Requirements for Windows HPC Server 2008 R2 in Failover Clusters

  • Article

Applies To: Windows HPC Server 2008 R2

This topic provides information about the requirements for deploying Windows® HPC Server 2008 R2 in one or more failover clusters where the servers are running Windows Server 2008 R2. This topic discusses requirements for the following configurations:

The system requirements for the configurations in this guide are a combination of the requirements for the following components, which are required for running Windows HPC Server 2008 R2 in the context of a failover cluster:

  • The failover clustering feature in Windows Server 2008 R2.

  • An installation of SQL Server 2008 Service Pack 1. The SQL Server can be a standalone SQL Server or a SQL Server failover cluster.

  • Windows HPC Server 2008 R2.

The following sections provide more detail about these requirements:

  • Hardware requirements for a failover cluster running Windows HPC Server 2008 R2

  • Software requirements for a failover cluster running Windows HPC Server 2008 R2

  • Domain account and SQL Server account requirements

  • Network infrastructure and SQL Server network access requirements

Hardware requirements for a failover cluster running Windows HPC Server 2008 R2

You need the following hardware for a failover cluster running Windows HPC Server 2008 R2:

  • Servers that are compatible with Windows Server 2008 R2: You need two servers for the failover cluster that runs the head node, and if the SQL Server that supports the head node is remote, one or more servers for the SQL Server installation. The failover cluster that runs the head node must contain only two servers.

    If your configuration also includes WCF broker nodes in failover clusters, each of those failover clusters must have at least two servers, and they can have as many as 16 servers. An HPC cluster can have as many as eight failover clusters supporting WCF broker nodes.

    We recommend that for each failover cluster, you use a set of matching computers that contain the same or similar components.

    The hardware requirements for the servers in the failover clusters are as follows:

    • Processor: The minimum is 1.4 GHz (x64 processor).

    • Memory: The minimum is 512 MB and the maximum is 2 TB.

    • Disk space: The minimum is 32 GB. For systems where the memory is larger than 16 GB, more disk space is required for paging, hibernation, and dump files.

    Important

    For Windows Server 2008 R2, Microsoft supports a failover cluster solution only if all the hardware components are marked as "Certified for Windows Server 2008 R2." In addition, the complete configuration (servers, network, and storage) must pass all tests in the Validate a Configuration Wizard, which is included in the Failover Cluster Manager snap-in.

    For additional information about server hardware requirements, see the following documents:

  • Network adapters and cables (for failover network communication): The network hardware, like other components in the Windows Server 2008 R2 failover cluster solution, must be compatible with Windows Server 2008 R2. If you use iSCSI, each of your network adapters must be dedicated to either network communication or iSCSI, not both.

    Important

    In the network infrastructure that connects the failover cluster servers, to avoid having single points of failure for the configuration that is described in this guide, you must use multiple, distinct networks. This is important because the load on the network that connects a head node to compute nodes can be high when the compute nodes are being deployed or when the application load is high.

    For Windows HPC Server 2008 R2, a public-only topology, also called Topology 5 (Public Only) is not recommended for the scenario described in this guide, because a single network is a single point of failure. For more information, see Network infrastructure and SQL Server network access requirements, later in this section.

    Note that if you connect the failover cluster servers by using a single network, the network will pass the redundancy requirement in the Validate a Configuration Wizard. However, the report from the Failover Cluster Setup Wizard will include a warning that the network should not have single points of failure.

  • Device controllers or appropriate adapters for the storage:

    • For Serial Attached SCSI or Fibre Channel: If you are using Serial Attached SCSI or Fibre Channel, all components of the storage stack should be identical in all servers in a failover cluster. It is required that the multipath I/O (MPIO) software and Device Specific Module (DSM) software components be identical. It is recommended that the mass-storage device controllers—that is, the host bus adapter (HBA), HBA drivers, and HBA firmware—that are attached to the cluster storage be identical. If you use dissimilar HBAs, you should verify with the storage vendor that you are following their supported or recommended configurations.

      Note

      With Windows Server 2008 R2, you cannot use parallel SCSI to connect the storage to the servers in the failover cluster. This was also true for Windows Server 2008.

    • For iSCSI: If you are using iSCSI, each server in the failover cluster must have one or more network adapters or host bus adapters that are dedicated to the cluster storage. The network that you use for iSCSI cannot be used for general network communication. The network adapters that you use to connect to the iSCSI storage target should be identical, and we recommend that you use Gigabit Ethernet or a faster connection.

      For iSCSI, you cannot use teamed network adapters, because they are not supported with iSCSI.

      For more information about iSCSI, see iSCSI Cluster Support: FAQ (https://go.microsoft.com/fwlink/?LinkId=61375).

  • Storage: You must use shared storage that is compatible with Windows Server 2008 R2. Storage requirements include the following:

    • For a failover cluster that runs the head node, at least two disks are needed.

      If you have a failover cluster that runs one or more WCF broker nodes, you must allow one disk per active cluster node (to support the instance of Message Queuing running on that active node). Also, if that failover cluster has an even number of nodes, an additional disk is needed for a disk witness. For details about how these disks are used, see Overview of Windows HPC Server 2008 R2 and SOA in Failover Clusters (https://go.microsoft.com/fwlink/?LinkId=198304).

      The disk witness is a disk in the failover cluster storage that is designated to hold a copy of the failover cluster configuration database. A failover cluster has a disk witness only if this is specified as part of the quorum configuration. For a failover cluster with an even number of nodes, a disk witness is part of the most common quorum configuration, Node and Disk Majority. Node and Disk Majority means that the servers in the failover cluster and the disk witness each contain copies of the failover cluster configuration, and a failover cluster can function as long as a majority of these copies are available.

    • If the storage that you are using has a proprietary installation process, use the instructions that are provided by your vendor.

    • To use the native disk support that is included in failover clustering, use basic disks, not dynamic disks.

    • We recommend that you format the partitions with the NTFS file system (for a disk witness, the partition must be NTFS).

    • For the partition style of the disk, you can use a master boot record (MBR) or a GUID partition table (GPT).

    • The disk where SQL Server will be installed must be uncompressed.

Software requirements for a failover cluster running Windows HPC Server 2008 R2

You need to use the following software for a failover cluster running Windows HPC Server 2008 R2:

  • To set up the head node (or WCF broker nodes, if you have them) in a failover cluster, you must use the Windows Server 2008 R2 Enterprise or Windows Server 2008 R2 Datacenter operating systems. In addition, Windows HPC Server 2008 R2 only supports x64, so all the servers must run the same x64 hardware version of the operating system. All servers should also have the same security updates and service packs.

    Important

    You cannot use Windows Server 2008 R2 HPC Edition as the operating system for a server that will be in a failover cluster.

  • If you plan to use service-oriented architecture (SOA), you must apply a hotfix rollup for Windows Communication Foundation (WCF) to each server that you want to include in the failover cluster that will run head node services.WCF is part of the .NET Framework 3.5 SP1 in Windows Server 2008 R2. For more information, see:

    A hotfix rollup is available for Windows Communication Foundation in the .NET Framework 3.5 SP1 for Windows 7 and Windows Server 2008 R2 (https://go.microsoft.com/fwlink/?LinkId=212793)

  • The SQL Server installation that supports the head node requires SQL Server 2008 Service Pack 1 (SP1). You can use any edition of SQL Server 2008 SP1 except SQL Server Compact 3.5, although we do not recommend using SQL Server 2008 Express for HPC clusters that have more than 256 nodes. If you decide to install SQL Server 2008 Express SP1 and you are deploying a large cluster, be aware that the database size is limited to 4 GB per database.

  • For best results, the servers that you configure as a head node in a failover cluster should not also be domain controllers. In other words, those servers cannot also have Active Directory Domain Services (a server role in Windows Server 2008 R2) installed.

  • In the HPC cluster, the head node cannot serve as a compute node or WCF broker node, and a WCF broker node cannot serve as a head node or compute node.

Domain account and SQL Server account requirements

You need the following domain attributes:

  • Domain role: No server running SQL Server 2008 SP1 should be a domain controller (to maintain security levels for SQL Server). If you create a SQL Server failover cluster, all servers in the failover cluster must be in the same Active Directory domain.

  • Account or accounts for configuring failover clusters: When you first create a failover cluster, you must be logged on to the domain with an account that has administrator rights and permissions on the servers in the failover cluster. The account does not need to be a Domain Admins account; it can be a Domain Users account that is in the Administrators group on the servers. In addition, if the account is not a Domain Admins account, the account (or the group that the account is a member of) must have the Create Computer Objects and Read All Properties permissions in the domain.

  • Accounts in Active Directory for the HPC cluster: In Active Directory Domain Services (AD DS), make sure that you have the following accounts for the HPC cluster:

    • Designated user account for adding nodes to the HPC cluster: We recommend that you create a specific user account that is designated as “installation credentials” for adding nodes (such as compute nodes) to the HPC cluster. For information about the permissions for this account, see the Windows HPC Server 2008 R2 Design and Deployment Guide, especially Step 3: Configure the Head Node (https://go.microsoft.com/fwlink/?LinkId=198319). Within the HPC cluster, this account will become a permanent member in the HPC cluster User role, which is why we recommend creating a designated account for this purpose.

    • Group accounts for HPC administrators and HPC users: Create two designated domain security groups, one for HPC users and one for HPC administrators. In HPC Cluster Manager, you should only add these domain security group accounts to the corresponding HPC cluster roles—User and Administrator. (The designated user account for “installation credentials” described in the previous item in this list is the only exception—when you specify that account, it is added to the HPC cluster User role.) For more information, see Step 6: Install HPC Pack 2008 R2 on a Server that Will Run Head Node Services.

  • Service account for SQL Server Services: You must assign a domain user account to each SQL Server service. You can use one domain account for all of the services. For more information, see Service Account (https://go.microsoft.com/fwlink/?LinkId=121044).

If you are using remote SQL Server, you also need the following attributes for your SQL Server instance:

  • Login on the SQL Server instance: On the SQL Server instance, you need to create a login for the domain security group account (in AD DS) that is used for HPC administrators. (This group account should contain the account of the person who will install the HPC cluster.)

  • Sysadmin role in SQL Server: You must be able to assign all SQL Server logins that are created for the HPC cluster to the sysadmin role in SQL Server.

Network infrastructure and SQL Server network access requirements

Windows HPC Server 2008 R2 supports multiple network topologies, which are designed to meet a wide range of user needs and performance, scaling, and access requirements. The topologies are distinguished by how many networks the cluster is connected to, and in what manner. In HPC clusters where you want to configure high availability, we recommend Topology 2 or Topology 4, depending on your application requirements. For best results, do not use Topology 5, which is a public-only topology, for such clusters.

Topology Description

1

Compute nodes isolated on private network

2

Compute nodes on public and private networks

3

Compute nodes isolated on private and MPI networks

4

Compute nodes on public, private, and MPI networks

5 (not recommended in a failover cluster)

Compute nodes on public network only

Network infrastructure requirements for failover clusters

Failover clusters have the following requirements for network infrastructure (in addition to the requirements listed in Hardware requirements for a failover cluster running Windows HPC Server 2008 R2, earlier in this topic):

  • Domain: All of the servers in the failover cluster or clusters must be in the same domain.

  • Network settings: In failover clusters, when you use identical network adapters for a network, also use identical communication settings on those adapters (for example, Speed, Duplex Mode, Flow Control, and Media Type). Also, compare settings between the network adapter and the switch it connects to and make sure that no settings are in conflict.

  • DNS: The servers in failover clusters require Domain Name System (DNS) for name resolution. The DNS dynamic update protocol can be used.

SQL Server network access requirements (for remote SQL Server only)

If you set up a remote SQL Server instance for your HPC cluster, you need the following network access for that SQL Server instance:

Additional references

Overview of Windows HPC Server 2008 R2 and SOA in Failover Clusters (https://go.microsoft.com/fwlink/?LinkId=198304)

Overview of Configuring the Head Node for Failover with Windows HPC Server 2008 R2 (https://go.microsoft.com/fwlink/?LinkId=198290)

Step 1: Install Windows Server 2008 R2 on Multiple Servers