replication-health-part-1.aspx.
You can also see this health information with PowerShell by using
Measure_VMReplication. As shown here, the output contains the same basic
information as in the replication health graphical interface. If you need a detailed level
of monitoring for Hyper-V Replica, one option is to use System Center Operations
Manager and the Hyper-V Management Pack, which monitors the health of Hyper-V
Replica and many other Hyper-V aspects.
PS C:> Measure‐VMReplication ‐ComputerName savdalhv21 ‐VMName savdalapp01 |
fl
VMName : savdalapp01
ReplicationState : Replicating
ReplicationHealth : Normal
LastReplicationTime : 12/10
/2013 9:30:17 AM
PendingReplicationSize : 4096
AverageReplicationLatency : 00:00:00
SuccessfulReplicationCount: 223
MissedReplicationCount : 0
If a virtual machine is the replica virtual machine and also has its own extended
replica, Measure-VMReplication will return information about its relationship with the
primary and extended replica. If you want to view information about only one of the
relationships, add -ReplicationRelationshipType and specify either Simple (to view
the primary relationship) or Extended (to see the extended relationship).
You can see the HRL files used by Hyper-V Replica if you look at the folder containing
the VHD files being replicated. It is possible to pause and resume replication, but be
careful to not pause replication for too long because this will cause the log files to
build up. You can also pause and resume the replication by using the Suspend-
VMReplication and Resume-VMReplication cmdlets. Once again, you can specify
ReplicationRelationshipType for a virtual machine that is the replica of the primary
and the source for the extended replica (in other words, the one in the middle).
This can also happen if a network break occurs between the primary and replica that
stops transmission of the log files. If the size of the log files is greater than 50 percent
of the size of the VHD file, a resynchronization is required. Resynchronization
performs a block-by-block comparison of the source of the replica, with only different
blocks being sent over the network. This is deemed more efficient than sending over
the accumulated log files, although there is a performance impact during a
resynchronization. Other scenarios can force a resynchronization, but they are rare.
See
http://blogs.technet.com/b/virtualization/archive/2013/05/10/resynchronization-of-
virtual-machines-in-hyper-v-replica.aspx, which is a great blog post and worth reading
for more information on the exact resynchronization process.