Website under construction

Server-to-server and cluster-to-cluster replication 
Storage Replica can connect two individual servers—sometimes called standalone replication—and 
their volumes. This storage can be serially attached SCSI JBOD, iSCSI target, SAN, or even local DAS, 
such as SCSI drives attached to a local RAID controller. Storage Replica can also replicate between two 
clusters as if they were two servers, with any shared storage the cluster considers acceptable. 
Replication occurs between two physical locations, such as different buildings on the same campus or 
different metropolitan datacenters. The replicated storage must be either NTFS or ReFS volumes. 
Figure 2-57 depicts a typical architecture used for implementing server-to-server replication using 
Storage Replica. On the left is the Redmond site with server SR-SRV01 and some storage (SAN, JBOD, 
or iSCSI). On the right is the Bellevue site with server SR-SRV02 and more storage. You can use 
Storage Replica to combine the servers and storage at these two sites into a partnership by 
asymmetrically replicating storage from one site to the other. 
Figure 2-57: 
Typical architecture used for implementing server-to-server replication 
In the server-to-server and cluster-to-cluster scenario, there is no graphical interface and no 
automatic failover management—all administration is manual and human-driven through the 
Windows PowerShell StorageReplica module. To ensure ease of provisioning, Storage Replica 
implements a simple system for configuring replication with a single command when possible. 
The Windows PowerShell StorageReplica module contains the following commands in Windows 
Server 2016: 
Get-Command -Module StorageReplica | FT –Auto 
CommandType Name Version Source 
----------- ---- ------- ------ 
Function Get-SRGroup 1.0 StorageReplica 
Function Get-SRPartnership 1.0 StorageReplica 
Function New-SRGroup 1.0 StorageReplica 
Function New-SRPartnership 1.0 StorageReplica 
Function Remove-SRGroup 1.0 StorageReplica 
Function Remove-SRPartnership 1.0 StorageReplica 
Function Set-SRGroup 1.0 StorageReplica 
Function Set-SRPartnership 1.0 StorageReplica 
Function Suspend-SRGroup 1.0 StorageReplica 
Function Sync-SRGroup 1.0 StorageReplica 
Function Test-SRTopology 1.0 StorageReplica
Configuring replication is as simple as providing the following information: 
New-SRPartnership -SourceComputerName np-sr-srv05 -SourceRGName rg01
-SourceVolumeName g: -SourceLogVolumeName h: -DestinationComputerName np-sr-srv06
-DestinationRGName rg02 -DestinationVolumeName g: -DestinationLogVolumeName h:
-LogSizeInBytes 16GB 

---

53 
CHAPTER 2 | Software-defined datacenter 
There are many options for the New-SRPartnership cmdlet, including creating asynchronous 
replication. You also can create replication in a more granular fashion by running New-SRGroup on 
each server and tying them together by using New-SRPartnership. You can add additional volumes to 
a replication group by using Set-SRGroup, and you can run more than one replication group on a 
server at a time. Storage Replica will include more cmdlets before the final release, including a cmdlet 
to determine how well replication will perform between two servers over a given network, how to 
optimally size the replication logs, and what the current write I/O load is on a server you propose for 
replication—all without the need to install or configure Storage Replica beforehand.