Microsoft Official Course
Module 6
Implementing Failover Clustering with Hyper-V
Module Overview
Overview of Failover Clustering
Configuring and Using Shared Storage
Implementing and Managing Failover Clustering
with Hyper-V
Lesson 1: Overview of Failover Clustering
What Is A Failover Cluster?
What are Failover and Failback?
Failover Cluster Networks Failover Cluster Storage What Is CSV? What Is a Quorum? What Is Encrypted Cluster Volume?
What Is A Failover Cluster?
Shared bus or iSCSI connection Clustered role highly available service
Cluster storage
Node 1 A dedicated network connects the nodes Node 2
A network connects the failover cluster and clients
Clients
What are Failover and Failback?
When a node fails, application is moved to a
different node
Operation is called failover During failover, all associated resources are moved
Failover occurs when: Node that hosts the application cannot be accessed One of the application resources fails Administrator forces a failover
Failback is moving an application back After the failed node becomes active again Can be automatic or manual
Failover Cluster Networks
Have at least two network paths between nodes
Multiple (fast) network cards are recommended Can use NIC Teaming and virtual network adapters
Cluster Network Types: Allow cluster network communication on this network
Allow clients to connect through this network
Do not allow cluster network communication Single network can support intranode and clients
Not recommended
Microsoft Failover Cluster Virtual Adapter Fault-tolerant connections across all available interfaces Similar to an internal NIC Teaming for clustering
Failover Cluster Storage
Shared storage provides consistent data When node fails, application data is still available Multiple nodes can access shared storage Shared storage options include: SAS iSCSI Fibre Channel Fiber Channel over Ethernet SMB 3.0 file shares
Only for Hyper-V, not failover cluster
Consider using CSV for shared storage
What Is CSV?
LUN is owned by single node at any time
Multiple nodes can access a CSV simultaneously Application failover without drive ownership change
Concurrent access to a single file system Single consistent file namespace across cluster nodes
Better use of disk space Virtual machine resources in a single logical location No special hardware is required Simple CSV setup
Hyper-V, SQL Server and file server can use CSV Use Storage Spaces to deploy CSV Deduplication and tiered storage spaces are supported
What Is a Quorum?
Consensus that enough nodes are available More than half of all votes Based on votes Nodes, file share, and disk can have a vote Cluster online, when a quorum has more than 50 percent of the votes Quorum modes: Node Majority Node and Disk Majority Node and File Share Majority No Majority: Disk Only
What Is a Quorum?
Node weight Can configure so not all nodes have a vote Useful in multisite cluster scenarios Works with all quorum modes except Disk Only
Dynamic Quorum Number of votes is adjusted dynamically Cluster can be available with less than 50% votes Enabled by default Works with all quorum modes, except Disk Only Tie breaker for 50 percent node split
Dynamic witness
What Is Encrypted Cluster Volume?
BitLocker-encrypted cluster disks Support for traditional failover disks and CSV Can be enabled before or after a disk is added Add the failover cluster AD DS identity as a
BitLocker protector to the target disk volumes Minimal performance impact Requirements
Windows Server 2012 or newer failover clustering Windows Server 2012 domain controller is accessible from all nodes BitLocker is installed Bitlocker PowerShell module or Manage-BDE
Lesson 2: Configuring and Using Shared Storage
Storing A Virtual Machine on an SMB 3.0 Shared
Folder Using Scale-Out File Server What Is iSCSI? What Is iSCSI Target Server? Demonstration: Using an iSCSI Target Server Considerations for Implementing iSCSI Storage Using Virtual Hard Disk Sharing as Shared Storage
Storing A Virtual Machine on an SMB 3.0 Shared Folder
Hyper-V supports SMB 3.0 or newer Hyper-V uses file share is used as cluster shared storage Continuously available shares are recommended All virtual machine storage can be on an SMB share Configuration, virtual hard disks, checkpoints SMB Share Applications profile should be used No access-based enumeration or share caching
Full permissions on NTFS folder and SMB share Hyper-V administrators Computer account of Hyper-V host Hyper-V cluster computer account, if Hyper-V clustered
Using Scale-Out File Server
File server role cluster can work in two modes: Scale-out file server cluster File server cluster for general use Benefits of scale-out file server cluster: Active-Active file shares
Increased bandwidth
CHKDSK with zero downtime CSV cache Simple management Automatic scale-out rebalancing
Clients redirected to the best node for access to a share Avoids unnecessary traffic redirection
What Is iSCSI?
Client-server SCSI transport protocol Sends SCSI commands over IP networks
iSCSI client runs iSCSI initiator
Any SCSI device can be accessed Windows Server supports only block storage
iSCSI target is Windows Server role service Can be added as failover cluster role Multipath I/O for redundancy iSCSI initiator is part of Windows client and
TCP/IP protocol
Block storage device is presented Used as locally attached disk
storage array
Windows Server operating systems
iSCSI target
What Is iSCSI Target Server?
Virtual disks use the .vhdx format Presented to iSCSI initiators as SCSI logical units
Limit initiators that can access logical units
Fixed size, dynamically expanding, or differencing Clear the virtual disk on allocation Size can be up to 64 TB
Can extend and shrink virtual disks online Enables application-consistent snapshots Includes Storage Management Initiative
Specification provider
Standards-based management Enables discovery and storage-based management
Demonstration: Using an iSCSI Target Server In this demonstration, you will see how to use an iSCSI target server
Considerations for Implementing iSCSI Storage
Use a fast network for accessing iSCSI storage Network should be dedicated for accessing storage Deploy the iSCSI target as a highly available role Including redundant network paths
Implement iSNS for discovering iSCSI devices
Implement security for iSCSI storage Limit initiators that can access iSCSI target Authenticate initiator connection Encrypt network traffic
Using Virtual Hard Disk Sharing as Shared Storage
A failover cluster runs inside virtual machines
A shared virtual disk used as a shared storage Virtual machines do not need access to iSCSI or FC SAN Presented as virtual SAS disk Can be used only for data
Requirements for shared virtual disk Virtual hard disk must be in .vhdx format Connected by using a virtual SCSI adapter Stored on a scale-out file server or CSV Supported operating systems in a virtual machine Windows Server 2012 or Windows Server 2012 R2
Using Virtual Hard Disk Sharing as Shared Storage
Comparing guest clustering options
Capability Supported storage How is storage presented in virtual machine Data flows through the Hyper-V switch Storage is configured at the Hyper-V host level Provides low latency and low CPU use Requires specific hardware Switch must be reconfigured when virtual machine is migrated Shared VHDX Storage Spaces, Serial attached SCSI, Fiber Channel, iSCSI, SMB Virtual Serial attached SCSI No Yes Yes (RDMA or Fiber Channel) No Virtual Fibre Channel Fiber Channel SAN Virtual Fiber Channel LUN No Yes Yes (Fibre Channel) Yes iSCSI iSCSI SAN iSCSI LUN Yes No No No
No
Yes
No
Lesson 3: Implementing and Managing Failover Clustering with Hyper-V
Implementing a Failover Cluster
Hyper-V Validation Tests
Demonstration: Creating a Failover Cluster Configuring Highly Available Virtual Machines Virtual Machine Monitoring Demonstration: Creating and Managing the Virtual
Machine Clustered Role What Is the Hyper-V Replica Broker Role? What Is CAU?
Implementing a Failover Cluster
Install Windows Server 2012 R2 servers Same hardware and same operating system on all nodes Configure shared storage All nodes should have access to shared storage
Install the failover clustering feature
Install the server role on the servers Only on servers that will host that highly available role
Validate and create a failover cluster
Configure a highly available clustered role Test clustered role failover
Hyper-V Validation Tests
Validate a Configuration Wizard Validates nodes before adding them to a failover cluster Required to have supported cluster configuration Run before creating a cluster and after any changes
Can limit validation to specific cluster storage
Hyper-V validation tests include
Compatibility of Virtual Fibre Channel SANs Integration Services Version Memory Resource Pool Compatibility Storage Resource Pool Compatibility Virtual Machine Network Configuration Virtual Machine Storage Configuration Matching Processor Manufacturers
Hyper-V Validation Tests
Hyper-V Validation Tests
Demonstration: Creating a Failover Cluster In this demonstration, you will see how to create a failover cluster
Configuring Highly Available Virtual Machines
High Availability Wizard Virtual machine storage is on shared storage
CSV or SMB 3.0 continuously available share
Virtual machine startup priority Higher priority is started before lower priority No auto start, must restart manually after failover Preferred owners Virtual machine will start on preferred Hyper-V host Start on possible owner only preferred owners are unavailable If preferred and possible owners are unavailable, virtual machine will move to other failover cluster node, but not start
Configuring Highly Available Virtual Machines
AntiAffinityClassNames Clustered roles in same AntiAffinityClassNames avoid same cluster node Prevents virtual machines from running on the same node Configured in Windows PowerShell or System Center 2012 Virtual Machine Manager Options to modify failover and failback settings Number of times to restart a clustered role Prevent failback of the clustered role to preferred node Virtual machine Policies settings Virtual machine Heartbeat monitoring
Requires integration services in virtual machine
Configuring Highly Available Virtual Machines
Configuring Highly Available Virtual Machines
Configuring Highly Available Virtual Machines
Configuring Highly Available Virtual Machines
Configuring Highly Available Virtual Machines
Virtual Machine Monitoring
Application health detection in virtual machine Monitor services through Service Control Manager
Configure service recovery to take no action
Monitor events in System, Application, or Security logs Windows Server 2012 or newer required
Configurable recovery actions Restart service Restart virtual machine Fail over virtual machine Virtual machine network and storage protection Failure of virtual hard disk or lost network connectivity Virtual machine moved to different cluster node
Virtual Machine Monitoring
Virtual Machine Monitoring
Virtual Machine Monitoring
Virtual Machine Monitoring
Demonstration: Creating and Managing the Virtual Machine Clustered Role In this demonstration, you will see how to create and manage a virtual machine clustered role.
What Is the Hyper-V Replica Broker Role?
ServerA
ServerB
Server1
ServerC
Server2
Replic a broke Server3 r
Failover cluster 1
Failover cluster 2
What Is CAU?
Automatically update cluster nodes Scans, downloads, and installs updates on all nodes Fails over clustered roles before installing the update
Minimal or no downtime
Restarts nodes after the updates are installed if needed Fails back clustered roles Customizable pre- and post-update actions
CAU can work in two modes Remote Updating mode Self Updating mode
Lab: Implementing Failover Clustering with Hyper-V
Exercise 1: Creating a Hyper-V Failover Cluster
Exercise 2: Managing a Hyper-V Failover Cluster
Logon Information Virtual machines: 20409A-LON-HOSTx, 20409A-LON-CLx, 20409A-LON-DC1, and 20409A-LON-SS1 Adatum\Administrator Pa$$w0rd
User name: Password:
Estimated Time: 60 minutes
Lab Scenario
A. Datum Corporation has implemented the Hyper-V virtualization platform in one of their subsidiaries. Most of the host servers in the subsidiary have been converted to Hyper-V hosts, including several servers that run critical business applications. These critical applications need to be available at all times, and the availability should not be affected by the failure of a single host machine. A. Datum has identified failover clustering as the best option for implementing this level of availability. You need to implement a high availability solution for these virtual machines by deploying failover clustering for the virtual machines. You also need to configure highly available virtual machines and virtual machine monitoring.
Lab Review
How can you control which servers can access
iSCSI virtual disks? When creating failover clustering in virtual machines, which storage types can virtual machines use as a shared storage? Where must you store a virtual hard disk if you want to use it as a shared storage for virtual machines?
Module Review and Takeaways
Review Questions
