Failover Cluster Infrastructure Requirements
Failover clusters depend on infrastructure services, and require that each server node must be in the same Active Directory domain and, if you use DNS, the nodes should use the same Domain Name System DNS servers for name resolution. We recommend that you install the same Windows Server 2008 features and roles on each node. In addition, you should not install the Active Directory Domain Services AD DS role on any of the cluster nodes. If you install the AD DS role on one of the nodes, you must...
The Cluster Validation Test Storage
The storage test validates Disk access latency SCSI device vital product data VPD The storage tests list and test the capabilities of all disks available to the cluster. These tests are comprehensive however, some specific tests may not run after the cluster is running nor in a multi-site cluster. The Validate a Configuration Wizard performs the following storage validation Lists all disks that are visible to one or more tested servers. The test lists Disks that can support clustering and can...
For more information see Whats New in Failover Clusters in Windows Server Azq
Storage features include the following features Parallel SCSI, which previous versions of server clusters supported, is no longer supported for shared disk devices in a failover cluster. In Windows Server 2008, the cluster disk driver is a stand-alone component that communicates with the Partition Manager driver PartMgr.sys , for managing clustered disks. During the installation process, the driver is registered as a system Plug-n-Play PNP device, and is started during the normal Windows...
Managing Cluster Nodes
Add nodes after you create a cluster Pause nodes, which prevents resources from running on that node Evict nodes from a cluster, which removes the node from the cluster configuration All of these actions are available from the Failover Cluster Management Actions pane There are three aspects to managing cluster nodes You can add a node to an established failover cluster by selecting Add Node from the Failover Cluster Management Actions pane. The Add Node Wizard will prompt you for information...
Windows Server NLB Features
Network connectivity with support for - Network Driver Interface Specification NDIS 6.0 - Multiple dedicated IP addresses per node Support for up to 32 computers per cluster Denial of service attack and timer starvation protection Support for rolling upgrades from Windows Server 2003 All NLB components support IPv6 addresses. You can configure the addresses as the primary cluster IP address, the dedicated IP addresses, and the virtual IP addresses. In addition, you can load balance IPv6 as...
Exercise Identifying Fibre Channel Storage Area Network SAN Components
The main task for this exercise is to identify components and points of failure Tailspin Toys is designing a simple Fibre Channel-based storage area network that will initially include a file server cluster. You are given the preliminary diagram of the new SAN, and have been asked to identify where the Host Bus Adaptors HBAs , switches, storage controller, and failure points might be. 1. Identify each component in the diagram using the word list shown. 2. Identify any single points of failure....
Overview of Requirements for Installing a Failover Cluster
Failover Cluster Server Hardware Requirements Failover Cluster Server Network Requirements Failover Cluster Storage Requirements Failover Cluster Infrastructure Requirements Failover Cluster Software Requirements Guidelines for Microsoft Support of Cluster Solutions Windows Server 2008 failover clusters have specific recommended hardware and software configurations that enable Microsoft to support the cluster. Failover clusters are intended to provide a higher level of service than stand-alone...
Exercise Clustering the File Services Role on Windows Server Core
Fabrikam has two servers with Server Core installed that you must configure as a failover cluster. You have to cluster a file service on these two servers. The Print Services role already is installed on VAN-CORE6Aand VAN-CORE6B. The iSCSI target also has been added to the server that you will use as the file share. The main tasks for this exercise are to Start the virtual machines, and then log on. Configure the iSCSI target mappings and shared drives on VAN-CORE6A. Configure the iSCSI target...
File share witness allows a node cluster with no shared disk
Single file server can serve as the witness for multiple clusters Single file server can serve as the witness for multiple clusters The Node and File Share Majority mode can help maintain quorum without adding another node to the cluster. To provide for a single site failure and allow automatic failover, the file share witness might need to exist at a third site. In a multi-site cluster, a single server can host the file share witness however, you must create a separate file share for each...
Failover Cluster Storage Requirements
All cluster storage device controllers must be identical, and use the same firmware version on each node Separate networks are required for client access, and to access iSCSI storage Storage drivers must be based on the Microsoft Storport driver model All multipathing solutions must be based on Microsoft Multipath I O MPIO Storage is one of the most critical components in a failover cluster and requires The mass-storage controllers match in each server node, in model, and in firmware versions....
How Resource Failover Works for Specific Resources
During failover, each of the resources is taken offline in the order set by the dependencies. After all resources are offline, the cluster moves the entire group ownership from the active node to the passive node. After the group ownership changes, the resources are started in the reverse order, which ensures that dependent resources are started first. The length of time for a failover depends on the cluster node's transaction load, and the resources that are being failed-over. For more...
Using Windows Server Backup to Back Up the Cluster Configuration
Windows Server Backup is the built-in backup and recovery software for Windows Server 2008. In order to complete a successful backup, be aware of the following For a backup to succeed in a failover cluster, the cluster must be running and must have quorum. In other words, enough nodes must be running and communicating perhaps with a witness disk or witness file share, depending on the quorum configuration, that the cluster has achieved quorum. You must back up all clustered applications. If you...
Configuring Cluster Resource Policies
k You can configure resources to respond in one of the following ways during a failure Attempt to restart on the same node X number of times for X amount of time If restart is unsuccessful, then fail over all resources in the application If restart is unsuccessful, restart the resource again after X amount of time You can customize a resource's response to a failover. For example, if a disk resource fails on the active node, you might want the entire group of resources to fail over to another...
Using Preferred Owners
Considerations for using preferred owners Preferred owners are set on the clustered application Multiple preferred owners can be set in an ordered list Setting preferred owners gives control over - The order in which an application will select a node to run on - Which applications can run on the same nodes in an Active Active configuration You can create a preferred owners list in order to control which nodes host clustered applications. Using preferred owners ensures that applications run on...
Benefits of Server Core Clustering
Server Core requires less maintenance than a full installation, because it installs fewer components. It also has a reduced attack surface, which allows fewer components to be exploited, and requires fewer updates if vulnerabilities are discovered. Server Core also requires less disk space and memory, which frees more resources for the clustered applications and services. For more information, see Windows Server 2008 Product Overview
Managing a Failover Cluster
Built-in management tools enable you to manage a cluster. This lesson outlines the failover cluster management options that help you maintain and configure the cluster. To manage the cluster, you can use either the Failover Cluster Management tool or the cluster.exe command line tool. Typically, you will use the Failover Cluster Management tool however, in addition to management tasks, you might use the command line tool for in-depth troubleshooting and logging activities. You must use the...
Scenario
Fabrikam has asked you to review cluster event logs and perform a test restore to validate the backup process for their existing file cluster. The Windows Server Backup feature is installed on both cluster nodes. Note Complete the steps to prepare for this lab after starting the virtual machines and logging on as described in Exercise 1, Task 1. 1. On VAN-SRV6A, click Start, click Administrative Tools, and then click Server Manager. 2. Expand Storage, and then click Disk Management. 3....
Key Points Bka
You use synchronous and asynchronous replication for different reasons. You use synchronous replication when data loss cannot be tolerated. Synchronous replication solutions cannot tolerate high-disk write latency, because the application waits for both storage solutions to acknowledge the data writes. The requirement for low latency disk writes also limits the distance between the storage systems because increased distance can cause higher latency. If the disk latency is high, the performance...
Considerations for Configuring NLB
Install multiple network cards to increase performance and provide redundancy Choose a cluster operations mode that is compatible with your network devices Use the load weight option on port rules to distribute client connections between cluster hosts Use single affinity for applications that require Secure Sockets Layer SSL connections, or multiple client connections Install the same server certificate on all cluster hosts that must support SSL connections Install the same server certificate...
Exercise Configuring Cluster Resources
Fabrikam has two servers configured as a failover cluster. You have to cluster the Print Services role on these two servers. The Print Services role already is installed on VAN-SRV6A and VAN-SRV6B. The main tasks for this exercise are to Change the preferred owner of VAN-Print to VAN-SRV6B. Change the failback settings to allow only failback to the preferred node between 01 00 and 04 00. Close all virtual machines, and discard undo disks. Task 1 Change the preferred owner of VAN-Print to...
Clustering Terminology
A server that participates in a cluster A device or service hosted on a cluster and accessed directly or indirectly by the application or end user A highly available clustering type with resources owned by a single server at a time A clustering type that distributes processing across a number of nodes A key component of clusters that withstands problems in hardware or software while continuing to operate The amount of time an application is unavailable due to an update or other maintenance The...
Tools for Troubleshooting NLB Clusters
Performance and server scalability issues troubleshooting tools such as Netdiag, Ping, Tracert, Nslookup, Netstat, ARP Network connectivity and name resolution
Comparing Unicast and Multicast NLB Cluster Operation Modes
The cluster operation mode is used to ensure that all cluster hosts receive all incoming client requests sent to the virtual IP address assigned to the cluster All cluster hosts share an identical unicast media access control MAC address The outgoing MAC address is modified, based on the cluster host's priority setting, to prevent switches from discovering that all cluster hosts have the same MAC address Each cluster host retains the original adapter MAC address The adapter is assigned a...
For more information see Windows Server Clustering Ori
Although complexity increases in a clustered environment, disaster recovery and update management capabilities outweigh these complexities. For example Disaster recovery improves in a cluster because application and hardware failures are recovered automatically by restarting the application on a healthy node. This enables the failed node to be repaired while the application remains available. In a stand-alone server configuration, you might need to reinstall or reconfigure an application on new...
Choosing a Quorum Mode
The most important reason for implementing failover clusters is to provide high availability for services and applications. Windows Server 2008 failover clusters use a quorum to ensure that enough members are available for the cluster to continue operating. In order to implement and manage failover clusters, you must understand the options that are available for configuring quorums. This lesson describes how quorum works in Windows Server 2008 failover clusters. The failover cluster quorum...
Task Perform an authoritative restore
1. Start an administrator-elevated command prompt. 2. Run wbadmin get versions to list the backups that have been completed. 3. Locate the backup that was just performed, by matching the Backup Time with the time that the backup was performed. Note the version identifier for the backup. 4. Run wbadmin start recovery -version 07 27 2007-19 21 -itemtype app -items cluster where the version information matches the version noted in the previous step, to start the restore. 5. After the restore is...
Using Event Viewer to Monitor Failover Clusters
Found in Applications and Services Logs Microsoft Windows FailoverClustering When problems arise in the cluster, use the Event Viewer to view events with a Critical, Error, or Warning severity level. Additionally, informational level events are logged to the Failover Clustering Operations log, which can be found in the Event Viewer in the Applications and Services Logs Microsoft Windows folder. Informational-level events are usually common cluster operations, such as cluster nodes leaving and...
Exercise Configuring Internet SCSI iSCSI Storage Connections
The main tasks for this exercise are to Start the virtual machines. Add the iSCSI target portal to 6423A-VAN-SRV3A. Add a persistent binding on 6423A-VAN-SRV3A. Add the iSCSI target portal to 6423A-VAN-SRV3B. Add a persistent binding on 6423A-VAN-SRV3B. Task 1 Start the virtual machines 1. Open the 6423A Lab Launcher. 2. Start the 6423A-VAN-DC1 virtual machine. 3. Start the 6423A-VAN-SRV3A virtual machine. 4. Start the 6423A-VAN-SRV3B virtual machine. Task 2 Add the iSCSI target portal to...
Using Clusterexe to Export the Cluster Log
Cluster.exe provides a cluster log for the entire cluster, or for a specific node in the cluster in a readable text format. The output from Cluster.exe is more flexible than Tracerpt.exe, because you can export trace logs for all nodes in the cluster and save them in a specified location. You can view the exported log using a text editor. To run successfully, Cluster.exe must be run from an administrator-elevated command prompt. Using the Reliability and Performance Monitor Snap-in The...
Using Event Tracing
Windows event tracing is a kernel component that is available early after startup, and late into shutdown. It is designed to allow for fast tracing and delivery of events to trace files and to consumers. Because it is designed to be fast, it allows only basic in-process filtering of events based on event attributes. The event trace log contains a comprehensive accounting of the failover cluster actions. Depending on how you want to view the data, use either Cluster.exe or Tracerpt.exe to access...
Key Points Hqr
To troubleshoot SAN issues, start by checking physical connections and each of the hardware component logs. Additionally, run the Validate a Configuration Wizard to verify that the current cluster configuration is still supportable. When you run the Validate a Configuration Wizard, ensure that the storage tests you select can be run on an online failover cluster. Several of the storage tests cause loss of service on the clustered disk when the tests are run. Troubleshooting Group and Resource...
Exercise Clustering the Print Services Role Using Failover Cluster Management
Fabrikam has two servers configured as a failover cluster. You have to cluster the Print Services role on these two servers. The Print Services role already is installed on VAN-SRV6A and VAN-SRV6B. The main tasks for this exercise are to Start the virtual machines, and then log on. Configure the printer disk. Cluster the Print Services role. Failover the VAN-Print clustered service from VAN-SRV6A to VAN-SRV6B. Task 1 Start the virtual machines, and then log on 1. On your host machine, click...