Thursday, December 18, 2014

New Features in Windows Server 2012 R2 Failover Clustering

New Features and Functionalities in Windows Server 2012 R2 Failover Clustering

New Features:

# Shared Virtual Hard Disk for Guest Clusters

# Virtual Machine drain on Shutdown

# Virtual Machine Network Health Detection

# Cluster Dashboard

# A Cluster that's detached from AD.

# Force Quorum Resilency

# Tie breaker for 50% node split.

# Configuration of Global Update Manager Mode

# Agility to disable IpSec encryption for communications between nodes in Cluster.  

Shared Virtual Hard Disk (.VHDX) For Guest Clusters:

Windows Server 2012 R2 Supports for Virtual Hard Disk sharing which allows a Virtual Hard Disk file (VHXD) to attach to Multiple VM's at same time through SCSI Controller. The Shared VHXD file is shown as shared serial attached SCSI (SAS) device to VM's and used as Shared Cluster Storage. The VHXD File should be store on CSV means it can be access directly by Hyper-V Hosts on local CSV or through Scale out File Server using SMB 3.0.

.VHXD File can be shared ideally for File Services running within VM, SQL Server Databases and Other Database File resides on guest Clusters.

Virtual Machine Drain on Shut down:

When You Use Failover Clustering in Server 2012 and earlier, Microsoft recommends that we should first move all VM's off a node before you shut it down because when you shut a node it induces the Cluster Controlled action of Quick migration on each VM. Quick Migration will put that VM in Saved State.

When VM is in saved state that means it is down. Your work is down until VM come back to online state. The reason of recommendation for moving all VM's off the node before shutting down is that you can use live migration to move those VM's so that their is no hamper to work. In Windows Server 2012 R2 Failover Clustering, Microsoft has change what happens after a node is shut down. The new Process has two new components. Drain on Shutdown and Best available Node placement.

If you shutdown Node without first putting it in to maintenance mode, Cluster will than automatically issue drain. During drain cluster uses Live migration to move VM's off node following machine priority. All VM's are moved including low priority VM's.

When Migrating VM's Cluster uses Best Available Node placement. Before Cluster start migrating VM's, it first checks availability of memory of remaining nodes. Using this information it places VM on best available Node. This results in to smooth run because it prevents moving high priority VM's to Node which not have enough Memory. This new process is enable by default. If you want to manually enable or disable it you can configure DrainOnShutdown  Cluster Common Property.

In Windows Server 2012 if their is a network disconnection at virtual machine level, the Virtual Machine continues to run on that computer even though Virtual machine may not available to users.