optimal in this configuration where
vMotion is used to migrate virtual machines between sites (for example, from Site A to Site B). After the
vMotion occurs, read and write I/O stemming from VMware Host B to the secondary Live Volume is proxied to
the Primary Live Volume controller through the asynchronous or synchronous replication link(s) between the
arrays. Automatic role swap allows SC Series storage to optimize the I/O after vMotion so that all or the
majority is flowing through the primary Live Volume.
Multiple sites and stretched clusters may also be configured with uniform storage presentation. For more
information on vMSC storage presentation, see section 8.9.
8.6 VMware vMotion and Live Volume
In a non-uniform storage presentation model, one method of controlling the location of the primary Live
Volume is to use vMotion and configure Live Volume automatic role swap (see Figure 46). vSphere Host A is
mapped to SC Series A and vSphere Host B is mapped to SC Series B. When a virtual machine running on a
Live Volume is migrated from Host A to Host B, Live Volume with automatic role swap observes that the
storage access has shifted from SC Series A to SC Series B. The SC Series array automatically swaps the
Secondary Live Volume to become the Primary Live Volume. Once this occurs, the virtual machine disk I/O
traverses a local path to the Primary Live Volume on SC Series B (using the Fixed or Round Robin MPIO
policy) instead of going through the Secondary Live Volume proxy across the replication link. The result
evades an inevitably higher cost in terms of increased latency. For stretched clusters or data centers,
consider the vMotion and Metro vMotion latency requirements between hosts. vMotion requires a round-trip
latency of 5 ms or less between hosts on the vMotion network. vSphere Metro Storage Cluster boosts the
allowable round-trip latency from 5 ms to 10 ms on the vMotion network between sites.
8.7 vSphere Metro Storage Cluster
vSphere Metro Storage Cluster (vMSC) is a specific VMware-certified design and is an implementation of
vSphere, compute, storage, and network infrastructure. The design is typically a stretched cluster
configuration over varying distances to provide high availability for virtualized workloads. This is achieved
through continuous availability of compute and storage resources provided by a combination of hardware
redundancy and automatic failover capabilities built into the vSphere hypervisor and the SC Series Live
Volume architecture. With the configuration of the Live Volume automatic failover feature, and by meeting
other solution requirements, SC Series storage is certified by VMware in supporting vSphere Metro Storage
Cluster in uniform, non-uniform, Fibre Channel, and iSCSI designs.