Avid® Interplay®Engine Failover Guide m a k e m a n a g e m ove | m e d i a ™ Avid ®
Legal Notices Product specifications are subject to change without notice and do not represent a commitment on the part of Avid Technology, Inc. The software described in this document is furnished under a license agreement. You can obtain a copy of that license by visiting Avid's Web site at www.avid.com. The terms of that license are also available in the product in the same directory as the software.
Copyright 1995, Trinity College Computing Center. Written by David Chappell. Permission to use, copy, modify, and distribute this software and its documentation for any purpose and without fee is hereby granted, provided that the above copyright notice appear in all copies and that both that copyright notice and this permission notice appear in supporting documentation. This software is provided "as is" without express or implied warranty. Copyright 1996 Daniel Dardailler.
This product includes portions of the Alloy Look & Feel software from Incors GmbH. This product includes software developed by the Apache Software Foundation (http://www.apache.org/). © DevelopMentor This product may include the JCifs library, for which the following notice applies: JCifs © Copyright 2004, The JCIFS Project, is licensed under LGPL (http://jcifs.samba.org/). See the LGPL.txt file in the Third Party Software directory on the installation CD.
GOT FOOTAGE? Editors — Filmmakers — Special Effects Artists — Game Developers — Animators — Educators — Broadcasters — Content creators of every genre — Just finished an incredible project and want to share it with the world? Send us your reels and we may use your footage in our show reel or demo!* For a copy of our release and Avid’s mailing address, go to www.avid.com/footage. *Note: Avid cannot guarantee the use of materials submitted.
Contents Using This Guide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Symbols and Conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 If You Need Help. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Viewing Help and Documentation on the Interplay Portal. . . . . . . . . . . . . . . . . . . . . 13 Accessing the Online Library . . . . . . . . . . . . . . .
Chapter 2 Creating a Microsoft Failover Cluster . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 Server Failover Installation Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 Before You Begin the Server Failover Installation . . . . . . . . . . . . . . . . . . . . . . . . . . 44 List of IP Addresses and Network Names . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 Preparing the Server for the Cluster Service . . . . . . . . . . . . . . . . . . . . . .
Chapter 3 Installing the Interplay Engine for a Failover Cluster . . . . . . . . . . . . . . . 85 Disabling Any Web Servers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 Installing the Interplay Engine on the First Node . . . . . . . . . . . . . . . . . . . . . . . . . . . 86 Preparation for Installing on the First Node . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86 Starting the Installation and Accepting the License Agreement . . . . . . . . . . . . .
Using This Guide Congratulations on the purchase of your Avid® Interplay™, a powerful system for managing media in a shared storage environment. This guide is intended for all Avid Interplay administrators who are responsible for installing, configuring, and maintaining an Avid Interplay Engine with the Automatic Server Failover module integrated. n The documentation describes the features and hardware of all models.
Symbol or Convention Meaning or Action Bold font Bold font is primarily used in task instructions to identify user interface items and keyboard sequences. Italic font Italic font is used to emphasize certain words and to indicate variables. Courier Bold font Courier Bold font identifies text that you type. Ctrl+key or mouse action Press and hold the first key while you press the last key or perform the mouse action. For example, Command+Option+C or Ctrl+drag.
Viewing Help and Documentation on the Interplay Portal Viewing Help and Documentation on the Interplay Portal You can quickly access the Interplay Help, PDF versions of the Interplay guides, and useful external links by viewing the Interplay User Information Center on the Interplay Portal. The Interplay Portal is a web site that runs on the Interplay Engine. You can access the Interplay User Information Center through a browser from any system in the Interplay environment.
Accessing the Online Library The Avid Interplay Online Library DVD contains all the Avid Interplay product documentation in PDF format. It includes multimedia tutorials for selected products, as well as other useful links. n You need Adobe® Reader® to view the documentation online. You can download the latest version from the Adobe web site. To access the online library from the Online Library DVD: 1. Insert the Online Library DVD into a DVD drive. 2.
1 Automatic Server Failover Introduction This chapter covers the following topics: • Server Failover Overview • How Server Failover Works • Installing the Failover Hardware Components • Clustering Terminology Server Failover Overview The automatic server failover mechanism in Avid Interplay allows client access to the Interplay Engine in the event of failures or during maintenance, with minimal impact on the availability.
1 Automatic Server Failover Introduction The following diagram illustrates the components of a cluster group, including sample IP addresses. For a list of required IP addresses and node names, see “List of IP Addresses and Network Names” on page 45. Cluster Group Intranet Resource groups Clustered services Failover Cluster 11.22.33.200 Node #1 Intranet: 11.22.33.44 Private: 10.10.10.10 Interplay Server (virtual) 11.22.33.201 Private Network MSDTC 11.22.33.202 Node #2 Intranet: 11.22.33.
Server Failover Configurations When the server is online, the resource monitor regularly checks its availability and automatically restarts the server or initiates a failover to the other node if a failure is detected. The exact behavior can be configured using the Windows Cluster Administrator console. Because clients connect to the virtual network name and IP address, which are also taken over by the failover node, the impact on the availability of the server is minimal.
1 Automatic Server Failover Introduction Redundant-Switch Configuration The following diagram illustrates the failover cluster architecture for an Avid Unity ISIS environment that uses two layer-3 switches. These switches are configured for failover protection through either HSRP (Hot Standby Router Protocol) or VRRP (Virtual Router Redundancy Protocol). The cluster nodes are connected to two subnets (VLAN 30), each on a different switch.
Server Failover Configurations The following table describes what happens in the redundant-switch configuration as a result of an outage: Type of Outage Result Hardware (CPU, network adapter, The cluster detects the outage and triggers failover to the remaining node. memory, cable, power supply) fails The Interplay Engine is still accessible. Network switch 1 (VLAN 30) fails External switches running VRRP/HSRP detect the outage and make the gateway available as needed.
1 Automatic Server Failover Introduction The following table describes what happens in the dual-connected configuration as a result of an outage: Type of Outage Result Hardware (CPU, network adapter, The cluster detects the outage and triggers failover to the remaining node. memory, cable, power supply) fails The Interplay Engine is still accessible.
Server Failover Requirements The following table describes what happens in the MediaNetwork configuration as a result of an outage: Type of Outage Result Hardware (CPU, network adapter, The cluster detects the outage and triggers failover to the remaining node. memory, cable, power supply) fails The Interplay Engine is still accessible. Network switch fails The cluster detects the outage and triggers failover but detects that the second node is also disconnected from the network, and both clusters fail.
1 Automatic Server Failover Introduction Antivirus Software You can run antivirus software on a cluster, if the antivirus software is cluster-aware. For information about cluster-aware versions of your antivirus software, contact the antivirus vendor. If you are running antivirus software on a cluster, make sure you exclude these locations from the virus scanning: Q:\ (Quorum disk), C:\Windows\Cluster, and S:\Workgroup_Databases (database).
Installing the Failover Hardware Components • “Failover Cluster Connections: Avid Unity MediaNetwork, Infortrend A16F-R221” on page 36 • “Failover Cluster Connections: Avid Unity MediaNetwork, Infortrend A16F-R2431” on page 39 SR2400 Slot Locations The SR2400 is supported as a server for the Interplay applications. This section describes the slot locations that are specific to the Interplay components in a cluster configuration.
1 Automatic Server Failover Introduction SR2500 Slot Locations (for Infortrend A16F-R221) The SR2500 is supported as a server for the Interplay applications. This section describes the slot locations that are specific to the Interplay components in a cluster configuration that uses the Infortrend model A16F-R221shared-storage RAID array. Use the following figure and table as guides to configuring an SR2500 system.
Installing the Failover Hardware Components SR2500 Slot Locations (for Infortrend A16F-R2431) The SR2500 is supported as a server for the Interplay applications. This section describes the slot locations that are specific to the Interplay components in a cluster configuration that uses the Infortrend A16F-R2431 shared-storage RAID array. Use the following figure and table as guides to configuring an SR2500 system.
1 Automatic Server Failover Introduction Failover Cluster Connections: Avid Unity ISIS, Redundant-Switch Configuration, Infortrend A16F-221 Make the following cable connections to add a failover cluster to an Avid Unity ISIS environment, using the redundant-switch configuration with an Infortrend A16F-R221 RAID array: • • First cluster node: - Left on-board network interface connector to layer-3 switch 1 (VLAN 30) - QLogic card connector to RAID array, Fibre Channel 1 left connector Second cluster no
Installing the Failover Hardware Components Failover Cluster Connections: Avid Unity ISIS, Redundant-Switch Configuration, SR2400, Infortrend A16F-R221 PCI adapter network interface right connector Interplay Engine Cluster Node 1 SR2400 Back Panel Right on-board network interface To Avid Network Switch 1 Fibre Channel 0 left connector QLogic card Left on-board network interface Private network for heartbeat Fibre Channel 1 left connector Cluster Shared-Storage RAID Array FC CH0 FC CH1 All switches s
1 Automatic Server Failover Introduction Failover Cluster Connections: Avid Unity ISIS Environment, Redundant-Switch Configuration, SR2500, Infortrend A16F-R221 PCI adapter network interface right connector Interplay Engine Cluster Node 1 Slot 3 Slot 2 Slot 1 Right on-board network interface To Avid Network Switch 1 Fibre Channel 0 left connector SR2500 Back Panel QLogic card Private network for heartbeat Left on-board network interface Fibre Channel 1 left connector Cluster Shared-Storage RAID Arra
Installing the Failover Hardware Components Failover Cluster Connections: Avid Unity ISIS, Redundant-Switch Configuration, Infortrend A16F-R2431 Make the following cable connections to add a failover cluster to an Avid Unity ISIS environment, using the redundant-switch configuration with an Infortrend A16F-R2431 RAID array: • • First cluster node: - Left on-board network interface connector to layer-3 switch 1 (VLAN 30) - ATTO 41XS card connector to RAID array, Fibre Channel 0 top-left connector Seco
1 Automatic Server Failover Introduction Failover Cluster Connections: Avid Unity ISIS, Redundant-Switch Configuration, SR2500, Infortrend A16F-R2431 PCI adapter network interface right connector Interplay Engine Cluster Node 1 Slot 3 Slot 2 Slot 1 SR2500 Back Panel Right on-board network interface ATTO 41XS card To Avid Network Switch 1 Left on-board network interface Fibre Channel 0 top left connector Cluster Shared-Storage RAID Array 1 ES A16F-R2431-1 BBU Status 1. Ctrl Status 2. C Dirty 3.
Installing the Failover Hardware Components Failover Cluster Connections: Avid Unity ISIS, Dual-Connected Configuration, Infortrend A16F-R221 Make the following cable connections to add a failover cluster to an Avid Unity ISIS environment, using the dual-connected configuration with an Infortrend A16F-R221RAID array: • • First cluster node: - Left on-board network interface connector to ISIS left subnet (VLAN 10) - Right on-board network interface connector to ISIS right subnet (VLAN 20) - QLogic ca
1 Automatic Server Failover Introduction Failover Cluster Connections: Avid Unity ISIS, Dual-Connected Configuration, SR2400, Infortrend A16F-R221 PCI adapter network interface right connector Interplay Engine Cluster Node 1 SR2400 Back Panel To ISIS left subnet Right on-board network interface Left on-board network interface To ISIS right subnet Fibre Channel 0 left connector QLogic card Private network for heartbeat Fibre Channel 1 left connector Cluster Shared-Storage RAID Array FC CH0 FC CH1 Al
Installing the Failover Hardware Components Failover Cluster Connections: Avid Unity ISIS, Dual-Connected Configuration, SR2500, Infortrend A16F-R221 PCI adapter network interface right connector Interplay Engine Cluster Node 1 Slot 3 Slot 2 Slot 1 SR2500 Back Panel To ISIS left subnet Right on-board network interface To ISIS right subnet Fibre Channel 0 left connector QLogic card Private network for heartbeat Left on-board network interface Fibre Channel 1 left connector Cluster Shared-Storage RAI
1 Automatic Server Failover Introduction Failover Cluster Connections: Avid Unity ISIS, Dual-Connected Configuration, Infortrend A16F-R2431 Make the following cable connections to add a failover cluster to an Avid Unity ISIS environment, using the dual-connected configuration with an Infortrend A16F-R2431 RAID array: • • First cluster node: - Left on-board network interface connector to ISIS left subnet (VLAN 10) - Right on-board network interface connector to ISIS right subnet (VLAN 20) - ATTO 41XS
Installing the Failover Hardware Components Failover Cluster Connections: Avid Unity ISIS, Dual-Connected Configuration, SR2500, Infortrend A16F-R2431 PCI adapter network interface right connector Interplay Engine Cluster Node 1 Slot 3 Slot 2 Slot 1 SR2500 Back Panel To ISIS left subnet Right on-board network interface ATTO 41XS card Left on-board network interface To ISIS right subnet Fibre Channel 0 top left connector Cluster Shared-Storage RAID Array 1 ES A16F-R2431-1 BBU Status 1.
1 Automatic Server Failover Introduction Failover Cluster Connections: Avid Unity MediaNetwork, Infortrend A16F-R221 Make the following cable connections to add a failover cluster to an Unity MediaNetwork environment, using the Infortrend A16F-R221RAID array: • • n First cluster node: - Left on-board network interface connector to Ethernet® public network on the Avid network switch - QLogic card connector to RAID array, Fibre Channel 1 left connector - ATTO card connector to Unity MediaNetwork FC s
Installing the Failover Hardware Components Failover Cluster Connections: Avid Unity MediaNetwork, SR2400, Infortrend A16F-R221 ATTO card To MediaNetwork FC switch Interplay Engine Cluster Node 1 SR2400 Back Panel To Ethernet Public Network Right on-board network interface QLogic card PCI adapter network interface - not used Left on-board network interface Private network for heartbeat Fibre Channel 1 left connector Fibre Channel 0 left connector Cluster Shared-Storage RAID Array FC CH1 FC CH0 A
1 Automatic Server Failover Introduction Failover Cluster Connections: Avid Unity MediaNetwork, SR2500, Infortrend A16F-R221 PCI adapter network interface - not used Interplay Engine Cluster Node 1 To Ethernet public network ATTO card To MediaNetwork FC switch Slot 3 Slot 2 Slot 1 Right on-board network interface SR2500 Back Panel QLogic card Left on-board network interface Private network for heartbeat Fibre Channel 1 left connector Fibre Channel 0 left connector Cluster Shared-Storage RAID Arra
Installing the Failover Hardware Components Failover Cluster Connections: Avid Unity MediaNetwork, Infortrend A16F-R2431 Make the following cable connections to add a failover cluster to an Unity MediaNetwork environment.
1 Automatic Server Failover Introduction Failover Cluster Connections: Avid Unity MediaNetwork, SR2500, Infortrend A16F-R2431 PCI adapter network interface - not used Interplay Engine Cluster Node 1 To Ethernet public network To MediaNetwork FC switch ATTO 41EL card Slot 3 Slot 2 Slot 1 SR2500 Back Panel Right on-board network interface ATTO 41XS card Left on-board network interface Fibre Channel 0 top left connector Cluster Shared-Storage RAID Array 0 1 ES A16F-R2431-1 BBU Status 1.
Clustering Terminology Clustering Terminology Clustering is not always straightforward, so it is important that you get familiar with the terminology of server clusters before you start. A good source of information is the Microsoft Technology Center for Clustering Services under: http://www.microsoft.com/windowsserver2003/technologies/clustering/default.mspx Detailed architecture documentation can be found here: http://www.microsoft.com/windowsserver2003/techinfo/overview/servercluster.
1 Automatic Server Failover Introduction 42
2 Creating a Microsoft Failover Cluster This chapter describes the processes for creating a Microsoft failover cluster for automatic server failover. It is crucial that you follow the instructions given in this chapter completely, otherwise the automatic server failover will not work.
2 Creating a Microsoft Failover Cluster n • Install the Interplay Engine on both nodes (see “Installing the Interplay Engine for a Failover Cluster” on page 85). • Test the complete installation (see “Testing the Complete Installation” on page 105). Do not install any other software on the cluster machines except the Interplay engine. For example, Media Indexer software needs to be installed on a different server.
Before You Begin the Server Failover Installation For information on creating a cluster user account, see the Microsoft document “Guide to Creating and Configuring a Server Cluster under Windows Server 2003.” http://www.microsoft.com/technet/prodtechnol/windowsserver2003/technologies/cl ustering/confclus.mspx. - n Cluster Installation and Administration Account: Create or select a user account to use during the installation process.
2 Creating a Microsoft Failover Cluster n n If your Active Directory domain or DNS includes more than one cluster, to avoid conflicts, you need to make sure the cluster names, MSDTC names, and IP addresses are different for each cluster. All names must be valid and unique network host names.
Before You Begin the Server Failover Installation IP Addresses and Node Names: ISIS Dual-Connected Configuration Node or Service Item Required Example Name Where Used First Cluster Node • 1 Host Name SECLUSTER1 • 2 ISIS IP addresses - public (one for left and one for right) • 1 IP address - private (Heartbeat) See “Configuring the Cluster Service on the First Node” on page 63 and “Creating a Resource Group for the Distributed Transaction Coordinator” on page 79.
2 Creating a Microsoft Failover Cluster IP Addresses and Node Names: MediaNetwork Configuration Node or Service Item Required Example Name Where Used First Cluster Node • 1 Host Name SECLUSTER1 • 1 MediaNetwork IP address public • 1 IP address - private (Heartbeat) See “Configuring the Cluster Service on the First Node” on page 63 and “Creating a Resource Group for the Distributed Transaction Coordinator” on page 79.
Preparing the Server for the Cluster Service • “Renaming the Local Area Network Interface on Each Node” on page 54 • “Removing Unnecessary Windows Components” on page 53 • “Renaming the Local Area Network Interface on Each Node” on page 54 • “Configuring the Private Network Adapter on Each Node” on page 56 • “Configuring the Binding Order Networks on Each Node” on page 59 • “Configuring the Public Network Adapter on Each Node” on page 61 • “Joining Both Servers to the Active Directory Domain”
2 Creating a Microsoft Failover Cluster To set the QLogic HBA link speed: 1. On the first node, click Start, and select Programs > QLogic Management Suite > San Surfer. The San Surfer FC HBA Manager dialog box opens. 2. In the left pane, select Port 1. 3. Click the Settings tab. 4. In the HBA Port Settings section, click the arrow pointer for the Data Rate list and change the default setting from Auto to 2 Gbps. 5. Click Save. 6. When prompted for a password, enter config and click OK. 7.
Preparing the Server for the Cluster Service Increasing the Boot Delay for Systems with ATTO Cards If your servers use an ATTO card to connect to the Infortrend, you need to increase the timeout value in the boot.ini file. This setting increases the time it takes for the server to boot, which might avoid a problem if the Infortrend requires a longer than usual time for a self-test and initializing. If a server boots before the Infortrend is ready, the cluster might fail.
2 Creating a Microsoft Failover Cluster 2. In the left pane, navigate to the appropriate channel on your host adapter. The NVRAM tab opens. 3. Click the arrow pointer for the Data Rate list and change the default setting from Auto to 4 Gb/sec. 4. Click Commit. 5. Reboot the system. 6. Open the Configuration tool again and verify that the Data Rate is set to 4 Gb/sec. 7. On the other node, repeat steps 1 through 6.
Preparing the Server for the Cluster Service Removing Unnecessary Windows Components You need to remove some unnecessary Windows components before you configure the cluster service. Which components you remove depends on the type of cluster you are configuring.
2 Creating a Microsoft Failover Cluster Renaming the Local Area Network Interface on Each Node You need to rename the LAN interface on each node to appropriately identify each network. Although you can use any name for the network connections, Avid suggests that you use the naming conventions provided in the table in the following procedure. Make sure you use the same name on both nodes. The names and network connections on both nodes must match. To rename the local area network connections: 1.
Preparing the Server for the Cluster Service Use the following illustration and table for reference. The illustration uses an SR2400 in an Avid Unity ISIS environment as an example. Left PCI adapter network interface SR2400 back view Right PCI adapter network interface Right on-board network interface This illustration shows Unity ISIS environment.
2 Creating a Microsoft Failover Cluster Naming Network Connections (Continued) Network Interface Avid Unity ISIS New Names (Redundant-switch configuration) Right PCI adapter network Private interface. This is a private network used for the heartbeat between the two servers in the cluster. Avid Unity ISIS New Names (Dual-connected configuration) Avid Unity MediaNetwork New Names Private Not used (Disabled) This is a private network used for the heartbeat between the two servers in the cluster. 4.
Preparing the Server for the Cluster Service 3. On the General tab, click the Internet Protocol (TCP/IP) check box. Make sure all other components are unchecked. Select this check box. All others are unchecked. 4. Select Internet Protocol (TCP/IP) and click Properties. The Internet Protocol (TCP/IP) Properties dialog box opens.
2 Creating a Microsoft Failover Cluster Type the private IP address for the node you are configuring. 5. On the General tab of the Internet Protocol (TCP/IP) Properties dialog box: n a. Select “Use the following IP address.” b. IP address: type the IP address for the Private network connection for the node you are configuring. See “List of IP Addresses and Network Names” on page 45.
Preparing the Server for the Cluster Service 7. On the DNS tab, make sure no values are defined and that the “Register this connection’s addresses in DNS” and “Use this connection’s DNS suffix in DNS registration” are not selected. 8. On the WINS tab, do the following: t Make sure no values are defined in the WINS addresses area. t Uncheck “Enable LMHOSTS Lookup”. t Select “Disable NetBIOS over TCP/IP.” 9. Click OK.
2 Creating a Microsoft Failover Cluster To configure the binding order networks: 1. On one node, open the Network Connections window. 2. Select Advanced > Advanced Settings. 3.
Preparing the Server for the Cluster Service - For an Avid Unity MediaNetwork environment use the following order: - Public - Private 4. Click OK. 5. Repeat this procedure on the other node and make sure the configuration matches on both nodes. Configuring the Public Network Adapter on Each Node Make sure you configure the IP address network interfaces for the Public Network Adapter as you normally would.
2 Creating a Microsoft Failover Cluster The following illustration shows the required names and drive letters. Configure disks as shown 5. Verify you can access the disk and that it is working by creating a file and deleting it. 6. Shut down the first node and start the second node. 7. On the second node, assign drive letters and names. You do not need to format the disks. 62 a. Open the Disk Management tool. Right-click the partition, select Change Drive Letter, and enter the appropriate letter.
Configuring the Cluster Service Configuring the Cluster Service Take the following steps to configure the cluster service: 1. Turn off the second node. 2. Configure the first node using the New Server Cluster Wizard. See “Configuring the Cluster Service on the First Node” on page 63 3. Validate the cluster service installation on the first node. See “Validating the Cluster Service on the First Node” on page 68. 4. Turn on the second node. Leave first node turned on. 5.
2 Creating a Microsoft Failover Cluster Type the Cluster service name. 8. Click Next. The Select Computer dialog box opens. n 64 You might be prompted for an account. If so, use a domain user account, such as the Cluster Installation Account referred to in “Before You Begin the Server Failover Installation” on page 44. Do not use the Cluster Service Account (Service Execution User).
Configuring the Cluster Service 9. In the Select Computer dialog box, in the Computer name text box, type the Cluster node host name of the first node. For example, use SECLUSTER1. See “List of IP Addresses and Network Names” on page 45. 10. Click Advanced. The Advanced Configuration Options dialog box opens. 11. Select Advanced (minimum) configuration, and click OK. 12. Click Next.
2 Creating a Microsoft Failover Cluster The setup process analyzes the node for hardware or software problems that might cause problems during installation. A warning icon displays next to “Checking Cluster feasibility.” In this case, the warnings do not indicate a problem. 13. Click Next after the analyze is complete and the Task Complete bar is green. 14. In the IP Address dialog box, type the Interplay Engine service ISIS IP address for the left side in the IP Address text box.
Configuring the Cluster Service 18. Click Quorum. The Cluster Configuration Quorum dialog box opens. 19. Select Disk Q: from the menu, and click OK. 20. Review the summary on the Proposed Cluster Configuration dialog box to verify all the information for creating the cluster is correct. 21. Click Next. The Creating the Cluster dialog box opens. 22. Review any errors during the cluster creation. n If red errors display, check the Cluster Service ISIS IP address you entered in step 14.
2 Creating a Microsoft Failover Cluster 23. Click Next. 24. Click Finish. Validating the Cluster Service on the First Node To validate the first node cluster installation: 1. Click Start and select Programs > Administrative Tools > Cluster Administrator. 2. In the left pane, click Resources to make sure all resources are online. Verify Resources Configuring the Cluster Service on the Second Node To configure the cluster service on the second node: 1.
Configuring the Cluster Service For example, use SECLUSTER2. See “List of IP Addresses and Network Names” on page 45. 7. Click Advanced. The Advanced Configuration Options dialog box opens. 8. Select Advanced (minimum) configuration, and click OK. 9. Click Next.
2 Creating a Microsoft Failover Cluster The setup process analyzes the node for hardware or software problems that might cause problems during installation. A warning icon displays next to “Checking Cluster feasibility.” In this case, the warnings do not indicate a problem. 10. Click Next after the analyze is complete and the Task Complete bar is green. 11. Type the password for the cluster service account. This account is used to start the cluster service. 12. Click Next. 13.
Configuring Rules for the Cluster Networks Configuring Rules for the Cluster Networks After the networks are configured on each node and the cluster service is configured, you need to configure the network roles to determine the function within the cluster. n This procedure uses Left-74 and Right-75 as examples of the public networks. If you are installing a dual-connection configuration, replace the numbers with your subnet numbers. To configure the rules for the cluster networks: 1.
2 Creating a Microsoft Failover Cluster 6. In the Public or Left-74 Properties dialog box, verify these options: - Name: Left-74 - Enable this network for cluster use - All communications (mixed network) 7. Click OK. 8. If you are installing a dual-connection configuration, in the left pane, click Cluster Configuration > Networks, and right-click Right-75 and select Properties. 9.
After Setting Up the Cluster 4. Verify the Private network is at the top of the list. You can use the Move Up and Move Down buttons to change the priority order. 5. Click OK. After Setting Up the Cluster After you finish setting up the cluster you need to verify that the quorum disk is using disk Q, set the startup times for each node, and test the cluster installation.
2 Creating a Microsoft Failover Cluster Verifying the Quorum Disk The Cluster Configuration Wizard automatically selects the disk used as the quorum device. Check to make sure the quorum device is using disk Q. To verify the quorum disk: 1. On either node, click Start and select Programs > Administrative Tools > Cluster Administrator. 2. In the left pane, right-click the cluster name at the top of the list and select Properties. 3. Click the Quorum tab and make sure Quorum resource displays Disk Q. 4.
After Setting Up the Cluster Setting the Startup Times on Each Node Avid recommends that you offset the startup time of each node’s operating system used during the power up of the cluster. n If you are configuring servers that use an ATTO card, these settings should display the values you set in the boot.ini file (see “Increasing the Boot Delay for Systems with ATTO Cards” on page 51). To set the time for displaying the list of operating systems (QLogic card): 1.
2 Creating a Microsoft Failover Cluster 5. Set the time to 15 seconds. 6. Click OK. 7. Repeat this procedure on the second node, but set the time to 75 seconds. To set the time for displaying the list of operating systems (ATTO card): 1. On the first node, click Start and right-click My Computer and select Properties. 2. Click the Advanced tab. 3. In the Startup And Recovery area, click Settings. The Setup and Recovery dialog box opens. 4. Select “Time to display list of operating systems.” 5.
After Setting Up the Cluster Testing the Cluster Installation You must test the cluster installation to make sure the failover process is working. To verify that resources will failover: 1. Click Start, and select Programs > Administrative Tools > Cluster Administrator. The Cluster Administrator window opens. Click Groups 2. In the left pane, open the Groups folder, right-click Cluster Group, and select Move Group. The group and all its resources are moved to the other node.
2 Creating a Microsoft Failover Cluster All resources are online Second node is now owner of the resources 3. Move the group back to node 1 after you finish testing the cluster installation. 4. Close the Cluster Administrator. Configuration of the cluster service on all nodes is complete and the cluster is fully operational. You can now install cluster resources, such as file shares, cluster aware services such as Distributed Transaction Coordinator.
Installing the Distributed Transaction Coordinator • “Creating a Physical Resource for the MSDTC Group” on page 82 • “Assigning Distributed Transaction Coordinator Resource to the MSDTC Group” on page 83 When performing these procedures Avid suggests you use the same entries shown in the procedure. These entries are from the list in section “List of IP Addresses and Network Names” on page 45.
2 Creating a Microsoft Failover Cluster 3. In the Name text box, type MSDTC. You can use any name for the group name, however Avid suggests you use MSDTC. 4. Click Next. The Preferred Owners dialog box opens. 5. Select both owners in the Available nodes list and add them to the Preferred owners list. 6. Click Finish. The group is created. Assigning an IP Address to the MSDTC Group To assign an IP address to MSDTC group: 1. In the Cluster Administrator, right-click MSDTC group and select New > Resource.
Installing the Distributed Transaction Coordinator 3. Complete the Possible Owners dialog box as follows: - Add the cluster server host names to the Possible owners lists. For example, SECLUSTER1 and SECLUSTER2. See “List of IP Addresses and Network Names” on page 45. 4. Complete the Dependencies dialog box as follows: - Leave the Resource dependencies list empty. 5. Complete TCP/IP Address Parameters dialog box as follows: - Address: type the IP address of the MSDTC service.
2 Creating a Microsoft Failover Cluster 8. Complete Network Name Parameters dialog box as follows: - Name: Type the virtual host name for the MSDTC group, for example, CLUSTERMSDTC. See “List of IP Addresses and Network Names” on page 45. Make sure to use a unique name for each Interplay Engine cluster on the network. - Uncheck “DNS Registration must succeed” and uncheck “Enable kerberos authentication.” 9. Click Finish.
Installing the Distributed Transaction Coordinator Assigning Distributed Transaction Coordinator Resource to the MSDTC Group To assign Distributed Transaction Coordinator Resource to MSDTC group: 1. In the Cluster Administrator, right-click MSDTC group and select New > Resource. 2. Complete the New Resource dialog box as follows: - Name: MSDTC Resource - Resource Type: Distributed Transaction Coordinator - Group: MSDTC 3. Click Next. 4.
2 Creating a Microsoft Failover Cluster To bring the MSDTC online: 1. Initialize the MSDTC Log file by doing the following: a. Bring MSDTC DISK R online: right-click MSDTC DISK R and select Bring Online. b. In the Command Window, run the following command on the node that is the owner to reset the log: msdtc -resetlog 2. Bring MSDTC Group online by right-clicking MSDTC, and selecting Bring Online.
3 Installing the Interplay Engine for a Failover Cluster After you set up and configure the cluster, you need to install the Interplay Engine software on both nodes.
3 Installing the Interplay Engine for a Failover Cluster Installing the Interplay Engine on the First Node The following sections provide procedures for installing the Interplay Engine on the first node. For a list of example entries, see “List of IP Addresses and Network Names” on page 45.
Installing the Interplay Engine on the First Node n When installing the Interplay Engine for the first time on a machine with cluster services, you are asked to choose between clustered and regular installation. The installation on the second node (or later updates) reuses the configuration from the first installation without allowing you to change the cluster-specific settings. In other words, it is not possible to change the configuration settings without uninstalling the Interplay Engine.
3 Installing the Interplay Engine for a Failover Cluster • “Specifying the Server User” on page 96 • “Specifying the Server Cache” on page 97 • “Enabling Email Notifications” on page 98 • “Installing the Interplay Engine for a Custom Installation on the First Node” on page 100 For information about updating the installation, see “Updating a Clustered Installation (Rolling Upgrade)” on page 107. Specifying Cluster Mode During a Custom Installation To specify cluster mode: 1.
Installing the Interplay Engine on the First Node 3. Select Cluster and click Next to continue the installation in cluster mode. The Specify Interplay Engine Details dialog box opens. Specifying the Interplay Engine Details In this dialog box, provide details about the Interplay Engine.
3 Installing the Interplay Engine for a Failover Cluster To specify the Interplay Engine details: 1. Type the following values: - Virtual IP address: This is the Interplay Engine service IP Address, not the Cluster service IP address. For a list of example names, see “List of IP Addresses and Network Names” on page 45. - Subnet Mask: The subnet mask on the local network. - Public Network: For a redundant-switch ISIS configuration or MediaNetwork configuration, type Public.
Installing the Interplay Engine on the First Node The Specify Interplay Engine Name dialog box opens. Specifying the Interplay Engine Service Name In this dialog box, type the name of the Interplay Engine service. To specify the Interplay Engine name: 1. Specify the public names for the Avid Interplay Engine service by typing the following values: - The Network Name will be associated with the virtual IP Address that you entered in the previous Interplay Engine Details dialog box.
3 Installing the Interplay Engine for a Failover Cluster The Specify Destination Location dialog box opens. Specifying the Destination Location In this dialog box specify the folder in which you want to install the Interplay Engine program files. To specify the destination location: 1. Avid recommends that you keep the default path C:\Program Files\Avid\Avid Interplay Engine. c Under no circumstances attempt to install to a shared disk; independent installations are required on both nodes.
Installing the Interplay Engine on the First Node Specifying the Default Database Folder In this dialog box specify the folder where the database data is stored. To specify the default database folder: 1. Type S:\Workgroup_Databases. Make sure the path specifies the shared drive (S:). This folder should reside on the shared drive that is owned by the resource group of the server. Avid strongly recommends using the shared drive resource so that it can be monitored and managed by the cluster service.
3 Installing the Interplay Engine for a Failover Cluster Specifying the Share Name In this dialog box specify a share name to be used for the database folder. To specify the share name: 1. Accept the default share name. Avid recommends you use the default share name WG_Database$. This name is visible on all client platforms, such as Windows 98, Windows ME, Windows NT Windows 2000 and Windows XP.The “$” at the end makes the share invisible if you browse through the network with the Windows Explorer.
Installing the Interplay Engine on the First Node Set for both nodes. Use this option for Interplay Archive Engine A Central Configuration Server (CCS) is an Avid Interplay Engine with a special module that is used to store server and database-spanning information. For more information, see the Avid Interplay Engine and Avid Interplay Archive Engine Administration Guide. To specify the server to act as the CCS server: 1.
3 Installing the Interplay Engine for a Failover Cluster Specifying the Server User In this dialog box, define the Cluster Service account (Server Execution User) used to run the Avid Interplay Engine. The Server Execution User is the Windows domain user that runs the Interplay Engine and the cluster service. This account is automatically added to the Local Administrators group on the server. This account must be the one that was used to set up the cluster service.
Installing the Interplay Engine on the First Node If necessary, you can change the name of the Server Execution User after the installation. For more information, see “Troubleshooting the Server Execution User Account” and “Re-creating the Server Execution User” in the Avid Interplay Engine and Avid Interplay Archive Engine Administration Guide and the Interplay ReadMe. Specifying the Server Cache In this dialog box, specify the path for the cache folder.
3 Installing the Interplay Engine for a Failover Cluster Enabling Email Notifications The first time you install the Avid Interplay Engine, the Enable Email Notification dialog box opens. The email notification feature sends emails to your administrator when special events, such as “Cluster Failure,” “Disk Full,” and “Out Of Memory” occur. Activate email notification if you want to receive emails on special events, server or cluster failures. To enable email notification: 1.
Installing the Interplay Engine on the First Node 2. Type the administrator's email address and the email address of the server, which is the sender. If an event, such as “Resource Failure” or “Disk Full” occurs on the server machine, the administrator receives an email from the sender's email account explaining the problem, so that the administrator can react to the problem. You also need to type the static IP address of your SMTP server.
3 Installing the Interplay Engine for a Failover Cluster Installing the Interplay Engine for a Custom Installation on the First Node In this dialog box, begin the installation of the engine software. To install the Interplay Engine software: 1. Click Next. Use the Back button to review or change the data you have entered. You can also terminate the installer using the Cancel button, because no changes have been done to the system yet.
Installing the Interplay Engine on the First Node 3. Do one of the following: t Click Finish. t Analyze and resolve any issues or failures reported. 4. Click OK if prompted for a restart the system. The installation procedure requires the machine to restart (up to twice). For this reason it is very important that the other node is shut down, otherwise the current node loses ownership of the Avid Workgroup resource group. This applies to the installation on the first node only.
3 Installing the Interplay Engine for a Failover Cluster Avid Workgroup Disk is online The Avid Workgroup Disk resource should be online and all other resources offline. 3. Bring the disk resource online manually before continuing if necessary. n Avid does not recommend starting the server at this stage yet, since it is not installed on the other node and a failover would be impossible. 4.
Installing the Interplay Engine on the First Node d. Complete the Dependencies dialog box as follows: e. f. Leave the Resource dependencies list empty. Complete TCP/IP Address Parameters dialog box as follows: - Address: type the second Interplay Engine service Avid Unity ISIS IP address.
3 Installing the Interplay Engine for a Failover Cluster Affect the group option j. Click OK. 5. When the installation is complete, leave this node running so that it maintains ownership of the resource group and proceed to “Installing the Interplay Engine on the Second Node” on page 104. Installing the Interplay Engine on the Second Node To install the Interplay Engine on the second node: 1.
Bringing the Interplay Engine Online 3. Click OK. 4. The same installation dialog boxes will open that you saw before, except for the cluster related settings that only need to be entered once. Enter the requested information and allow the installation to proceed. c Make sure you use the installation mode that you used for the first node and enter the same information throughout the installer. Using different values results in a corrupted installation. 5.
3 Installing the Interplay Engine for a Failover Cluster To test the complete installation: 1. To start the server, bring the resource group online; this starts the Interplay Engine and its affiliated services. After starting the Avid Interplay Engine on the first node, the Cluster Administrator should look similar to the following figure. 2. Start an Interplay Administrator, install the licenses if needed, create a test database and add some files to it.
Updating a Clustered Installation (Rolling Upgrade) Updating a Clustered Installation (Rolling Upgrade) A major benefit of a clustered installation is that you can perform “rolling upgrades.” You can keep a node in production while updating the installation on the other, then move the resource over and update the second node as well. n For information about updating specific versions of the Interplay Engine and a cluster, see the Avid Interplay ReadMe.
3 Installing the Interplay Engine for a Failover Cluster 5. During the installation, the installer displays a dialog box that asks you to move the Avid Workgroup Server group to the second node. Move the group, then click OK in the installation dialog box to continue. Restart as requested and continue with Part 2 of the installation. The installer will ask you to restart again after Part 2. 6. You might want to test the final result of the update by moving the server back to the first node.
Uninstalling the Interplay Engine on a Clustered System 6. When you are asked if you want to restart the system, click Yes. 7. At the end of the uninstallation process, if you are asked to restart the system, click Yes. 8. After the uninstallation on the inactive node is complete, wait until the last restart is done. Then open the Cluster Administrator on the active node and make sure the inactive node is shown as online.
3 Installing the Interplay Engine for a Failover Cluster 110
4 Automatic Server Failover Tips and Rules This chapter provides some important tips and rules to use when configuring the automatic server failover. Don't Access the Machines Directly Don’t access the machines (nodes) directly. Use the virtual network name or IP address that has been assigned to the Interplay Engine resource group (see “List of IP Addresses and Network Names” on page 45). Never use the actual physical names or IP addresses of the machines that are part of the cluster.
4 Automatic Server Failover Tips and Rules Do Not Change the Interplay Engine Server Execution User The domain account that was entered when setting up the cluster (the Cluster Service Account —see “Before You Begin the Server Failover Installation” on page 44) also has to be the Server Execution User of the Interplay Engine. Given that you cannot easily change the cluster user, the Interplay Engine execution user has to stay fixed as well.
Changing the CCS If you specify the wrong Central Configuration Server (CCS), you can change the setting later on the server machine in the Windows Registry under: HKEY_LOCAL_MACHINE/Software/Avid Technology/Workgroup/DatabaseServer The string value CMS specifies the server. Make sure to set the CMS to a valid entry while the Interplay Engine is online, otherwise your changes to the registry won't be effective.
4 Automatic Server Failover Tips and Rules 114
ABCDEFGHIJKLMNOPQRSTUVWXYZ Index A B Active Directory domain adding cluster servers 61 Antivirus software running on a failover cluster 21 Apache web server on failover cluster 85 ATTO card setting boot delay 51 setting link speed 51 Avid Unity environment ISIS failover cluster connections SR2400 (illustration) 29 ISIS failover cluster connections SR2500 (illustration) 28, 30, 34 MediaNetwork failover cluster connections SR2500 ( illustration) 38, 40 SR2400 server slot locations (failover cluster) 23 SR
Index ABCDEFGHIJKLMNOPQRSTUVWXYZ D E default database location for failover cluster 93 installing on first node 86 Server Execution User, specifying for failover cluster 96 share name for failover cluster 94 Interplay Portal viewing 13 IP addresses (failover cluster) assigning to MSDTC group 80 private network adapter 56 public network adapter 61 required 45 Email notification setting for failover cluster 98 L F License agreement (failover server) 87 License requirements failover cluster system 21
ABCDEFGHIJKLMNOPQRSTUVWXYZ P QLogic card setting link speed 49 setting spinup delay 49 Quorum disk 43 configuring 63 verifying 74 Quorum resource defined 41 Server Execution User changing 111 specifying for failover cluster 96 Server Failover overview 15 See also Failover cluster Service name examples for failover cluster 45 Services dependencies 111 Shared drive configuring for failover cluster 61 specifying for Interplay Engine 89 Slot locations SR2400 server (failover cluster) 23 SR2500 server (failov
Index ABCDEFGHIJKLMNOPQRSTUVWXYZ W Web servers disabling 85 Windows Cluster Administrator console 16 118
