Administrator Guide

Live Volume support for Microsoft Windows/Hyper-V
75 Dell EMC SC Series: Synchronous Replication and Live Volume | CML1064
9.7 Hyper-V and Live Volume
The full Live Volume feature set works well with Microsoft Hyper-V in both clustered and non-clustered
environments. These Live Volume features include support for synchronous and asynchronous replication,
load balancing, disaster avoidance, managed replication, pre-defined DR plans, and other features covered in
this document.
Note: Live Volume automatic failover (LV-AFO) supports host and guest clusters running Windows Server
2012 or newer, or Hyper-V 2012 or newer. More on LV-AFO is covered in section 9.9.
9.7.1 VM live migration and Live Volume automatic role swap
With clustered Hyper-V servers running Windows Server 2008 R2 and newer, live migration of virtual
machines to another node can trigger a Live Volume automatic role swap. In order for automatic role swap to
work (using a two-node Hyper-V cluster as an example), each node is mapped only to the SC Series array
that is local to it. Therefore, when a VM workload is live migrated to the second node, the I/O will follow, and
for a brief time, the I/O be proxied over secondary data paths until the auto role swap thresholds are
exceeded. As noted previously, the Live Volume automatic role swap feature will not work when using MPIO
Round Robin (all paths optimal) because the minimum I/O thresholds that trigger a role swap will not be
exceeded.
9.7.2 Single-site configuration
In a single-site configuration, multiple Hyper-V servers can be mapped to both SC Series arrays for uniform
server mappings. If latency and bandwidth are not a concern over secondary data paths, Round Robin is
typically the best (and simplest) MPIO policy choice. If latency and bandwidth are a concern to non-optimal
paths, configure the Live Volume to report non-optimized paths (SCOS 7.3 and newer) and leverage Live
Volume ALUA if the OS fully supports Live Volume ALUA. This is the default setting for new Live Volumes
configured with SCOS 7.3. Failover Only can also be used to limit I/O to a preferred optimized data path, but
is typically more complicated to set up and maintain than Round Robin.
9.7.3 Multi-site configuration
In a multi-site stretch-cluster configuration, typically the individual Hyper-V hosts are mapped only to the SC
Series array at that particular site, for a non-uniform mapping configuration. In this scenario, the best MPIO
policy would typically be Round Robin for each Hyper-V host if the host does not fully support Live Volume
ALUA (unpatched versions of Windows Server 2016 and prior). If the host OS fully supports Live Volume
ALUA (Windows Server 2016 with the March 2018 cumulative update applied and newer), Round Robin with
Subset is a better choice (with Report Non-optimized paths enabled for each Live Volume).