Improvements in Failover Clustering in Windows Server 2012

Post on 01-Jun-2015

3.557 views 4 download

Tags:

description

Presentation by Michael Van Horenbeeck: http://twitter.com/mvanhorenbeeck. Video recording available here: http://technet.microsoft.com/en-us/video/windows-server-2012-improvements-in-failover-clustering.

Transcript of Improvements in Failover Clustering in Windows Server 2012

Failover Clustering ImprovementsWindows Server 2012

Michael Van HorenbeeckTechnology ConsultantXylos@mvanhorenbeeckMichael.van.horenbeeck@xylos.com

Agenda

Introduction to failover clustering1What’s new in Windows Server 2012 ?2

Continuous Availability3Q & A4

Introduction to failover clustering

A trip down memory lane

Failover Clustering

What?

Multiple individual computers working together to increase the availability (and scalability) of a clustered service.

Why?

Support business needs by avoiding downtime (increasing availability)

Result

+ Complexity+ Management+ Cost

Welcome to the future!

Failover Clustering features and improvements in Windows Server 2012

Failover Clustering in Windows Server 2012Infrastructure for the Private Cloud

Hyper-V

Management of the private cloud

Infrastructure of the private cloud

Platform of the private cloud

Scalability

Build larger clusters Up to 64 nodes Up to 4000 VMs !

Improvements to storage-model

Support for Hyper-V Replica

Failover Policies

Automated management

. . .

Scale Up

Scale Out

4,000 VM’s in a single cluster

64 nodes in a cluster

Manageability

81 Commands!

Get-Command –Module FailoverClusters

Automated Management

All Roles: Priority Settings Automated Node Draining

Hyper-V Specific Preemption of Virtual Machines Memory-Aware VM Placement Live Migration Queuing (Anti-Affinity Rules)

Clustered Scheduled Tasks

Can only be managed through PowerShell

Task Scheduler (GUI)can be used to view tasks

Register-ClusteredScheduledTaskSet/Get-ClusteredScheduledTask

Automated Node Draining

One-click action to drain all roles off a node

Simplifies maintenance / patching

Scriptable (PowerShell)

‘Intelligent’ Supports all clustered roles Takes appropriate actions, based on workload

E.g. VM’s will be live-migrated; File Server resources will be moved

Cluster Configuration

Cluster validation improvements New tests Validation runs faster Targeted validation: only run tests you need

Upgrade / Migration: Import / Export Virtual Machines Migrate configuration information from clusters running:

Windows Server 2008 Windows Server 2008 R2 Windows Server 2012

Cluster management

SIMPLICITY

DEMO

Clustered Shared Volumes (CSV)

Allows multiple nodes to have simultaneous access to the same file system

Runs ‘on top of’ NTFS

No unmount/mount operation required at failover

Shared StorageVHD VHD VHD

Metadata

CSV Improvements

Simplified setup! Support Scale-Out File Servers Increased Security (Bitlocker) Ready for other Windows Server 2012 Storage &

Network technologies like: Storage Spaces Thin Provisioning SMB 3.0 features …

And many, many more…

Setup shared storage

SIMPLICITY

DEMO

Hyper-V

Features

Deeper integration into Failover Clustering Manager E.g. Live Migration Queuing

Hyper-V enables Guest Clustering using FC (formerly only possible through iSCSI)

VM Mobility

VHD

Concurrent Live Migrations:

Multiple simultaneous LM’s for a given source or

target

Live Migration Queuing:In-box tools queue &

manage large numbers of VMs

Storage Live Migration:Moves VHD’s from one disk to

anotherHyper-V Replica:

Point-in-time replication of VHD’s for disaster recovery

VHD

VM Monitoring Application Health detection inside virtual machine

Clustering in the host identifies app failures and takes recovery action Failover Clustering Reboots VM Failover Clustering Fails over VM to another node

Generic health monitoring for any application Monitor services through Service Control Manager Generation of specific Event ID’s

VM Monitoring vs Guest Clustering

does not replace guest clustering Is a quick, simple and light-weight solution to monitor apps

running in VMs e.g Print Server Role

VM Monitoring Guest Clustering

Application Health Monitoring

Application Mobility

Simplified Configuration

Event Monitoring

Monitoring “Print Spooler”

Clustered Print Servers

Print Spooler service is no longer a clustered resource

In Windows Server 2012 it’s a VM running on a node in a Hyper-V cluster: No more fooling around with drivers! Reduced complexity

Live-Migration of VM to other node in cluster

Continuous Availability

The marriage between Failover Clustering & SMB 3.0 ?

Windows Server 2012 Continuous Availability Enables evolution from “High Availability” to

“Continuous Availability” for critical workloads.

At the same time: Keep costs under control Work with “what you know” Keep it simple

“Smart solutions” (e.g. DHCP Availability)

Achieving continuous availability…

… is not done by a single component!

Node 2Node N

Node 2Node N

Hyper-V

VSwitch

VM

VM

VM

Mgmt OS

External

Storage

Arrays

Shared JBOD

Node 1

HBA

HBA

Platform

Storage

Clustering

NIC

NIC

NIC

NIC

File Server & NetworkingStack

VMBus

Data Manageme

nt

VM

Node 1

Scal

e-Out

Scal

e-Out

Contributing improvements

Better AD DS integration Removed dependency on startup Support for delegated administration

Multi-Site Support Dynamic Quorum Specify node vote

The marriage between SMB 3.0 & Failover Clustering

SMB 3.0 Improvements

SMB Direct Offloading tasks from CPU to NIC > better performance

SMB Multichannel Aggregation of available paths

Fault tolerance Higher Bandwidth

SMB Transparent Failover

SMB Scale Out

Scale-Out File Servers Active/Active

Simultaneous access to a single share through all cluster nodes

Multiple shares still useful to isolate traffic

Targeted for Server Applications Server applications with few

metadata operations - Hyper-V and SQL Server

Bandwidth intensive applications – Increase available bandwidth by adding cluster nodes

Simplified and easy management Single logical file server Single file system namespaces No cluster disk resources to

manage

Single File System Namespace

Cluster Shared Volumes

Single Logical File Server (\\FS\Share)

Hyper-V Cluster

(Up to 63 nodes)

File Server Cluster(Up to 4 nodes)

Data Center Network

Scale-Out File Server Benefits

Active / Active File Shares Increased Bandwidth

Bandwidth = Bandwidth x Nodes CHKDSK with zero-downtime

Self-repairing “on the fly checking”

CSV Cache Easier management

No need for separate clusters with separate resources

Failover pre-Windows 2012

Historically not really transparent

Works great for traditional workloads

Storage expected to be continuously available

Some downsides: file handles lost during *over

leading to disconnects / timeouts File Server Cluster

File Server Node A

File Server Node B

\\fs1\share

1

2\\fs1\share

3

SQL Server

Failover share and connections and handles lost

2

Normal operation1

Administrator intervention needed to recover

3

Transparent failover in Windows 2012

Failover transparent to server application Only small IO delay during failover

Supports planned and unplanned failovers

Resilient for both file and directory operations

File Server Cluster

SQL Server

Failover share - connections and handles lost,temporary stall of IO

2

2

Normal operation1

Connections and handles auto-recoveredApplication IO continues with no errors

3

1 3

File Server Node A

File Server Node B

\\fs1\share \\fs1\share

SMB Transparent Failover: the details

SMB 3.0 Server

SMB 3.0 Client SMB 3.0 Server

User

Kernel

User

Kernel

WitnessService

WitnessClient

Witness Protocol(new)

SMB2 Redirector

File System

Resume Key Filter

SMB3 Server

SMB 3.0Operation

replayState

persistence

User

Kernel

• Persistent end-user experience

• Witness protocol to overcome traditional ‘issues’ (i.e. time-outs)

• Requires both client & server to run SMB 3.0

Continuous Available File Shares

DEMO

Cluster-Aware Updating

Cluster Maintenance

Prior to Windows Server 2012 -> Lots of manual work

1. Take offline / move resources2. Patch3. (Restart)4. Put back online5. Fail over resources ….

Take offline

Patch

Restart

Put online

Failover Resourc

es

Cluster-Aware Updating

Automated updating of all nodes in a cluster without affecting service (availability)

Combined with Continuous Available workloads = zero-impact

Works with existing technologies (Windows Update, WSUS, …)

Windows Update

Update Cluster(Orchestrator)

Node Drained

Admin

Initiates Check & Update

Cluster

Cluster Aware Updating

Coming later… ;-)

DEMO

Failover clustering in Windows Server 2012:

The road to your own private cloud!

Call to Action

Check www.microsoftvirtualacademy.com and learn more about Windows Server 2012

Try Windows Server 2012 out TODAY and download the Release Candidate from:http://technet.microsoft.com/en-US/evalcenter/hh670538.aspx

© 2012 Microsoft Corporation. All rights reserved. Microsoft, Windows, Windows Vista and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries.The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.

Appendix

Picture Credits

Slide 39: http://www.flickr.com/photos/emmebi89/2918973663/

Sources “Building Windows 8” ~ http://blogs.msdn.com/b/b8/ “Clustering and High-Availability” http://blogs.msdn.com/b/clustering/ “Windows Server Blog” ~ http://blogs.technet.com/b/windowsserver/