Dell MD3600f, MD3620f Owner's Manual

Download

Page 1

Dell PowerVault MD3600f and

MD3620f Storage Arrays With

Microsoft Windows Server

Failover Clusters

Hardware Installation

and Troubleshooting

Guide

Page 2

Notes and Cautions

NOTE: A NOTE indicates important information that helps you make better use of

your computer.

CAUTION: A CAUTION indicates potential damage to hardware or loss of data if

instructions are not followed.

__________________

Reproduction of these materials in any manner whatsoever without the written permission of Dell Inc. is strictly forbidden.

Trademarks used in this text: Dell™, the DELL logo, PowerEdge™, and PowerVault™ are trademarks of Dell Inc. Microsoft Corporation in the United States and/or other countries.

Other trademarks and trade names may be used in this publication to refer to either the entities claiming the marks and names or their products. Dell Inc. disclaims any proprietary interest in trademarks and trade names other than its own.

2011 - 07 Rev. A00

, Windows®, and Windows Server® are registered trademarks of Microsoft

Page 3

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . 7

Overview . . . . . . . . . . . . . . . . . . . . . . . . . . 7

Cluster Solution

. . . . . . . . . . . . . . . . . . . . . . 8

Cluster Requirements . . . . . . . . . . . . . . . . . . . 8

Cluster Nodes

. . . . . . . . . . . . . . . . . . . . . 9

Cluster Storage . . . . . . . . . . . . . . . . . . . 10

Cluster Storage Management Software

Supported Cluster Configurations

Direct-Attached Cluster

SAN-Attached Cluster

. . . . . . . . . . . . . . 12

. . . . . . . . . . . . . . . 13

. . . . . . . . . . . . 12

. . . . . . 10

Configuration Order for Direct-Attached and SAN-Attached Connections

Direct-Attached Configuration Order

. . . . . . . . . . . . . 14

. . . . . . . 15

SAN-Attached Configuration Order . . . . . . . . 15

Introduction

This document provides information for installing and managing your cluster solution using Dell PowerVault MD3600f and MD3620f storage systems. It is intended for experienced IT professionals who configure a cluster solution and for trained service technicians who perform upgrade and maintenance procedures. This document also addresses readers who are new to clustering.

Overview

The Dell PowerEdge Cluster with Microsoft Windows Server Failover Clustering combines specific hardware and software components to provide enhanced availability for applications and services that run on the cluster. A failover cluster is designed to reduce the possibility of any single point of failure within the system that can cause the clustered applications or services to become unavailable. It is recommended that you use redundant components like system and storage power supplies, connections between the nodes and the storage array(s), connections to client systems, or other systems in the multi-tier enterprise application architecture in your cluster.

This guide addresses the configuration of your Dell MD3600f and MD3620f Fibre

Channel storage arrays for use with one or more Windows Server failover clusters. It provides information and specific configuration tasks that enable you to deploy the shared storage for your cluster.

For more information on deploying your cluster, see the Dell Failover Clusters with Microsoft Windows Server Installation and Troubleshooting Guide at support.dell.com/manuals.

NOTE: Throughout this document:

• Windows Server 2008 refers to Microsoft Windows Server 2008 x64 Enterprise Edition or Microsoft Windows Server 2008 R2 x64 Enterprise Edition.

• Dell PowerVault MD36x0f storage array refers to both Dell PowerVault MD3600f and Dell PowerVault MD3620f storage arrays.

For a list of recommended operating systems, hardware components, and driver or firmware versions for your Dell Windows Server Failover Cluster, see the Dell Cluster Configuration Support Matrices at dell.com/ha.

Introduction 7

Page 8

Cluster Solution

Your Fibre Channel cluster implements a minimum of two-node clustering and a maximum of sixteen-node clustering and provides the following features:

• 8 Gbps and 4 Gbps Fibre Channel technology.

• High availability of system services and resources to network clients.

• Redundant paths to the shared storage.

• Failure recovery for applications and services.

• Flexible maintenance capabilities, allowing you to repair, maintain, or upgrade a cluster node without taking the entire cluster offline.

Implementing advantages:

•

Flexibility

switches without degrading the signal.

•

Availability

providing multiple data paths and greater availability for clients.

•

Connectivity

Computer System Interface (SCSI) or Serial Attached SCSI (SAS). Because Fibre Channel devices are hot swappable, you can add or remove devices from the nodes without bringing down the cluster.

Fibre

Channel technology in a cluster provides the following

—Fibre Channel allows a distance of up to 10 km between

—Fibre Channel components use redundant connections,

—Fibre Channel allows more device connections than Small

Cluster Requirements

Your cluster requires the following components:

• Servers (cluster nodes)

• Storage and storage management software

8 Introduction

Page 9

Cluster Nodes

Table 1-1 lists hardware requirements for the cluster nodes.

Table 1-1. Cluster Node Requirements

Component Minimum Requirement

Processor At least one processor for each cluster node.

Cluster Nodes A minimum of two identical PowerEdge systems.

RAM At least 1 GB RAM on each cluster node.

Host Bus Adapter (HBA) ports

Network Interface Cards (NICs) (public and private)

Internal Disk Controller

Two Fibre Channel HBAs per node, unless the server employs an integrated or supported dual-port Fibre Channel HBA.

At least two NICs—one NIC for the public network and another NIC for the private network.

NOTE: It is recommended that the NICs on each public network are

identical and that the NICs on each private network are identical.

One controller connected to internal disks for each node. Use any supported Redundant Array of Independent Disks (RAID) controller or disk controller.

Two physical disks are required for mirroring (RAID 1) and at least three are required for disk striping with parity (RAID 5).

NOTE: It is recommended that you use hardware-based RAID or

software-based disk-fault tolerance for the internal drives.

NOTE: For more information about supported systems, HBAs, and operating system

versions, see the Dell Cluster Configuration Support Matrices at dell.com/ha.

Introduction 9

Page 10

Cluster Storage

Table 1-2 provides the configuration requirements for the shared storage system.

Table 1-2. Cluster Storage Requirements

Hardware Components

Supported storage systems

Minimum Requirement

• One Dell PowerVault MD3600f or MD3620f RAID enclosure.

• Any combination of up to seven Dell PowerVault MD1200 and/or MD1220 expansion enclosures.

NOTE: The number of hard drives must not exceed 96.

Cluster nodes All nodes must be directly attached to a single storage system

or attached to one or more storage systems through a SAN.

Switch and cable At least two 8 Gbps Fibre Channel switches in a SAN-attached

environment.

Power and cooling requirements

Physical disks At least two physical disks in the PowerVault MD3600f or

Multiple clusters and stand-alone systems

NOTE: RAID 0 and independent disks can be used but are not recommended for a

high-availability system because they do not offer data redundancy if a disk failure occurs.

Two integrated hot-swappable power supply/cooling fan modules.

MD3620f RAID enclosure.

In a switch-attached configuration, clusters and stand-alone systems can share one or more PowerVault MD3600f or MD3620f systems.

Cluster Storage Management Software

Dell PowerVault Modular Disk Storage Manager

The software runs on the management station or any host attached to the array to centrally manage the PowerVault MD3600f and MD3620f RAID enclosures. You can use Dell PowerVault Modular Disk Storage Manager (MDSM) to perform tasks such as creating disk groups, creating and mapping virtual disks, monitoring the enclosure status, and downloading firmware.

10 Introduction

Page 11

MDSM is a graphical user interface (GUI) with wizard-guided tools and a task-based structure. MDSM is designed to:

• Reduce the complexity when you install, configure, manage, and perform diagnostic tasks for the storage arrays.

• Contain an event monitoring service that is used to send alerts when a critical problem with the storage array occurs.

• Provide a command line interface (CLI) to run commands from an operating system prompt.

Modular Disk Storage Manager Agent

This software resides on each cluster node to collect system-based topology data that can be managed by the MDSM.

Multipath I/O (MPIO) Software

Multipath I/O software (also referred to as the failover driver) is installed on each cluster node. The software manages the redundant data path between the system and the RAID enclosure. For the MPIO software to correctly manage a redundant path, the configuration must provide for redundant HBAs and cabling.

The MPIO software identifies the existence of multiple paths to a virtual disk and establishes a preferred path to that disk. If any component in the preferred path fails, the MPIO software automatically re-routes I/O requests to the alternate path so that the storage array continues to operate without interruption.

Introduction 11

Page 12

Advanced Features

Advanced features for the PowerVault MD3600f and MD3620f RAID storage systems include:

•

Snapshot Virtual Disk

for backup, testing, or data processing without affecting the contents of the source virtual disk.

•

Virtual Disk Copy

disk to the target virtual disk in a storage array. You can use Virtual Disk Copy to back up data, copy data from disk groups that use smaller-capacity physical disks to disk groups using greater capacity physical disks, or restore snapshot virtual disk data to the source virtual disk.

•

Upgrading to High-Performance Tier

system beyond that of a MD36 performance level.

•

Remote Replication

storage arrays in separate locations.

NOTE: For more information on deploying the correct options in the cluster

environment, see "Using Advanced (Premium) PowerVault Modular Disk Storage Manager Features" on page 63.

—Captures point-in-time images of a virtual disk

—Generates a full copy of data from the source virtual

—Increases the performance of the

0f series array operating at the standard

—Enables real-time replication of data between two

Supported Cluster Configurations

The following sections describe the supported cluster configurations.

Direct-Attached Cluster

In a direct-attached cluster, all the nodes of the cluster are directly attached to a single storage system. In this configuration, the RAID controllers on the storage system are connected by cables directly to the Fibre Channel HBA ports in the nodes.

Figure 1-1 shows a basic direct-attached, single-cluster configuration.

12 Introduction

Page 13

Figure 1-1. Direct-Attached, Single-Cluster Configuration

private network

public network

cluster node

Fibre Channel connections

cluster node

storage system

NOTE: The configuration can have up to 4 nodes. The nodes can be:

• One cluster (up to 4 nodes)

• Two two-node clusters

• One cluster and stand-alone server(s)

SAN-Attached Cluster

In a SAN-attached cluster, all nodes are attached to a single storage system or to multiple storage systems through a SAN using redundant switch fabrics. SAN-attached clusters are superior to direct-attached clusters in configuration flexibility, expandability, and performance.

Figure 1-2 shows a SAN-attached cluster.

Introduction 13

Page 14

Figure 1-2. SAN-Attached Cluster

cluster node

private network

Fibre Channel connections

storage system

Fibre Channel switch

public network

Fibre Channel connections

cluster node

NOTE: The configuration can have up to 64 nodes. The nodes can be:

• One cluster (up to 16 nodes)

• Multiple clusters

• Multiple clusters and stand-alone server(s)

Configuration Order for Direct-Attached and SAN-Attached Connections

This section describes the configuration steps for both direct-attached and SAN-attached connections. These steps assume that you are setting up a Fibre Channel storage for the first time.

NOTE: If you are adding a Fibre Channel storage array or if your host server is

already configured to access Fibre Channel storage, some of the steps in this section may not apply. Before proceeding, see the Dell Cluster Configuration Support Matrices at dell.com/ha to confirm that your existing hardware components and Host Bus Adapter (HBA) firmware and BIOS levels are supported.

14 Introduction

Page 15

Direct-Attached Configuration Order

Install the supported HBAs on your cluster nodes. See "Installing Supported Fibre Channel HBAs" on page 44.

Cable the cluster nodes to the storage array. See "Cabling Storage for Your Direct-Attached Cluster" on page 24.

Install the required HBA drivers and firmware versions listed in the

Cluster Configuration Support Matrices

Install and configure the MD Storage Manager software (included with

dell.com/ha

your storage array) on your cluster nodes. See "Installing the Storage Management Software" on page 47.

Using MD Storage Manager (MDSM), configure the host servers, storage arrays, and virtual disks. See "Configuring the Shared Storage System" on page 48.

Activate and configure premium features (if applicable).

SAN-Attached Configuration Order

NOTE: A SAN-attached configuration is required to use the Remote Replication

premium feature. Remote Replication is not supported in direct-attached configurations.

Install the supported HBAs on your cluster nodes. See "Installing Supported Fibre Channel HBAs" on page 44.

Cable the cluster nodes to the Fibre Channel switches. See "Cabling a SAN-Attached Cluster to an MD36x0f Storage System" on page 33 and "Remote Replication" on page 65.

Install the required HBA drivers and firmware versions listed in the

Cluster Configuration Support Matrices

Install and configure the MD Storage Manager software (included with your storage array) on your cluster nodes. See "Installing the Storage Management Software" on page 47.

Cable the storage array to the Fibre Channel switches. See "Cabling Storage for Your SAN-Attached Cluster" on page 30.

Configure zoning on all Fibre Channel switches. See "Setting Up Zoning on the Fibre Channel Switch Hardware" on page 47.

dell.com/ha

Dell

Introduction 15

Page 16

NOTE: All equipment attached to the switch must be powered on before

establishing zoning. For additional switch hardware requirements, see the manufacturer’s documentation.

Using MDSM, configure the cluster nodes, storage arrays, and virtual disks. See "Configuring the Shared Storage System" on page 48.

Activate and configure premium features (if applicable).

Cabling Your Cluster Hardware

The following sections provide information on cabling various components of your cluster.

Cabling the Mouse, Keyboard, and Monitor

When installing a cluster configuration in a rack, you must include a switch box to connect the mouse, keyboard, and monitor to the nodes. See the documentation included with your rack for instructions on cabling each node's connections to the switch box.

Cabling the Power Supplies

To ensure that the specific power requirements are satisfied, see the documentation for each component in your cluster solution.

It is recommended that you adhere to the following guidelines to protect your cluster solution from power-related failures:

• For nodes with multiple power supplies, plug each power supply into a separate AC circuit.

• Use uninterruptible power supplies (UPS).

• For some environments, consider having backup generators and power from separate electrical substations.

Figure 2-1 illustrates a recommended method for power cabling of a cluster solution consisting of two Dell PowerEdge systems and one storage system. To ensure redundancy, the primary power supplies of all the components are grouped onto one or two circuits and the redundant power supplies are grouped onto a different circuit.

Cabling Your Cluster Hardware 19

Page 20

Figure 2-1. Power Cabling Example

primary power supplies on one AC power strip [or one AC PDU (not shown)]

redundant power supplies on one AC power strip [or one AC PDU (not shown)]

MD36x0f RAID controller module 0

MD36x0f RAID controller module 1

cluster node 1

cluster node 2

NOTE: This illustration is intended only to demonstrate the power distribution of the

components.

20 Cabling Your Cluster Hardware

Page 21

Cabling Your Public and Private Networks

The network adapters in the cluster nodes provide at least two network connections for each node. These connections are described in Table 2-1.

Table 2-1. Network Connections

Network Connection Description

Public Network All connections to the client LAN.

At least one public network must be configured for mixed mode (public mode and private mode) for private network failover.

Private Network A dedicated connection for sharing cluster health and

status information between the cluster nodes.

Network adapters connected to the LAN can also provide redundancy at the communications level in case the cluster interconnect fails.

For more information on private network redundancy, see your Microsoft Failover Clustering documentation.

Figure 2-2 shows an example of network adapter cabling in which dedicated network adapters in each node are connected to the public network and the remaining network adapters are connected to each other (for the private network).

Cabling Your Cluster Hardware 21

Page 22

Figure 2-2. Example of Network Cabling Connection

public network

private network adapter

cluster node 1 cluster node 2

private network

Cabling Your Public Network

Any network adapter supported by a system running TCP/IP may be used to connect to the public network segments. You can install additional network adapters to support additional public network segments or to provide redundancy in the event of a faulty primary network adapter or switch port.

Cabling Your Private Network

The private network connection to the cluster nodes is provided by a second or subsequent network adapter that is installed in each node. This network is used for intra-cluster communications.

Table 2-2 lists the required hardware components and connection method for three possible private network configurations.

22 Cabling Your Cluster Hardware

Page 23

Table 2-2. Private Network Hardware Components and Connections

Method Hardware Components Connection

Network switch Gigabit or 10 Gigabit

Ethernet network adapters and switches

Point-to-Point (two node cluster only)

NOTE: Throughout this document, Ethernet refers to either Gigabit Ethernet or

10 Gigabit Ethernet.

Copper Gigabit or 10 Gigabit Ethernet network adapters with RJ-45 connectors

Copper 10 Gigabit Ethernet network adapters with SFP+ connectors

Optical Gigabit or 10 Gigabit Ethernet network adapters with LC connectors

Depending on the hardware, connect the CAT5e or CAT6 cables, the multimode optical cables with Local Connectors (LCs), or the twinax cables from the network adapters in the nodes to a switch.

Connect a standard CAT5e or CAT6 Ethernet cable between the network adapters in both nodes.

Connect a twinax cable between the network adapters in both nodes.

Connect a multi-mode optical cable between the network adapters in both nodes.

Using Dual-Port Network Adapters for Your Private Network

You can configure your cluster to use the public network as a failover for private network communications. However, if dual-port network adapters are used, do not use two ports simultaneously to support both the public and private networks.

Cabling Your Cluster Hardware 23

Page 24

NIC Teaming

Network Interface Card (NIC) teaming combines two or more NICs to provide load balancing and/or fault tolerance. Your cluster supports NIC teaming, but only in a public network; NIC teaming is not supported in a private network.

NOTE: Use the same brand of NICs in a team. You cannot mix brands of teaming

drivers.

Cabling the Storage Systems

This section provides information on cabling your cluster to a storage system in a direct-attached configuration or to one or more storage systems in a SAN-attached configuration.

Cabling Storage for Your Direct-Attached Cluster

A direct-attached cluster configuration consists of redundant Fibre Channel host bus adapter (HBA) ports cabled directly to a Dell PowerVault MD36x0f storage system. If a component (for example, port, cable, or the storage controller) fails in the storage path, the MPIO software automatically re-routes the I/O requests to the alternate path so that the storage array continues to operate without interruption. The configuration with two single-port HBA provides higher availability. An HBA failure does not cause the failover cluster to move cluster resources to the other cluster node.

Figure 2-3 shows an example of a direct-attached, single cluster configuration with redundant HBA ports installed in each cluster node.

24 Cabling Your Cluster Hardware

Page 25

Figure 2-3. Direct-Attached Cluster Configuration

cluster node

private network

public network

cluster node

storage system

Fibre Channel connections

Cabling Your Cluster Hardware 25

Page 26

Cabling a Cluster to a Dell PowerVault MD36x0f Storage System

Each cluster node attaches to the storage system using two Fibre optic cables with duplex local connector (LC) multimode connectors that attach to the HBA ports in the cluster nodes and the MD36x0f RAID controller module ports in the MD36x0f storage system. These connectors consist of two individual Fibre optic connectors with indexed tabs that must be aligned properly into the HBA ports and the MD36x0f RAID controller module.

CAUTION: Do not remove the connector covers until you are ready to insert the

connectors into the HBA port, MD36x0f RAID controller module port, or tape library port.

Cabling a Two-Node Direct-Attached Cluster to an MD36x0f Storage System

Connect cluster node 1 to the storage system:

Install a cable from cluster node 1 HBA port 0 to the fibre channel port on MD36

Install a cable from cluster node 1 HBA port 1 to the fibre channel port on MD36

Connect cluster node 2 to the storage system:

Install a cable from cluster node 2 HBA port 0 to the fibre channel port on MD36

Install a cable from cluster node 2 HBA port 1 to the fibre channel port on MD36

0f RAID controller module 0, port In-0.

0f RAID controller module 1, port In-0.

0f RAID controller module 0, port In-1.

0f RAID controller module 1, port In-1.

26 Cabling Your Cluster Hardware

Page 27

Figure 2-4. Cabling a Two-Node Direct-Attached Cluster to a PowerVault MD36x0f

Corporate, public, or private network

cluster node1

cluster node2

SAS out port (2)

RAID controller module 0

RAID controller module 1

Storage System

NOTE: The SAS out port provides SAS connection for cabling to MD1200 or

MD1220 expansion enclosure(s).

Cabling a Four-Node Direct-Attached Cluster to an MD36x0f Storage System

You can configure a cluster with more than two nodes in a direct-attached configuration using a PowerVault MD36x0f storage system. The MD36x0f storage system can support up to four-node direct-attached cluster.

The following example explains how to cable a four-node direct-attached cluster:

Connect cluster node 1 to the storage system:

Install a cable from cluster node 1 HBA port 0 to the Fibre Channel port on the MD36

0f RAID controller module 0, port In-0.

Cabling Your Cluster Hardware 27

Page 28

Install a cable from cluster node 1 HBA port 1 to the Fibre Channel port on the MD36

Connect cluster node 2 to the storage system:

Install a cable from cluster node 2 HBA port 0 to the Fibre Channel port on the MD36

Install a cable from cluster node 2 HBA port 1 to the Fibre Channel port on the MD36

Connect cluster node 3 to the storage system:

Install a cable from cluster node 3 HBA port 0 to the Fibre Channel port on the MD36

Install a cable from cluster node 3 HBA port 1 to the Fibre Channel port on the MD36

Connect cluster node 4 to the storage system:

Install a cable from cluster node 4 HBA port 0 to the Fibre Channel port on the MD36

Install a cable from cluster node 4 HBA port 1 to the Fibre Channel port on the MD36

0f RAID controller module 1, port In-0.

0f RAID controller module 0, port In-1.

0f RAID controller module 1, port In-1.

0f RAID controller module 0, port In-2.

0f RAID controller module 1, port In-2.

0f RAID controller module 0, port In-3.

0f RAID controller module 1, port In-3.

28 Cabling Your Cluster Hardware

Page 29

Figure 2-5 shows a four-cluster node connected to two RAID controller

PowerVault MD360x0f series storage array

Corporate, public, or private network

modules. Since each cluster node has redundant paths, loss of a single path still allows access to the storage array through the alternate path.

Figure 2-5. Cabling a Four-Node Direct-Attached Cluster to a PowerVault MD36x0f Storage System

Cabling Multiple Clusters to an MD36x0f Storage System

The high number of available Fibre Channel ports on the system also allows you to configure four-node clusters, 2 two-node clusters or a mix of cluster(s) and non-clustered server(s) in a direct-attached configuration.

Cabling Your Cluster Hardware 29

MD36x0f

storage

Page 30

Cabling Two Two-Node Direct-Attached Clusters to an MD36x0f Storage System

The following steps are an example of how to cable a two two-node cluster.

In the first cluster, connect cluster node 1 to the storage system:

Install a cable from cluster node 1 HBA port 0 to the Fibre Channel port on the MD36

Install a cable from cluster node 1 HBA port 1 to the Fibre Channel port on the MD36

In the first cluster, connect cluster node 2 to the storage system:

Install a cable from cluster node 2 HBA port 0 to the Fibre Channel port on the MD36

Install a cable from cluster node 2 HBA port 1 to the Fibre Channel port on the MD36

In the second cluster, connect cluster node 1 to the storage system:

Install a cable from cluster node 1 HBA port 0 to the Fibre Channel port on the MD36

Install a cable from cluster node 1 HBA port 1 to the Fibre Channel port on the MD36

In the second cluster, connect cluster node 2 to the storage system:

Install a cable from cluster node 2 HBA port 0 to the Fibre Channel port on the MD36

Install a cable from cluster node 2 HBA port 1 to the Fibre Channel port on the MD36

0f RAID controller module 0, port In-0.

0f RAID controller module 1, port In-0.

0f RAID controller module 0, port In-1.

0f RAID controller module 1, port In-1.

0f the RAID controller module 0, port In-2.

0f RAID controller module 1, port In-2.

0f RAID controller module 0, port In-3.

0f RAID controller module 1, port In-3.

Cabling Storage for Your SAN-Attached Cluster

A SAN-attached cluster is a cluster configuration where all cluster nodes that are attached to a single storage system or to multiple storage systems through SAN use a redundant switch fabric.

SAN-attached cluster configurations provide more flexibility, expandability, and performance than direct-attached configurations.

For more information on Fibre Channel switch fabrics, see "Implementing Zoning on a Fibre Channel Switched Fabric" on page 45.

Figure 2-6 shows an example of a two node SAN-attached cluster.

Figure 2-7 shows an example of an sixteen-node SAN-attached cluster.

30 Cabling Your Cluster Hardware

Page 31

Similar cabling concepts can be applied to clusters that contain a different

cluster node

private network

Fibre Channel connections

storage system

Fibre Channel switch

public network

Fibre Channel connections

number of nodes.

NOTE: The connections listed in this section are representative of one proven

method of ensuring redundancy in the connections between the cluster nodes and the storage system. Other methods that achieve the same type of redundant connectivity may be acceptable.

Figure 2-6. Two-Node SAN-Attached Cluster

Cabling Your Cluster Hardware 31

Page 32

Figure 2-7. Sixteen-Node SAN-Attached Cluster

public network

storage system

cluster nodes (2-16)

Fibre Channel switch

private network

32 Cabling Your Cluster Hardware

Page 33

Cabling a SAN-Attached Cluster to an MD36x0f Storage System

The cluster nodes attach to the storage system using a redundant switch fabric and Fibre optic cables with duplex LC multimode connectors.

The switches, the HBA ports in the cluster nodes, and the RAID controller module ports in the storage system use duplex LC multimode connectors. The connectors consist of two individual fibre optic connectors with indexed tabs that must be inserted and aligned properly in the small form-factor pluggable (SFP) module connectors on the Fibre Channel switches and the connectors on the cluster nodes and storage systems.

Each HBA port is cabled to a port on a Fibre Channel switch. One or more cables connect from the outgoing ports on a switch to a RAID controller module on an MD36x0f storage system.

NOTE: Adding more cables from the storage system to the switches can increase

the I/O bandwidth and high availability of data.

Figure 2-8 illustrates a method for cabling a SAN-attached cluster to an MD36x0f storage system.

Cabling a SAN-Attached Cluster to the MD36x0f Storage System

Connect cluster node 1 to the SAN:

Connect a cable from HBA port 0 to the Fibre Channel switch (Switch fabric A).

Connect a cable from HBA port 1 to the Fibre Channel switch (Switch fabric B).

Repeat step 1 for each additional cluster node.

Connect the storage system to the SAN:

Connect a cable from Fibre Channel switch (Switch fabric A) to the Fibre Channel port on the MD36

0f RAID controller module 1, port

In-1.

Connect a cable from Fibre Channel switch (Switch fabric A) to the Fibre Channel port on the MD36

0f RAID controller module 1, port

In-3.

Connect a cable from Fibre Channel switch (Switch fabric A) to the Fibre Channel port on the MD36

0f RAID controller module 0, port

In-0.

Cabling Your Cluster Hardware 33

Page 34

Connect a cable from Fibre Channel switch (Switch fabric A) to the Fibre Channel port on the MD36

0f RAID controller module 0, port

In-2.

Connect a cable from Fibre Channel switch (Switch fabric B) to the Fibre Channel port on the MD36

0f RAID controller module 0, port

In-1.

Connect a cable from Fibre Channel switch 1 (Switch fabric B) to the Fibre Channel port on the MD36

0f RAID controller module 0, port

In-3.

Connect a cable from Fibre Channel switch 1 (Switch fabric B) to the Fibre Channel port on the MD36

0f RAID controller module 1, port

In-0.

Connect a cable from Fibre Channel switch 1 (Switch fabric B) to the Fibre Channel port on the MD36

0f RAID controller module 1, port

In-2.

34 Cabling Your Cluster Hardware

Page 35

Figure 2-8. Cabling a SAN-Attached Cluster to the MD36x0f

cluster node 1

cluster node 2

switch fabric A

switch fabric B

corporate, public, or private network

RAID controller module 0

RAID controller module 1

Cabling Your Cluster Hardware 35

Page 36

Table 2-3. Zoning Configuration Information

Switch Fabric A Switch Fabric B

Zone_1_Cluster Node_1_HBA_0 Zone_3_Cluster Node_1_HBA_1

Cluster Node_1_HBA_0 Cluster Node_1_HBA_1

Array_Ctrl-0-0 Array_Ctrl-0-1

Array_Ctrl-0-2 Array_Ctrl-0-3

Array_Ctrl-1-1 Array_Ctrl-1-0

Array_Ctrl-1-3 Array_Ctrl-1-2

Zone_2_Cluster Node_2_HBA_0 Zone_4_Cluster Node_2_HBA_1

Cluster Node_2_HBA_0 Cluster Node_2_HBA_1

Array_Ctrl-0-0 Array_Ctrl-0-1

Array_Ctrl-0-2 Array_Ctrl-0-3

Array_Ctrl-1-1 Array_Ctrl-1-0

Array_Ctrl-1-3 Array_Ctrl-1-2

Cabling Multiple SAN-Attached Clusters to an MD36x0f Storage System

To cable multiple clusters to the storage system, connect the cluster nodes to the appropriate Fibre Channel switches and then connect the Fibre Channel switches to the appropriate MD36x0f RAID controller modules.

For rules and guidelines for SAN-attached clusters, see the Dell Cluster Configuration Support Matrices at dell.com/ha.

Connecting a PowerEdge Cluster to a Tape Library

To provide additional backup for your cluster, you can add tape backup devices to your cluster configuration. The Dell PowerVault tape libraries may contain an integrated Fibre Channel bridge or Storage Network Controller (SNC) that connects directly to your Fibre Channel switch.

Figure 2-9 shows a supported Failover Cluster configuration using redundant Fibre Channel switches and a tape library. In this configuration, each of the cluster nodes can access the tape library to provide backup for your local disk resources, as well as your cluster disk resources. This configuration allows you to add more servers and storage systems in the future, if needed. For more information on configuring these components, see the storage and tape backup documentation at support.dell.com/manuals.

36 Cabling Your Cluster Hardware

Page 37

NOTE: While tape libraries can be connected to multiple fabrics, they do not

cluster node

tape library

storage system

cluster node

private network

Fibre Channel switch

provide path failover.

Figure 2-9. Cabling a Storage System and a Tape Library

Cabling Your Cluster Hardware 37

Page 38

Connecting a PowerEdge Cluster to Multiple PowerVault MD3600f or MD3620f Storage Systems

You can increase your cluster storage capacity by attaching multiple storage systems to your cluster using redundant Fibre Channel switch fabrics. The PowerEdge cluster systems support configurations with multiple PowerVault MD3600f or MD3620f storage systems attached to clustered systems. In this scenario, the Failover Cluster software can fail over disk drives in any clusterattached shared storage system between the cluster nodes.

38 Cabling Your Cluster Hardware

Page 39

public network

private network

2 to 16 cluster nodes

switch fabric A

MD36x0f RAID controller module 0

MD36x0f RAID controller module 1

storage array 1

storage array 2

switch fabric B

Figure 2-10. SAN-Attached Cluster Configuration With Multiple Storage Arrays

Cabling Your Cluster Hardware 39

Page 40

When attaching multiple PowerVault MD3600f and MD3620f storage systems with your cluster, the following rules apply:

• A maximum of four Power Vault

MD3600f and MD3620f

storage systems

per cluster.

• The shared storage systems and firmware must be identical. Using dissimilar storage systems and firmware for your shared storage is not supported.

• Microsoft Windows limits access to drives using limited drive letters which is 22. A maximum of 22 drive letters (E to Z) can be used for your storage system disks.

NOTE: Drive letters A through D are reserved for local disks.

• Windows Server 2008 Enterprise Edition supports mount points, allowing greater than 22 drives per cluster.

40 Cabling Your Cluster Hardware

Page 41

Preparing Your Systems for Clustering

CAUTION: Only trained service technicians are authorized to remove and access

any of the components inside the system. See the safety information that shipped with your system for complete information about safety precautions, working inside the system, and protecting against electrostatic discharge.

Cluster Configuration Overview

Ensure that your site can handle the cluster’s power requirements.

Contact your sales representative for information about your region's power requirements.

Install the servers, the shared storage array(s), and the interconnect switches (for example, in an equipment rack), and ensure that all these components are turned on.

NOTE: For more information on step 3 through step 7 and step 10 through

step 12, see the “Preparing your systems for clustering” section of the

Dell Failover Clusters with Microsoft Windows Server 2008 Installation and Troubleshooting Guide at support.dell.com/manuals.

Deploy the operating system (including any relevant service pack and hotfixes), network adapter drivers, and storage adapter drivers (including the MPIO software) on each of the servers that must become cluster nodes. Depending on the deployment method that is used, it may be necessary to provide a network connection to successfully complete this step.

NOTE: You can record the Cluster configuration to the Cluster Data Form to

help in planning and deployment of your cluster. For more information, see the "Cluster Data Form" on page 73 and "Zoning Configuration Form" on page 77.

Establish the physical network topology and the TCP/IP settings for network adapters on each server node to provide access to the cluster public and private networks.

Preparing Your Systems for Clustering 41

Page 42

Configure each server node as a member server in the same Windows Active Directory Domain.

NOTE: You can configure the cluster nodes as Domain Controllers. For more

information, see the “Selecting a Domain Model” section of the Dell Failover

Clusters with Microsoft Windows Server 2008 Installation and Troubleshooting Guide at support.dell.com/manuals.

Establish the physical storage topology and any required storage network settings to provide connectivity between the storage array and the servers that must be configured as cluster nodes. Configure the storage system(s) as described in your storage system documentation.

Use storage array management tools to create at least one logical unit number (LUN). The LUN is used as a witness disk for Microsoft Windows Server 2008 Failover cluster. Ensure that this LUN is presented to the servers that must be configured as cluster nodes.

NOTE: It is recommended that you configure the LUN on a single node, for

security reasons, as mentioned in step 8 when you are setting up the cluster. Later, you can configure the LUN as mentioned in step 9 so that other cluster nodes can access it.

Select one of the systems and form a new failover cluster by configuring the cluster name, cluster management IP, and quorum resource. For more information, see "Preparing Your Systems for Clustering" on page 41.

NOTE: For Windows Server 2008 Failover Clusters, run the Cluster Validation

Wizard to ensure that your system is ready to form the cluster.

Join the remaining node(s) to the failover cluster. For more information, see "Preparing Your Systems for Clustering" on page 41.

Configure roles for cluster networks. Take any network interfaces that are used for other purposes outside of the cluster and out of the control of the cluster.

42 Preparing Your Systems for Clustering

Page 43

Test the failover capabilities of your new cluster.

NOTE: You can also use the Cluster Validation Wizard.

Configure highly-available applications and services on your failover cluster. Depending on your configuration, this may also require providing additional LUNs to the cluster or creating new cluster resource groups. Test the failover capabilities of the new resources.

Configure client systems to access the highly available applications and services that are hosted on your failover cluster.

Installation Overview

Each node in your Dell Windows Server failover cluster must have the same release, edition, service pack, and processor architecture of the Windows Server operating system installed. For example, all nodes in your cluster may be configured with Windows Server 2008 R2, Enterprise x64 Edition. If the operating system varies among nodes, it is not possible to configure a failover cluster successfully. It is recommended to establish system roles prior to configuring a failover cluster, depending on the operating system configured on your cluster.

For more information on deploying your cluster with the Windows Server 2008 operating systems, see the Dell Failover Clusters with Microsoft Windows Server 2008 Installation and Troubleshooting Guide at support.dell.com/manuals.

Preparing Your Systems for Clustering 43

Page 44

The following sub-sections describe steps that enable you to establish communication between the cluster nodes and your shared MD3600f or MD3620f storage array(s), and to present disks from the storage array to the cluster:

Installing Supported Fibre Channel HBAs

Implementing Zoning on a Fibre Channel Switched Fabric

Installing the Storage Management Software

Configuring the Shared Storage System

Configuring a Failover Cluster

Installing Supported Fibre Channel HBAs

The storage array supports a specific set of Fibre Channel HBAs, each requiring a unique driver and firmware level. For more information on supported HBAs with the drivers, firmware, and BIOS versions, see the Dell Cluster Configuration Support Matrices at dell.com/ha.

If an HBA is already installed on your host server, verify that it is listed in the Support Matrices and operates at the driver and/or firmware level specified. If the HBA is not listed in the Support Matrices, it should not be used to connect to the storage array.

CAUTION: Connecting the storage array to a non-supported host server HBA or

installing non-qualified HBA drivers or firmware can cause instability or loss of access to your data.

Installing Fibre Channel HBA on Your Host Server

For instructions about physically installing an HBA in your host server and loading supported drivers and firmware, see the documentation supplied with your HBA hardware.

Customizing HBA Drivers and Firmware Settings

In most cases, the default manufacturer HBA driver and firmware settings are sufficient to support Fibre Channel communication with the storage array.

44 Preparing Your Systems for Clustering

Page 45

However, in the case of certain HBA manufacturers, the Dell Cluster Configuration Support Matrices at dell.com/ha may contain information

about specific driver and firmware settings necessary to optimize storage arrays. Review the Support Matrices to ensure that you are aware of any customized settings that may affect the performance of your HBA.

NOTE: Most Fibre Channel HBAs require a link timeout and node timeout value of

10 seconds. This value can be set using the HBA manufacturer's management utilities, for example, Emulex's OneCommand Manager and QLogic's Enterprise Fabric Suite.

For more information about installing and configuring HBAs, see the following:

•

Emulex HBAs Support website at

• QLogic HBAs—QLogic support website at website at

• Brocade HBAs—Brocade support website at Support website at

For more information about supported HBA controllers and drivers, see the Dell Cluster Configuration Support Matrices at dell.com/ha.

—Emulex support website at

support.dell.com

emulex.com

qlogic.com

brocade.com

or the Dell

or the Dell Support

or the Dell

Implementing Zoning on a Fibre Channel Switched Fabric

A Fibre Channel switch allows you to connect multiple devices to the SAN without compromising bandwidth or data throughput. Switch zoning implemented on each Fibre Channel switch hardware component using the manufacturer’s embedded utility, allows you to divide port traffic between the host server and storage array to maintain the highest level of redundancy.

Figure 2-8 shows each connection to the cluster node routed to an individual logical zone on the SAN. Two physical switches provide an alternate (redundant) data path to the storage array. If a single HBA, switch or RAID controller fails, the host server can still access data on the storage array through one of the alternate paths provided by the switches.

NOTE: For a list of supported Fibre Channel switches, see the Dell Cluster

Configuration Support Matrices at dell.com/ha.

Preparing Your Systems for Clustering 45

Page 46

World Wide Name Zoning

There are several different switch zoning techniques used across various SANs. When configuring zoning with your storage array, use 64 bit World Wide Name (WWN) port zoning to uniquely identify each component in your Fibre Channel switch fabric. An advantage of using WWN port zoning (also called soft zoning) is that it allows you to remove and/or replace cabling to and from the fabric without reconfiguring the existing zone logic.

When implemented on the Fibre Channel switch, port WWN zoning causes the host server to query the World Wide Identifier (WWID) name of all attached physical disks and/or expansion enclosures each time it access the switch fabric. By zoning specific ports on the host server, switch and storage array together, the switch allows the host server to see only devices included in that zone, decreasing the amount of time it would otherwise take for the host server to query attached, but out-of-zone, devices.

Switch Zoning Requirements

The storage array imposes specific requirements that must be followed when setting up Fibre Channel switch zoning:

• If a Fibre Channel switch is used to connect your host server and storage array, it must be zoned. Un-zoned or open switches cannot be used.

• Always use WWN port zoning, not based on the device.

• No more than one HBA may be connected to a single logical switch zone.

• No more than four paths (port-to-port segments) can be established from a single, physical host server to a single RAID controller.

• A zone can contain multiple targets.

• A zone can span multiple storage arrays.

• If the Remote Replication premium feature is activated, a separate zone for each replication port is required. Only data traffic related to Remote Replication can move through that zone. For more details and requirements regarding Remote Replication, see the

support.dell.com/manuals

hard zoning

that blocks or grants access

Owner’s Manual

46 Preparing Your Systems for Clustering

Page 47

Setting Up Zoning on the Fibre Channel Switch Hardware

Setting up zoning on Fibre Channel switches varies greatly between manufacturers. For detailed information on how to set up zoning on your switch, see the manufacturer’s product documentation or technical support website.

Table 3-1 provides a list of WWN identifiers that you can find in the Dell PowerVault MD36x0f cluster environment.

Table 3-1. Port Worldwide Names in a SAN Environment

Identifier Description

xx:xx:xx:2b:xx:xx:xx:xx PowerVault MD36x0f RAID controller module

xx:xx:00:00:C9:xx:xx:xx Emulex HBA ports

xx:xx:00:E0:8B:xx:xx:xx QLogic HBA ports (non-embedded)

xx:xx:00:0F:1F:xx:xx:xx Dell 2362M HBA port

xx:xx:xx:60:45:xx:xx:xx PowerVault 132T and 136T tape libraries

xx:xx:xx:E0:02:xx:xx:xx PowerVault 128T tape autoloader

xx:xx:xx:C0:01:xx:xx:xx PowerVault 160T tape library and Fibre

Channel tape drives

xx:xx:xx:C0:97:xx:xx:xx PowerVault ML6000 Fibre Channel tape drives

CAUTION: When you replace a Fibre Channel HBA in a PowerEdge server,

reconfigure your zones to provide continuous client data access. Additionally, when you replace a switch module, reconfigure your zones to prevent data loss or corruption.

Installing the Storage Management Software

The PowerVault MD3600f and MD3620f storage software installer provides features that include the core software, providers, and optional utilities. The core software feature includes the host-based storage agent, MPIO software, and MDSM application used to configure, manage and monitor the storage array solution. The providers feature includes a provider for the Microsoft Virtual Disk Service (VDS) and Microsoft Volume Shadow-Copy Service (VSS) frameworks.

Preparing Your Systems for Clustering 47

Page 48

NOTE: For more information about the Microsoft VDS, Microsoft VSS providers,

see the Dell PowerVault MD3600f and MD3620f Storage Arrays Owner's Manual.

Follow these steps to install the Storage Management Software:

Close all other programs before installing any new software.

Insert the PowerVault MD36x0f resource media.

Depending on the outrun settings of the operating system, the

PowerVault MD36

prompt may be displayed to run the PowerVault MD launcher is not displayed and there is no prompt to run the

md_launcher.exe

run the

Select

Select one of the following installation options:

• Full (recommended)—This package installs core software, providers,

• Host Only—This package includes the host-based storage agent,

• Management Station—This package includes the MD Storage

• Custom—This option allows you to select specific components.

• Reboot each host server.

md_launcher.exe

Install MD36x0f Storage Software

and utilities. It includes the necessary host-based storage agent, MPIO software, MD Storage Manager, providers, and optional utilities.

MPIO software, and optional utilities required to configure the host.

Manager, providers, and optional utilities.

0f Resource DVD

file, navigate to the root of the resource media and

file.

window may be displayed or a

md_launcher.exe

file. If the

Dell

Configuring the Shared Storage System

See Table 1-2 for a list of supported Dell Powervault MD36x0f storage systems. Before you begin configuring your storage system, you must fill out the "Zoning Configuration Form" on page 77. Gathering this type of information about your cluster node, HBAs, and storage before starting the configuration steps helps you complete the process faster.

48 Preparing Your Systems for Clustering

Page 49

Automatic Storage Array Discovery

Launch MDSM.

The

Enterprise Management

Management

•

Devices

•

Setup

window contains the following tabs:

tab—Provides information about the storage arrays.

tab—Presents the initial setup tasks that guide you through

window opens. The

Enterprise

adding storage arrays and configuring alerts.

If this is the first storage array to be set up, the

Add New Storage Array

window appears.

Select

Automatic

and click OK.

It may take several minutes for the discovery process to complete.

You can manage the array by launching the from the

Enterprise Management

window. The

Array Management

window

window provides management functions for a single storage array. You can have multiple

Array Management

windows open simultaneously to

manage different storage arrays.

To launch the

Enterprise Management

Array Management

window, and double-click on the relevant storage

window, click on the

Devices

tab from

array.

Defining a Host

If the Host Context Agent is running on the host, the hosts and the host ports connected to the storage array are automatically detected by MDSM and appear in the Mappings tab in the Array Management window, under the Default Group.

After the storage software is installed, all the hosts should show up. If one host is not detected, it can be manually added:

In the

Array Management

window, select the

Mappings

tab and select the

appropriate storage array.

Perform the following actions:

Select

Select the

MappingsDefineHost

Setup

tab and click

Mappings

tab.

Manually Define Hosts

Preparing Your Systems for Clustering 49

Page 50

Right-click the root node (storage array name), or

Host Group

the host, and select

The

Specify Host Name

Host name

Select the relevant option in

the this storage array

The

Specify Host Port Identifiers

Select the relevant option to add a host port identifier to the host, you can

node in the

Define Host

To po l og y

pane to which you want to add

from the pop-up menu.

window is displayed.

, enter an up to 30 character alphanumeric name.

Do you plan to use the storage partitions in

, and click

window is displayed.

Default Group

select:

•

Add by selecting a known unsolicited host port identifier unsolicited host port identifiers

, select the relevant host port

—In

identifier.

•

Add by creating a new host port identifier identifier

, enter a 16-character name and an

identifier (up to 30 characters only), and click

NOTE: The host port identifier name must contain only the letters A through F.

Click

The

Specify Host Type

Host

type, select the relevant operating system for the host.

The

Host Group Question

window is displayed.

—In

New host port

Alias

for the host port

Add

node,

Known

Yo u ca n s el e ct :

•

Yes

—this host shares access to the same virtual disks with other hosts.

•

—this host does NOT share access to the same virtual disks with

other hosts.

Click

If you select select

Enter the name of the host group or select an existing host group and click

The

Preview

Yes

, the

, go to step 11.

window is displayed.

Specify Host Group

window is displayed. If you

50 Preparing Your Systems for Clustering

Page 51

Click

Finish

Creating a Host Group

A host group is a logical entity of two or more hosts that share access to specific virtual disks on the storage array.

To create host groups:

In the

Array Management window

In the

To po l og y

Perform one of the following actions:

• Select

pane, select the storage array or the default group.

Mappings Define Host Group

• Right-click the storage array or

Host Group

Type the name of the new host group in the

from the pop-up menu.

, select the

Mappings

Default Group

Enter New Host Group Name

tab.

and select

Define

field.

Select the appropriate hosts in the

Add

Click OK. The host group is added to the storage array.

Creating Disk Groups and Virtual Disks

Select Hosts to Add Area

field and click

In some cases, the virtual disks may have been bound when the system was shipped. However, it is important that you install the management software and verify that the desired virtual disk configuration exists.

You can manage your virtual disks remotely using PowerVault Modular Disk Storage Manager. A minimum of one virtual disk is required for an active/passive cluster configuration and at least two virtual disks are required for an active/active cluster configuration.

Disk groups are created in the non-configured capacity of a storage array and virtual disks are created in the free capacity of a disk group. The hosts attached to the storage array read and write data to the virtual disks.

NOTE: Before you create virtual disks, you must first organize the physical disks

into disk groups and configure host access. You can then create virtual disks within a disk group.

Preparing Your Systems for Clustering 51

Page 52

To create a virtual disk, use one of the following methods:

• Automatic configuration

• Manual configuration

Create disk groups using automatic configuration as follows:

To start the

• To create a disk group from unconfigured capacity in the storage

array—On the

Disk Group Create

Alternatively, you can right-click the and select

• To create a disk group from unassigned physical disks in the storage

array—On the disks of the same physical disk type, and select

Alternatively, you can right-click the unassigned physical disks, and select

Create Disk Group Wizard

Logical

Create Disk Group

Physical

Create Disk Group

tab, select

from the pop-up menu.

tab, select one or more unassigned physical

from the pop-up menu.

, perform one of these actions:

Unconfigured Capacity

Disk Group Create

node

node

• To create a secure disk group—On the

unassigned security capable physical disks of the same physical disk type, and select

Alternatively, you can right-click the unassigned security capable physical disks, and select The

Create Disk Group

Click

. The

displayed.

Type a name (up to 30 characters) for the disk group in field.

Select the appropriate configuration method of Physical Disk selection from the following:

• Automatic (see step 6)

• Manual (see step 7)

Click

For automatic configuration, the displayed.

52 Preparing Your Systems for Clustering

Disk Group Create

Create Disk Group

window is displayed.

Disk Group Name and Physical Disk Selection

RAID Level and Capacity

Physical

tab, select one or more

from the pop-up menu.

Disk Group Name

window is

Page 53

Select the appropriate RAID level in the

Select Raid Level

can select RAID levels 0, 1/10, 6, and 5. Depending on your RAID level selection, the physical disks available for the selected RAID level is displayed in the

In the

Select Capacity

and click

For manual configuration, the

Select Capacity

Finish

table.

table, select the relevant disk group capacity

Manual Physical Disk Selection

displayed.

Select the appropriate RAID level in Select RAID level. You can select RAID levels 0, 1/10, 6, and 5. Depending on your RAID level selection, the physical disks available for the selected RAID level is displayed in

Unselected Physical Disks

In the

Unselected Physical Disks

disks and click

NOTE: You can select multiple physical disks at the same time by holding

<Ctrl> or <Shift> and selecting additional physical disks.

Click

Calculate Capacity

Click

Finish

Add

. A message is displayed confirming that the disk group is

table.

table, select the appropriate physical

to view the capacity of the new disk group.

successfully created and that you must create at least one virtual disk before you can use the capacity of the new disk group.

To create virtual disks:

Choose one of these methods to start the

Create Virtual Disk Wizard

• To create a virtual disk from unconfigured capacity in the storage array—On the select

Virtual Disk Create

Logical

tab, select an unconfigured capacity node and

Alternatively, right-click the unconfigured capacity node and select

Create Virtual Disk

from the pop-up menu.

field. You

window is

• To create a virtual disk from free capacity on a disk group—On the

Logical Create

Alternatively, right-click the free capacity

Virtual Disk

tab, select a free capacity node and select

node and select

from the pop-up menu.

Preparing Your Systems for Clustering 53

Virtual Disk

Create

Page 54

• To create a virtual disk from unassigned physical disks in the storage array—On the disks of the same physical disk type, and select

Physical

tab, select one or more unassigned physical

Virtual Disk Create

Alternatively, right-click the unassigned physical disks, and select

Create Virtual Disk

• To create a secure virtual disk—On the

from the pop-up menu.

Physical

tab, select one or more unassigned security capable physical disks of the same physical disk type, and select

Virtual Disk Create

Alternatively, you can right-click the unassigned security capable physical disks and select

Create Virtual Disk

from the pop-up menu.

If you chose an unconfigured capacity node or unassigned physical disks to create a virtual disk:

On the

Create a disk group by using the The

Disk Group Required

window, click

Create Virtual Disk Wizard

Yes

Create Disk Group Wizard

is displayed after you create the

disk group.

If you chose a free capacity node, the

Create Virtual Disk

window is

displayed.

Click

. The

Specify Capacity/Name

Select the appropriate unit for memory from the enter the capacity of the virtual disk in the

window is displayed.

Units

drop-down list and

New Virtual Disk Capacity

field.

Enter a character name (up to 30 characters) for the virtual disk in the

Virtual Disk Name

In the

Advanced Virtual Disk Parameters

field.

field, you can select:

• Use recommended settings

• Customize settings

Click

In the

Customize Advanced Virtual Disk Parameters

window, select the appropriate Virtual Disk I/O Characteristics type from the following options:

• File system (typical)

• Database

54 Preparing Your Systems for Clustering

Page 55

• Multimedia

•Custom

NOTE: If you select Custom, you must select an appropriate segment size.

Select the appropriate Preferred RAID controller module.

For more information on creating disk groups and virtual disks, see the Dell PowerVault Modular Disk Storage Manager User’s Guide at support.dell.com/manuals.

It is recommended that you create at least one virtual disk for each application. If multiple NTFS volumes are created on a single virtual disk using Windows Disk Management, the volumes failover together, rather than individually from node-to-node.

NOTE: It is recommended that you use a RAID level other than RAID 0 (which is

commonly called striping). RAID 0 configurations provide very high performance, but do not provide the level of availability required for the quorum resource. See the documentation for your storage system for more information about setting up RAID levels for the system.

Creating Host-to-Virtual Disk Mappings

Create host-to-virtual disk mappings to assign virtual disks to the host groups containing cluster nodes as follows:

In the

Array Management

In the

To po l og y

pane, select:

window, select the

Mappings

tab.

•Default Group

• Undefined Mappings Node

• Individual Defined Mapping

• Host Group

•Host

In the tabular, select

Define Additional Mapping

Select the appropriate host group from the

Logical Unit Number

Mappings Define Additional Mapping

window is displayed.

Host Group

Host

field, select a LUN. The supported LUNs are 0

. The

field.

through 255.

Preparing Your Systems for Clustering 55

Page 56

Select the virtual disk to be mapped in the

Virtual Disk

section lists the names and capacity of the virtual disks that

Virtual Disk

section. The

are available for mapping based on the selected host group or selected host.

Click

Add

NOTE: The Add button is inactive until a host group or host, LUN, and virtual disk

are selected.

To define additional mappings, repeat step 4 through step 7.

NOTE: After a virtual disk has been mapped once, it is no longer available in the

Virtual Disk area.

Click

Mappings

. The mappings are saved. The

pane in the

Mappings

tab are updated to display the mappings.

To po l og y

pane and the

Defined

Troubleshooting Tools

The Dell PowerVault MDSM establishes communication with each managed array and determines the current array status. When a problem occurs on a storage array, the MDSM provides several ways to troubleshoot the problem.

Event Log

You can use the Event Log Viewer to view a detailed list of events that occur in a storage array. The event log is stored on reserved areas on the storage array disks. It records configuration events and storage array component failures.

CAUTION: Use this option only under the guidance of your Technical Support

representative.

The event log stores approximately 8000 events before it replaces an event with a new event. If you want to keep a record of the events, you may save them or clear them from the event log.

The Event Log window shows the following types of event views:

• Summary view—Shows an event summary in a table form.

• Detail view—Shows details about a selected event.

To view the event log:

In the

Array Management

View Event Log

. The

window, select

Event Log

Advanced Troubleshooting

is displayed. By default, the summary

view is displayed.

56 Preparing Your Systems for Clustering

Page 57

Select

View Details

to view the details of each selected log entry. A

Detail

pane is added to the event log that contains information about the log item. You can view the details of a single log entry at a time.

To save the event log:

Click

Save As

Navigate to the relevant folder and enter the relevant file name.

Click

Save

Click

Clear All

Click

. The

Save Events

dialog box is displayed.

to erase all log entries from the event log.

to exit the event log.

For more information, see the PowerVault Modular Disk Storage Manager online help topics.

Recovery Guru

The Recovery Guru is a component of MDSM that diagnoses critical events on the storage array and recommends step-by-step recovery procedures to resolve problems.

To di s play t he Recovery Guru window in the

Array Management

window

perform one of the following actions:

•Click

•On the

• From the

Recovery Guru

Support

Attention

Status

tab, click

Recover from Failure

pane on the

Summary

tab, click

Storage Array Needs

You can detect a problem using the following indicators:

• Non-Optimal status icons

• Alert notification messages that are sent to the appropriate destinations

• Hardware indicator lights

The status icons return to Optimal status when problems are resolved.

Storage Profile

The storage array profile provides a description of all components and properties of the storage array. The storage array profile also provides the option to save the storage array profile information in a text file. You can also

Preparing Your Systems for Clustering 57

Page 58

use the storage array profile as an aid during recovery or as an overview of the current configuration of the storage array. Create a new copy of the storage array profile if your configuration changes.

To open the storage array profile in the

Array Management

window,

perform one of the following actions:

• Select

• Select the

Storage Array View Profile

Summary

tab and click

Storage Array Profile

in the

Status

area.

• Select the

The

Storage Array Profile

Support

tab and click

View Storage Array Profile

screen is displayed. The

Storage Array Profile

screen contains several tabs, and the title of each tab corresponds to the subject of the information contained.

Perform one of these actions in the

Storage Array Profile

screen:

• View detailed information (Go to step 3).

• Search the storage array profile (Go to step 4).

• Save the storage array profile (Go to step 5).

• Close the storage array profile (Go to step 6).

Select one of the tabs, and use the horizontal scroll bar and the vertical scroll bar to view the storage array profile information. You can use other steps in this procedure to search the storage array profile, to save the storage array profile, or to close the storage array profile.

To search the storage array profile:

Click

Find

In the

Find

text box, type the term that you want to search. If the term is located on the current tab, it is highlighted in the storage array profile information.

NOTE: The search is limited to the current tab. If you want to search for the

term in other tabs, select the tab and click Find again.

Click

Find

To save the storage array profile:

Click

To save all sections of the storage array profile, select

again to search for additional occurrences of the term.

Save As

58 Preparing Your Systems for Clustering

All Sections

Page 59

To save information from particular sections of the storage array profile, select the

Select Sections

option and click on the check boxes

corresponding to the sections that you want to save.

Select an appropriate directory.

File Name

field, type a file name of your choice. To associate the file with a particular software application that opens it, specify a file extension, for example

NOTE: The file is saved as ASCII text.

Click

Save

To exit the storage array profile, click

Status Icons

.text

Status icons identify the six possible health status conditions of the storage array. For every non-optimal status icon, use the Recovery Guru to detect and troubleshoot the problem. The six possible health status conditions are described below:

• Optimal—Every component in the managed array is in the desired working condition.

• Needs Attention—A problem exists with the managed array that requires intervention to correct it.

• Fixing—The managed array is currently changing to an

Needs Attention

condition has been corrected and the

Optimal

status.

• Unresponsive—The storage management station cannot communicate with the array, one controller, or both controllers in the storage array. Wait at least five minutes for the storage array to return to an

Optimal

status

following a recovery procedure.

• Unsupported—The node is not supported by this version of MDSM.

• Software Unsupported—The storage array is running a level of software that is no longer supported by MDSM.

Preparing Your Systems for Clustering 59

Page 60

Configuring the RAID Level for the Shared Storage Subsystem

The virtual disks in your shared storage subsystem must be configured into disk groups or virtual disks using the Dell PowerVault MDSM software. All virtual disks, especially if they are used for the quorum resource, must be bound and must incorporate the appropriate RAID level to ensure high availability.

NOTE: It is recommended that you use a RAID level other than RAID 0 (which is

called striping). RAID 0 configurations provide very high performance, but do not provide the level of availability required for the quorum resource. See the documentation for your storage system for more information about setting up RAID levels for your system.

Windows Operating System and Dynamic Volumes

The Windows operating system does not support dynamic disks (upgraded disks) or volumes as shared cluster storage. If the shared cluster storage is configured as a dynamic disk, the Cluster Configuration wizard is not able to discover the disks, preventing the cluster and network clients from accessing the disks.

Assigning Drive Letters and Mount Points

A mount point is a drive attached to an empty folder on an NTFS volume. A mount point functions the same as a normal drive but is assigned a label or name instead of a drive letter. Using mount points, a cluster can support more shared disks than the number of available drive letters.

The cluster installation procedure does not automatically add the mount point into the disks managed by the cluster. To add the mount point to the cluster, create a physical disk resource in the cluster resource group for each mount point. Ensure that the new physical disk resource is in the same cluster resource group and is dependent on the root disk (that is, the disk from which the mount point is attached).

NOTE: When mounting a drive to an NTFS volume, do not create mount points from

the quorum resource or between the clustered disks and the local disks. Mount points must be in the same cluster resource group and must be dependent on the root disk.

60 Preparing Your Systems for Clustering

Page 61

Naming and Formatting Drives on the Shared Storage System

Each virtual disk being created in the PowerVault Modular Disk Storage Manager becomes a physical disk in Windows Disk Management. For each physical disk, perform the following:

• Write the disk signature

• Create the partition

• Assign the drive letter

• Format the partition with NTFS

CAUTION: The drive letters are manually assigned from the second node, the

shared disks are simultaneously accessible from both nodes. To ensure file system integrity and prevent possible data loss before you install the Microsoft Failover Clustering software, prevent any I/O activity to the shared drives by performing the following procedure on one node at a time and ensuring that the other node is shut down.

The number of drive letters required by individual servers in a cluster may vary. It is recommended that the shared drives be named in reverse alphabetical order beginning with the letter z. To assign drive letters and format drives on the shared storage system, perform the following steps:

Turn off node 2 and open

Allow Windows to enter a signature on all new physical or logical drives.

NOTE: Do not upgrade or convert your disks to dynamic disks.

Locate the icon for the first unnamed, unformatted drive on the shared storage system.

Right-click the icon and select unformatted drives are not visible, verify the following:

• The zoning is correct (for

• The LUNs have been assigned to the hosts.

• The storage system is properly cabled to the servers.

In the dialog box, create a partition with the size of the entire drive (the default) and then click

Disk Management

Create

from the submenu. If the

SAN-attached configuration

on node 1.

s only)

NOTE: A virtual disk that is mapped or assigned from the storage system to a

cluster node(s) is represented as a physical disk within the Windows operating system on each node. Microsoft Cluster allows only one node to

Preparing Your Systems for Clustering 61

Page 62

access a given physical disk resource at a time. Therefore, if a disk is partitioned and contains multiple NTFS volumes, concurrent access to different volumes is only possible from the cluster node controlling the physical disk resource. If two NTFS volumes need to be controlled by different nodes, these volumes must reside on separate disks.

Click

Yes

to confirm the partition.

With the mouse pointer on the same icon, right-click and select

Drive Letter and Path

Assign a drive letter to an NTFS volume or create a mount point.

from the submenu.

To assign a drive letter to an NTFS volume:

Click

Edit

and select the letter you want to assign to the drive (for

example, z).

Click OK.

Go to step 9.

To create a mount point:

Click

Add

Click

Mount

in the following empty NTFS folder.

Type the path to an empty folder on an NTFS volume, or click to locate it.

Click OK.

Go to step 9.

Click

Yes

to confirm the changes.

Right-click the drive icon again and select

Under

Volume La b e l

, enter a descriptive name for the new volume; for

Format

from the submenu.

example, Disk_Z or Email_Data.

In the dialog box, change the file system to and click

Start

NTFS

, select

Change

Browse

Quick Format

NOTE: The NTFS file system format is required for shared-disk resources

under Microsoft Cluster.

Click OK at the warning.

Click

to acknowledge that the format is complete.

Click

to close the dialog box.

62 Preparing Your Systems for Clustering

Page 63

Repeat step 3 through step 15 for each remaining drive.

Disk Management

Turn off node 1.

Turn on node 2.

On node 2, open

Ensure that the drive letters for node 2 are correct and re-assign the drive letters, if necessary. To re-assign the drive letters, repeat step 7 through step 9.

Disk Management

Using Advanced (Premium) PowerVault Modular Disk Storage Manager Features

PowerVault Modular Disk Storage Manager includes the following advanced features:

• Snapshot Virtual Disk

•Virtual Disk Copy

• The High Performance Tier

• Remote Replication

To install and enable these premium features, you must purchase a feature key file for each feature and specify the storage array that must host them. For more information, see the Premium Feature Activation card that shipped along with your Dell PowerVault MD3600f or MD3620f storage system.

These premium features increase the high availability for your cluster solution. It is essential that you follow the instructions below to ensure proper cluster operations.

Snapshot Virtual Disk

Snapshot Virtual Disk captures point-in-time images of a virtual disk for backup, testing, or data processing without affecting the contents of the source virtual disk. You can use either Simple Path or Advanced Path to create a snapshot for your cluster disk. The Snapshot Virtual Disk can be mapped to the primary node (the node owning the source disk) or the secondary node (the node not owning the source disk) for backup, testing, or data processing.

Preparing Your Systems for Clustering 63

Page 64

CAUTION: Avoid mapping the Snapshot Virtual Disk to more than one node in the

cluster at any point of time. The Snapshot Virtual Disk is not managed by Failover Cluster Manager, so mapping the Snapshot Virtual Disk to the host group or both nodes in the cluster may allow both nodes to access data concurrently and thus cause data corruption.

You can use a Microsoft Volume Shadow-copy Service (VSS) application to create and map snapshots. If you are using MDSM instead, you must follow the procedures described below.

To map the Snapshot Virtual Disk to the primary node:

Use Host-to-Virtual Disk Mapping in the Modular Disk Storage Manager. This ensures that a different disk signature is assigned properly to the Snapshot Virtual Disk.

Use Windows Disk Management to re-scan for the Snapshot Virtual Disk, assign the drive letter, and start accessing the drive.

NOTE: The disks may be re-scanned several times for the Snapshot Virtual

Disk to be detected by Windows Disk Management. If the Snapshot Virtual Disk is not detected, wait for a few minutes and re-scan the disks. Repeat the process until the Snapshot Virtual Disk is detected; do not reboot the server.

To map the Snapshot Virtual Disk to the secondary node (the node not owning the source disk):

CAUTION: Attempts to map the Snapshot Virtual Disk to the secondary node, prior

to obtaining the signature from the primary node, may cause the operating system to misidentify the Snapshot Virtual Disk as an existing system volume and that may result in data loss or an inaccessible Snapshot Virtual Disk.

Map the Snapshot Virtual Disk to the primary node to ensure that the snapshot is assigned a new disk signature.

Use Modular Disk Storage Manager to unmap the Snapshot Virtual Disk from the primary node.

Map the Snapshot Virtual Disk to the secondary node and start accessing it.

NOTE: For a cluster configuration with multiple Snapshot Virtual Disks, each virtual

disk must be mapped to the node owning the associated source disk first. The primary node for a Snapshot Virtual Disk may not be the primary node for another Snapshot Virtual Disk.

64 Preparing Your Systems for Clustering

Page 65

Virtual Disk Copy

Virtual Disk Copy generates a full copy of data from the source virtual disk to the target virtual disk in a storage array. You can use Virtual Disk Copy to back up data, copy data from disk groups that use smaller-capacity physical disks to disk groups using greater-capacity physical disks, or restore Snapshot Virtual Disk data to the source virtual disk.

To create a Virtual Disk Copy of a Microsoft Cluster shared disk:

Create a Snapshot Virtual Disk using the cluster shared disk as a source disk.

Do not map that Snapshot Virtual Disk to any cluster node. Then, use the newly created Snapshot Virtual Disk as the source disk for the Virtual Disk Copy.

NOTE: When you attempt to create a Virtual Disk Copy of a Microsoft Cluster

shared disk directly, the operation fails and the following error is displayed:

The operation cannot complete because the selected virtual disk is not a source virtual disk candidate

If the cluster shared disk fails and you want to restore it from the target virtual disk, use Failover Cluster Manager to change the status of the cluster group containing the failed disk to offline. Use one of the following methods:

• Use Virtual Disk Copy to transfer the data from the target virtual disk to the cluster shared disk.

• Unassign the cluster shared disk from the host group and map the target virtual disk to the host group.

Remote Replication

Remote Replication is a storage array premium feature that provides online, real-time replication of data between storage arrays. When this feature is activated, a dedicated port (port 3) on each RAID controller is reserved exclusively for Remote Replication traffic. No other data traffic is allowed on that port until the Remote Replication feature is deactivated. A dedicated zone on each Fibre Channel switch is also required. For cabling examples, configurations, host bus adapter (HBA) installation and zoning, see Configuring Fibre Channel with the Dell MD3600f Series Storage Array at support.dell.com/manuals.

Preparing Your Systems for Clustering 65

Page 66

Configuring a Failover Cluster

You can configure the operating system services on your Windows Server failover cluster after you have established the private and public networks and assigned the shared disks from the storage array to the cluster nodes. The procedures for configuring the failover cluster are different depending on the Windows Server operating system you use.

For more i n f ormation on deploying your cluster, see the Dell Failover Clusters with Microsoft Windows Server 2008 Installation and Troubleshooting Guide at support.dell.com/manuals.

66 Preparing Your Systems for Clustering

Page 67

Troubleshooting

Table A-1 describes the general cluster problems, the probable causes, and solutions for each problem.

Table A-1. General Cluster Troubleshooting

Problem Probable Cause Corrective Action

The nodes cannot access the storage system, or the cluster software is not functioning with the storage system.

The storage system is not cabled properly to the nodes or the cabling between the storage components is incorrect.

One of the cables is faulty. Replace the faulty cable.

Host Group or Host-toVirtual Disk Mappings is not created correctly.

Ensure that the cables are connected properly from the node to the storage system. For more information, see "Cabling Your Cluster Hardware" on page 19.

Verify the following:

• Host Group is created and the cluster nodes are added to the Host Group.

• Host-to-Virtual Disk Mapping is created and the virtual disks are assigned to the Host Group containing the cluster nodes.

Troubleshooting 67

Page 68

Table A-1. General Cluster Troubleshooting

Problem Probable Cause Corrective Action

One of the nodes takes a long time to join the cluster.

One of the nodes fails to join the cluster.

The node-to-node network has failed due to a cabling or hardware failure.

Long delays in node-tonode communications may be normal.

One or more nodes may have the Internet Connection Firewall enabled, blocking Remote Procedure Call (RPC) communications between the nodes.

(continued)

Check the network cabling. Ensure that the node-to-node interconnection and the public network are connected to the correct NICs.

Verify that the nodes can communicate with each other by running the ping command from each node to the other node. Try both the host name and IP address when using the ping command.

Configure the Internet Connection Firewall to allow communications that are required by the Microsoft Failover Clustering and the clustered applications or services. For more information, see the Microsoft Knowledge Base article KB883398 at support.microsoft.com.

68 Troubleshooting

Page 69

Table A-1. General Cluster Troubleshooting

Problem Probable Cause Corrective Action

Attempts to connect to a cluster using Failover Cluster Manager fail.

• The Cluster Service has not been started.

• A cluster has not

(continued)

Verify that Cluster Service is running and that a cluster has been formed.

been formed on the system.

• The system has just booted and services are still starting.

You are prompted to configure one network instead of two during Microsoft Failover Cluster installation.

The cluster network name is not responding on the network because the Internet Connection Firewall is enabled on one or more nodes

The TCP/IP configuration is incorrect.

The private (point-topoint) network is disconnected.

Configure the Internet Connection Firewall to allow communications that are required by Microsoft Cluster and the clustered applications or services. For more information, see the Microsoft Knowledge Base article KB883398 at support.microsoft.com

The node-to-node network and public network must be assigned static IP addresses on different subnets. For more information about assigning the network IPs, see "Assigning Static IP Addresses to Your Cluster Resources and Components" in the Dell Failover Clusters with

Microsoft Windows Server 2008 Installation and Troubleshooting Guide at

support.dell.com/manuals.

Ensure that all systems are powered on so that the NICs in the private network are available.

Troubleshooting 69

Page 70

Table A-1. General Cluster Troubleshooting

(continued)

Problem Probable Cause Corrective Action

Unable to add a node to the cluster.

The new node cannot access the shared disks.

Ensure that the new cluster node can enumerate the cluster disks using Windows Disk Administration. If the disks do not appear in Disk Administration, check the following:

• All cable connections.

• The Access Control settings on the attached storage systems.

One or more nodes may have the Internet Connection Firewall enabled, blocking RPC communications between the nodes.

Configure the Internet Connection Firewall to allow communications that are required by the Microsoft Cluster and the clustered applications or services. For more information, see the Microsoft Knowledge Base article KB883398 at support.microsoft.com.

Public network clients cannot access the applications or services that are provided by the cluster.

One or more nodes may have the Internet Connection Firewall enabled, blocking RPC communications between the nodes.

Configure the Internet Connection Firewall to allow communications that are required by the Microsoft Cluster and the clustered applications or services.

For more information, see the Microsoft Knowledge Base article KB883398 at support.microsoft.com.

Virtual Disk Copy operation fails.

The Virtual Disk Copy operation uses the cluster disk as the source disk.

Perform a Virtual Disk Copy operation on the cluster share disk:

Create a snapshot of the disk.

Perform a Virtual Disk Copy of the snapshot virtual disk.

70 Troubleshooting

Page 71

Table A-1. General Cluster Troubleshooting

(continued)

Problem Probable Cause Corrective Action

Unable to assign the drive letter to the snapshot virtual disk.

Unable to access the snapshot virtual disk.

System Error Log displays a warning with event 59 from partmgr stating that the snapshot virtual disk is

The snapshot virtual disk has been erroneously mapped to the node that does not own the source disk.

Unmap the snapshot virtual disk from the node not owning the source disk.

Assign it to the node that owns the source disk. For more information, see "Using Advanced (Premium) PowerVault Modular Disk Storage Manager Features" on page 63.

a redundant path of a cluster disk.

Troubleshooting 71

Page 72

72 Troubleshooting

Page 73

Cluster Data Form

You can attach the following form in a convenient location near each cluster node or rack to record information about the cluster. Use the form when you call for technical support.

Table B-1. Cluster Configuration Information

Cluster Information Cluster Solution

Cluster name and IP address

Server type

Installer

Date installed

Applications

Location

Notes

Cluster Data Form 73

Page 74

Table B-2. Cluster Node Configuration Information

Node Name Service Tag

Number

Public IP Address Private IP Address

Table B-3. Additional Network Information

Additional Networks

74 Cluster Data Form

Page 75

Table B-4. Storage Array Configuration Information

Array Array Service Tag IP Address Number of Attached Disk

Array Enclosures (DAEs)

Virtual Disks

Cluster Data Form 75

Page 76

76 Cluster Data Form

Page 77

Zoning Configuration Form

Node HBA WWPNs

or Alias Names

Storage WWPNs or Alias Names

Zone Name Zone Set for

Configuration Name

DELL CONFIDENTIAL – PRELIMINARY 7/25/11 - FOR PROOF ONLY

Template Last Updated - 2/7/2007 Zoning Configuration Form 77

Page 78

FILE LOCATION:

C:\Users\philip_john\Documents\HA_Clusters_Compellent\FalconHead\Sources\appC.fm

DELL CONFIDENTIAL – PRELIMINARY 7/25/11 - FOR PROOF ONLY

78 Zoning Configuration Form

Page 79

Index

advanced features

MD3600f and MD3620f storage

systems, 12

virtual disk copy, 12

assigning

drive letters and mount points, 60

cabling

mouse, keyboard, and monitor, 19 power supplies, 19

cluster configurations

direct-attached, 12, 24 SAN-attached, 13

cluster data form, 73

cluster storage requirements, 10

configuring

failover cluster, 66 shared storage system, 48

Dell PowerVault MD36x0f

storage system

Cabling a Two-Node Cluster, 26

about, 24

event log, 56

host bus adapter

configuring the Fibre Channel

HBA, 44

installing and configuring

storage management software, 47

MDSM, 10

Modular Disk Storage

Manager, 10

Modular Disk Storage Manager

Agent, 11

multipath software, MPIO

software, 11

direct-attached cluster

Index 79

Page 80

NIC teaming, 24

supported cluster

configurations, 12

operating system

installing, 43

PowerVault MD36x0f

Cabling a Multi-Node Cluster, 27

PowerVault MD36x0f storage

systems Cabling Multiple Clusters, 29 Cabling Multiple SAN-Attached

Clusters, 36 Cabling Storage, 30 Zoning, 36

recovery guru, 57

SAN-Attached Cluster, 13

SAN-attached cluster

about, 30

snapshot virtual disk, 63

tape library

connecting to a PowerEdge

cluster, 36

troubleshooting

general cluster, 67

virtual disk copy, 65

Windows Server 2008,

Enterprise Edition

installing, 43

zones

implementing on a Fibre Channel

switched fabric, 45

zoning configuration form, 77

status icons, 59

storage profile, 57

80 Index

Dell MD3600f, MD3620f Owner's Manual

Contents

Introduction

Overview

Cluster Solution

Cluster Requirements

Cluster Nodes

Cluster Storage

Cluster Storage Management Software

Dell PowerVault Modular Disk Storage Manager

Modular Disk Storage Manager Agent

Multipath I/O (MPIO) Software

Advanced Features

Supported Cluster Configurations

Direct-Attached Cluster

SAN-Attached Cluster

Configuration Order for Direct-Attached and SAN-Attached Connections

Direct-Attached Configuration Order

SAN-Attached Configuration Order

Other Documents You May Need

Cabling Your Cluster Hardware

Cabling the Mouse, Keyboard, and Monitor

Cabling the Power Supplies

Cabling Your Public and Private Networks

Cabling Your Public Network

Cabling Your Private Network

Using Dual-Port Network Adapters for Your Private Network

NIC Teaming

Cabling the Storage Systems

Cabling Storage for Your Direct-Attached Cluster

Cabling a Cluster to a Dell PowerVault MD36x0f Storage System

Cabling a Two-Node Direct-Attached Cluster to an MD36x0f Storage System

Cabling a Four-Node Direct-Attached Cluster to an MD36x0f Storage System

Cabling Multiple Clusters to an MD36x0f Storage System

Cabling Two Two-Node Direct-Attached Clusters to an MD36x0f Storage System

Cabling Storage for Your SAN-Attached Cluster

Cabling a SAN-Attached Cluster to an MD36x0f Storage System

Cabling a SAN-Attached Cluster to the MD36x0f Storage System

Cabling Multiple SAN-Attached Clusters to an MD36x0f Storage System

Connecting a PowerEdge Cluster to a Tape Library

Connecting a PowerEdge Cluster to Multiple PowerVault MD3600f or MD3620f Storage Systems

Preparing Your Systems for Clustering

Cluster Configuration Overview

Installation Overview

Installing Supported Fibre Channel HBAs

Installing Fibre Channel HBA on Your Host Server

Customizing HBA Drivers and Firmware Settings

Implementing Zoning on a Fibre Channel Switched Fabric

World Wide Name Zoning

Switch Zoning Requirements

Setting Up Zoning on the Fibre Channel Switch Hardware

Installing the Storage Management Software

Configuring the Shared Storage System

Automatic Storage Array Discovery

Defining a Host

Creating a Host Group

Creating Disk Groups and Virtual Disks

Creating Host-to-Virtual Disk Mappings

Troubleshooting Tools

Event Log

Recovery Guru

Storage Profile

Status Icons

Configuring the RAID Level for the Shared Storage Subsystem

Windows Operating System and Dynamic Volumes

Assigning Drive Letters and Mount Points

Naming and Formatting Drives on the Shared Storage System

Using Advanced (Premium) PowerVault Modular Disk Storage Manager Features

Snapshot Virtual Disk

Virtual Disk Copy

Remote Replication

Configuring a Failover Cluster

Troubleshooting

Cluster Data Form

Zoning Configuration Form

Index