HP Serviceguard SolutionsFamily data sheet

High availability and disaster recovery for the Converged Infrastructure Do you know how much unplanned downtime your organization experiences every month? If not, what would you estimate your monthly downtime to be? If you have a number in mind, chances are that your estimate is considerably lower than the unplanned downtime reported by organizations similar to yours that actually measure it.

And the problem could be getting worsesince 2005, average unplanned downtime per month in large organizations with more than 2,500 users has increased 69 percent, from 8.7 to 14.7 hours per month.* Those same organizations reported they would lose on average $128,536 per hour if their most mission-critical business systems were severely impacted by an outage.

The increase in average hours of unplanned downtime each month is sobering. As IT becomes more integral to the outcomes of the business, more applications are being classified as mission-critical and complexity

is risingso rapidly that you may not even fully track that complexity or completely digest the implications.

Assessing and addressing this problem successfully requires viewing downtime and complexity as two intertwined issues. HP Serviceguard Solutions enable robust high availability (HA) and disaster recovery (DR) for HP-UX 11i.

HP Serviceguard Solutions can simplify your daily operations to drive down incidents resulting from unnecessary complexity while reducing the impacts of maintenance and upgrades on your operations. The costs of downtime and the impacts to your organization can be radically reduced by employing Serviceguard Solutions to maintain the health of your infrastructure transparently to enable nearly continuous operations. On top of this, the portfolio includes tools to enhance the manageability of the entire environment. This data sheet will give you a solid understanding of the features and benefits of the various products within the Serviceguard Solutions family.

* Gartner Dataquest Insight: Unplanned Downtime Rising for Mission-Critical Applications, October 2008

Table of contents

Serviceguard ..........................................................3Serviceguard Manager ............................................6Serviceguard Toolkits ...............................................8Serviceguard NFS Toolkit ........................................8Serviceguard Enterprise Cluster Master Toolkit............8New! Serviceguard Toolkit for Oracle Data Guard .....8Serviceguard Developers Toolbox ............................8

Serviceguard Extensions .........................................10New! Serviceguard Extension for Oracle E-Business Suite ..................................................10

Serviceguard Extension for SAP .............................10Serviceguard Extension for RAC ............................ 11

Serviceguard Disaster Recovery Solutions .................14Extended Distance Cluster .....................................14Metrocluster ........................................................15Continentalclusters ............................................... 17

Serviceguard Storage Management Suite .................19

Hub

Quorum service

Node A

Root

Node A disk

Node A mirror

Node B disk

Node B mirror

Mirror

Root

Mirror

Node B

3

ServiceguardHP Serviceguard high availability clustering software is the foundation of every Serviceguard business continuity solution.

Serviceguard packages a critical service such as an application or database together with its associated resources, and monitors the entire package for any faults related to hardware, software, OS, virtualization layer, virtual machine guests, network, and storage. When a failure is detected, Serviceguard gracefully shuts down the application, restarts it on an available server in the cluster, connects it to the appropriate storage device, and resumes normal operationsall in mere seconds.

Serviceguard also includes enhanced package dependencies, which help administrators craft and control the dependencies between applications in a standardized way. This reduces complexity and provides a more realistic model of the environment, since real environments need to manage multiple tiers of applications together in a controlled and well-defined way.

Serviceguard cluster configurations deliver high availability using redundant hardware to remove single points of failure. This architecturetypically implemented at one site in a single data centeris sometimes called a local cluster (see Figure 1).

CapabilitiesComprehensive protection against hardware,

software, virtualization, network, and storage faults Advanced cluster arbitration mechanisms to prevent

data corruption and loss Seamless integration with the HP Insight Dynamics

VSE to double resource utilization and flex resources in real time to maintain service levels after a failover or disaster

Full support for both physical and virtualized server environments

Complete integration with utility pricing to synchronize movement of hardware resources and software licenses along with a service during failover

Cluster File System works in conjunction with Oracle ODM to improve I/O performance significantly and simplify application installation and ongoing management, particularly for Oracle Real Application Cluster databases. ODM can be used with both single-instance Oracle on a regular file system or CFS and with Oracle RAC over a CFS

Full scalability for large clusters with no degradation in failover speed

Easy to install and configure Simple cluster customization and replication

Figure 1: A basic Serviceguard cluster

4Table 1: HP Serviceguard 11.20 features and benefits

Features Benefits

New! Cluster-wide device file names

Ability to access the same storage device (disk or LUN) from all nodes in a cluster with the same name reducing time and effort when managing clustered storage by nearly 90%

New! Easy deployment Using a single command, volume groups, logical volumes, file systems, and mount point directories can be created and distributed to all package allowing a cluster to be created in five minutes

New! Multi-node Package Start Performance enhancements

90% faster application start-up by starting packages in parallel instead of sequentially

New! Live Application Detach Halt a node or the cluster without halting the packages that are currently running on those nodes to perform Serviceguard or Online Hardware changes while the application continues to run resulting in reduced planned downtime

New! Cluster Verification Verify that packages in the cluster are able to failover and that the cluster is configured correctly with a single and easy to use command that can also be automated to further reduce unplanned failures due to human errors

New! NFS file type support More storage configuration flexibility with Package Manager enhancement allowing support of NFS in a modular package as a file type required for certain applications

Integration with HP-UX 11i Dynamic Root Disk (DRD)

Significantly reduce planned downtime when performing upgrades to Serviceguard by performing upgrades on clones instead of active systems. Downtime is avoided except for that incurred when rebooting from the upgraded clone

Fast cluster reformation Reduces downtime and its associated revenue impact. Delivers full scalability of failover speed regardless of cluster size and the fastest failover on the market

IP level network monitoring across routers and subnets

More comprehensive and faster recovery from network failures; very flexible networking configurations; robust and rapid network detection for a large variety of failure scenarios; transparently restores network communications with redundant interfaces

Fully IPv6 compliant clustering Critical for organizations that require IP addresses for growth; supports federal mandates for IPv6; allows full IPv6 or mixed IPv4/IPv6 clusters; supports IPv6 for any subnet configured into the cluster, including subnets that carry the cluster heartbeat, and those connected to a quorum server

Robust monitoring Comprehensive protection against hardware, software, virtualization, network, and storage faults

Advanced cluster arbitration mechanisms

Prevents data corruption and loss. Flexible configuration options include lock LUN, lock disk, and quorum server.

Centralized logging and advanced log-filtering features

Enables a highly available log-consolidation server

Rapid automatic detection and recovery time

Prevents failed nodes from jeopardizing the integrity of application data

Ability to survive multiple node failures

Increases application availability and reduces operator error

Fast failback Provides superior protection

Rolling upgrades Minimize application downtime during system maintenance and upgrade

Multiple cluster configurationsactive-active, active-standby, and rotating standby

Provides flexible configuration options and full support for extensive virtualization models

Integrated workload balancing capabilities for static resource environments

Helps ensure that servers dont become overloaded after a failure in fixed resource environments by placing packages with fixed weights onto nodes with predefined capacities in event of a failure

Complete integration with HP Insight DynamicsVSE

Enables you to double resource utilization and flex resources in real time to maintain service levels after a failover or disaster. Helps to ensure that service-level objectives (SLOs) are maintained during planned and unplanned downtime

Full support for both physical and virtualized server environments

Cluster physical servers, nPartitions, vPartitions or HP Integrity Virtual Machines to re-act to software/hardware failures while assisting with consolidation

Complete integration with HP Utility Pricing

Synchronize movement of hardware resources and software licenses along with a service during failover

5Technical specifications HP Serviceguard A.11.20

Cluster types Active/Active, Active/Standby, Rotating Standby

Quorum Yes

Maximum nodes Fibre Channel: 16, Single-path SCSI: 2, Multi-path SCSI: 4

Operating systems HP-UX 11i v3

Supported nodes All HP Integrity and HP 9000 servers (including mixed clusters) HP Integrity Virtual Machines, vPars, nPars

Fibre Channel storage HP StorageWorks Modular Storage Array (MSA), HP StorageWorks Enterprise Virtual Array (EVA)HP StorageWorks XPEMC SymmetrixEMC CLARiiON

SCSI storage HP StorageWorks Modular Storage Array

File systems Base VxFS, Full VxFS, CFS, and NFS

Volume Managers LVM, VxVM, CVM

Data replication HP Continuous Access for XP and EVA Disk ArraysVeritas MirroringMirrorDisk/UXEMC SRDF

Application Integration Toolkits andExtensions

Serviceguard NFS ToolkitEnterprise Cluster Master ToolkitServiceguard Extension for SAPServiceguard Extension for RACServiceguard Extension for Oracle E-Business SuiteServiceguard Toolkit for Oracle Data GuardHP Developers Toolbox

Disaster recovery Yes (described later)

Interconnects Gigabit Ethernet, HyperFabric, Fibre Channel, InfiniBand, 10GbE

Management HP Serviceguard Manager, HP OpenView, HP System Insight Manager (described later)

6Serviceguard ManagerHP Serviceguard Manager is a graphical user interface (GUI) that provides configuration, administration, and monitoring capabilities for Serviceguard, Serviceguard Extension for SAP, Serviceguard Extension for RAC, Metrocluster, and Continentalclusters.

Serviceguard Manager provides you with a summary of the health of the cluster including the status of each node and each package. Operators see a color-coded summary of the health and configuration of all clusters, including the status of each node and each package, and have the ability to drill down and proactively manage clusters, systems (nodes), and packages that run applications.

Serviceguard Manager delivers effective control of infrastructure and applications with end-to-end management from hardware chassis to applications through integration with HP Systems Insight Manager (SIM), HP Virtualization Manager, and HP OpenView.

Serviceguard Manager is a Web-based, HP System Management Homepage (HP SMH) plug-in application.

CapabilitiesCluster Topology Map to view the cluster status at

a glanceMaintains consistency of configuration files on

cluster member nodes

Provides administration for clusters, nodes, and packages

Command line and GUI options for management and configuration of multiple clusters

Centralized logging and advanced log-filtering features enable a highly available log-consolidation server

Command fan-out tools to execute shell commands and distribute files across your Serviceguard clusters

Supports administration functions including: Run, halt, or move a package through drag and drop (standalone version)

Run or halt a node or a cluster

Easy to understand confirmation dialogs

Monitors functions to act on changing status Alerts Panel reports cluster status in one place

SNMP event browser displays all SNMP traps generated by Serviceguard nodes (standalone version)

Auto-Refresh provides automatic updates including property sheets

Provides command line preview of configuration operations

Package dependency graph

7Table 2: HP Serviceguard Manager B.03.00 features and benefits

Features Benefits

New! Easy Deployment Simplified creation of clusters using a graphical wizard that creates a simple cluster with just a few clicks

New! Easy Storage Configuration Wizard to automatic discover available file systems and volume groups across a cluster to accelerate storage configuration by up to 50%

New! SGeSAP management Intuitive wizard to cluster SAP using Serviceguard Extension for SAP (SGeSAP) for, step by step configuration and monitoring

New! Cluster Verification Quickly validate changes done to the cluster to ensure a cluster is configured correctly and will behave as expected

New! Live Application Detach Halt a node or the cluster to perform Serviceguard or hardware changes while applications continue to run

New! Metrocluster management enhancements

Metrocluster modular package configuration, monitoring and administration including the Metrocluster site aware clusters capability

Integrated package dependencies Precisely control dependencies between packages without custom scripting for orderly start-up/shutdown of inter-dependent applications to save up to 25% on administrator time and effort

Online Cluster Maintenance Change the configuration of the quorum server, the lock disk, the lock LUN or change an interface from IPv4 to IPv6, or vice versaall while the cluster is running

Online Package Maintenance Perform configuration changes to modular and legacy packages while the package is running

Command preview Preview the effect on packages of certain actions or events before they actually occur

Partial Start-up Maintenance Mode Allows a package to be put into a partially started state for maintenance purposes and easier iterative testing during deployment

Command Fan-out Tools with DSAU

Executes shell commands and distribute files across your HP Serviceguard clusterssend the same command to all systems in your configuration in a single action

Configuration synchronization with DSAU

Specify a specific server as your configuration master with all other systems defined as clients. The configuration master retains copies of all files that you want synchronized across your clients. Synchronization actions can include updating client files from the configuration master, executing shell commands, or checking for certain processes

Log consolidation with DSAU Logs from all systems you manage, whether in a cluster or not, can send their logs (system, package, and cluster) to a single location on the consolidation node from which they are easily monitored. GUI front-end and also integrated into Serviceguard Manager.

Cluster Topology Map Comprehensive monitoring, color-coded, graphically-intuitive icons, online status, and configuration information for multiple clusters and their member nodes. HP Serviceguard Manager enables a user to view cluster status at a glance and manage a cluster via graphical map; alerts panel reports cluster status in one place; auto-refresh provides automatic updates including property sheets.

Unified manageability Fully integrated with HP Systems Management Homepage and HP Systems Insight Manager. Identify nodes that are VMware guests, Integrity VM hosts or Integrity VM guests, nPartitions or vPars with context sensitive links to Partition Manager (on nPartitions) and VM Manager (on HP Integrity Virtual Machines)

HP Serviceguard Manager B.03.00

Supported clusters Plug-inServiceguardversionA.11.17.01orlater

ServiceguardExtensionforRACA.11.17.01orlater

ServiceguardExtensionforSAPB.05.10orlater

EnterpriseClusterMasterToolkitB.06.00orlater

NFSToolkitA.11.31.07orlater

Metrocluster(allversionscompatiblewithServiceguardversionsA.11.17.01orlater)

Continentalclusters(allversionscompatiblewithServiceguardversionsA.11.17.01orlater)

Software requirements Plug-inAnyWebbrowsersupportedbyHPSMH

Technical specifications

8Serviceguard ToolkitsHP Serviceguard Toolkits enable plug and play application integration. Toolkits describe how Serviceguard should monitor a specific application or service, and what to do in the event of a failure. Because of the unique way in which Serviceguard is designed, our simple Toolkit framework presents a standardized, simple mechanism to integrate virtually any application or service into a cluster.

CapabilitiesConfidence that services are integratedcorrectlyFully supported by HPReduces time and effortA common look and feel for each

clusteredapplication

Serviceguard NFS ToolkitHP Serviceguard Network File Server (NFS) Toolkit includes the configuration files and control scripts that allow you to use Serviceguard to set up highly available NFS servers. The configuration files are customized for each end users environment. Also included is a special monitor script that is used to monitor the health of NFS services. If a service fails, the monitor issues a restart on the local node or triggers a node switch.

Serviceguard Enterprise Cluster Master ToolkitHP Serviceguard Enterprise Cluster Master Toolkit (ECMT) accelerates the integration of major industry databases and Internet servers into a Serviceguard cluster. This bundle includes Serviceguard Toolkits for Oracle single-instance database, Apache, Tomcat, CIFS/9000, DB2, MySQL, and Sybase.

New! Serviceguard Toolkit for Oracle Data GuardThis Toolkit works in conjunction with Serviceguard to provide additional levels of availability for Data Guard itself by monitoring system resources like network, volume groups, and file systems, and initiating failover when needed. Using the Serviceguard Toolkit for Oracle Data Guard, Data Guard can then be integrated with Serviceguard Solutions for Disaster Recovery to provide disaster protection for Oracle databases as well as for the applications that use the databases.

Serviceguard Developers ToolboxHP Serviceguard Developers Toolbox enables ISVs and customers to develop their own toolkits quickly with a validation loop back to the HP development team at no charge.

9Technical specifications HP Serviceguard Enterprise Cluster Master Toolkit B.06.00

Databases

Oracle 10g R1, 10g R2, 11g R1, 11g R2 (on HP Integrity with HP-UX 11iv3) with or without ASM

IBM DB2 V9.1 (HP 9000 and HP Integrity)

V9.5 on HP Integrity servers only

V9.7 on HP Integrity servers

MySQL Server 5.0.56 and later

Sybase ASE 15.0.2 and later

Internet Servers

HP Apache B.2.0.55.03 and later

HP CIFS/9000 A.02.03.01 and later

HP Tomcat B.5.5.9.04 and later

Table 3: HP Serviceguard Solutions 11.20 Toolkit features and benefits

Features Benefits

Pre-written, tested and supported Toolkits for error-proof clustering of key applications and databases

Extensible infrastructure to accommodate any application and its resources. Available for all major services, applications and databasesfully supported, tested, and qualified by HP

ServiceguardEnterpriseClusterMasterToolkit

ServiceguardNFSToolkit

New! HP Serviceguard Toolkit for Oracle Data Guard

Common look and feel for each cluster service

Reduces time and effort, and dependency on expertise

Serviceguard Developers Toolbox Facilitates quick, easy, and correct integration of an application into a Serviceguard cluster with a validation loop back to HP

10

Table 4: HP Serviceguard Extension for SAP B.05.00 features and benefits

Features Benefits

Ease of management Modular package support for easier configuration. Integrated with Serviceguard Manager for ease of configuration and administration. A wizard graphical user interface helps the user to configure a modular SGeSAP package and automatically discovers SAP applications and fills in parameters for the user whenever possible

Co-developed with, fully tested and endorsed by SAP 100% confidence in the resiliency and robustness of your SAP environment

Automatic upgrades if purchased with a support contract No need to devote staff to updating or maintaining SAP integration into the cluster

Comprehensive availability and monitoring Protects the SAP Central Instance, SAP System Central Services, SAP Dialog Instances, SAP Replication Instances, SAP liveCache and underlying Oracle or SAP MaxDB databases. Software monitors are available for SAP enqueue services, SAP enqueue replication and SAP MaxDB database services

Online maintenance Reduces downtime even when you are performing maintenance or installing patches

Integrated with HP Serviceguard Disaster Recovery Solutions Protection in the event of site outages or natural and man-made disasters

Integrated with HP Serviceguard Extension for RAC (SGeRAC) Reduces the complexity for integrating SAP environments with Oracle RAC to ease deployment and configuration of high-end complex SAP environments that require and can sustain heavy workloads

Enables SAP virtualization Enables SAP virtualization

Utilizes SAP Replicated Enqueue technology Fully automated and transparent zero-impact failover for SAP Single Points Of Failure (SPOFs) in the software stack

Helps to optimize resource utilization No need for a dedicated secondary node to serve as backupduring normal operation, the secondary node can function as an application server in the production SAP environment or run as a separate SAP development or test instance

Hot Standby Service for SAP liveCache Reduce recovery time for SAP liveCache from hours to less than two minutesendorsed by SAP as the first solution of its kind

Serviceguard ExtensionsHP Serviceguard Toolkits enable plug and play application integration. Toolkits describe how Serviceguard should monitor a specific application or service, and what to do in the event of a failure. Because of the unique way in which Serviceguard is designed, our simple Toolkit framework presents a standardized, simple mechanism to integrate virtually any application or service into a cluster.

New! Serviceguard Extension for Oracle E-Business SuiteHP Serviceguard Extension for Oracle E-Business Suite (SGeEBS) delivers fast and foolproof high availability for Oracles E-Business Suite (EBS) of enterprise resource planning (ERP), customer relationship management (CRM), and supply-chain management (SCM) applications relying on Oracle database. SGeEBS handles all of the complexities of Oracle EBS for you to automatically and transparently address both planned and unplanned downtime.

Serviceguard Extension for SAPHP Serviceguard Extension for SAP (SGeSAP) extendsServiceguards powerful failover capabilities to SAP environments. It continuously monitors the health of each SAP component and automatically responds to failures or threshold violations.

SGeSAP provides a single, uniform interface to cluster SAP Netweaver systems and ABAP-stack and/or JAVA-stack based SAP applications in a vast range of supported release versions. The Serviceguard Extension for SAP also clusters underlying databases. On demand, SAP liveCache cluster packages or SAP Master Data Management packages can be created.

You can combine all clustered components of a single SAP system into one failover package for simplicity and convenience. There is also the full flexibility to split components up into several packages to avoid unwanted dependencies and to lower potential failover times. Multiple SAP applications of different types and release versions can be consolidated in a single cluster.

SGeSAP B.05.00 requires the use of either the HP Serviceguard NFS Toolkit or the HP Serviceguard Cluster File System.

CapabilitiesCo-developed with, fully tested, and recognized

bySAPAutomatic upgrades if purchased with supportProtects the SAP Central Instance, SAP System Central

Services, SAP Dialog Instances, SAP Replication Instances, SAP liveCache, SAP MDM, and underlying Oracle, IBM DB/2, or SAP MaxDB databases

Reduces downtime while performing maintenance

11

Can be integrated with HP Metrocluster for disaster-tolerant SAP environments

Can be integrated with HP Serviceguard Extension for RAC

Enables SAP virtualization capabilities by moving redundant SAP Application Server Instances between hosts to quickly adapt to changing resource demands or maintenance needs

Ability to utilize SAP Replicated Enqueue technology to become a complete cluster solution which provides fully automated and transparent zero-impact failover for SAP Single Points Of Failure (SPOFs) in the software stack

No need for a dedicated secondary node to serve as backupduring normal operation, the secondary node can function as an application server in the production SAP environment or run as a separate SAP development or test instance

Reduce recovery time for SAP liveCache from hours to less than two minutes

Serviceguard Extension for RACTo create a completely available solution, the application environment should be designed to remove all single points of failure and to reduce the impact of various component failures. Serviceguard Extension for RAC (SGeRAC) provides a comprehensive and consistent continuity solution for the database, non-database data, and higher-level applications for Oracle 10g and 11g RAC environments.

High availability clusters configured with Oracle Real Application Cluster software are known as RAC clusters. In Figure 2, two loosely coupled systems (each one known as a node) are running separate instances of Oracle software that read data from and write data to a shared set of disks. Clients connect to one node or the other through LAN. RAC on HP-UX lets you maintain a single database image that is accessed by the HP servers in parallel, thereby gaining added processing power without the need to administer separate databases. SGeRAC provides a highly available database that continues to operate even if one hardware component should fail. RAC cluster configuration is designed so that in the event of a node failure, another node with a separate instance of Oracle can continue processing transactions.

Clients

Heartbeat/Data subset

DLM and Heartbeat subset

RAC Inst 2

Redundant RAC data,logs, and control files

Disk Array

PackageB

RAC Inst 1

PackageA Cache fusionRoot

Mirror

Root

Mirror

Node 1 Node 2

Figure 2: Sample SGeRAC configuration

12

Table 5: HP Serviceguard Extension for RAC 11.20 features and benefits

Features Benefits

Support for Cluster wide Device Special Files (cDSFs) Simplified storage configuration through cluster wide device file name. cDSF ensures that each storage device used by the cluster (or any group of nodes within the cluster) has a unique file name

Support for Cluster Interconnect Monitoring on IPv4 and IPv6 subnets

Further improve monitoring of subnet networks required in Oracle RAC to respond more quickly to NIC outages or subnet problems that might otherwise cause nodes to be evicted with a subsequent loss of RAC service

Enablement of improved RAC VIP management and failover performance and VIP network monitoring on IPv4 and IPv6

Simplifies network configuration in a RAC environment and enhances the availability of networks needed by Oracle clients.

GMS Authorization Prevent unauthorized users from accessing cluster information and registering or deregistering group memberships to further secure the environment so that only certain users can modify it according to your security policies

Joint support from HP and Oracle Single point of contact, comprehensive inclusion of both HP and Oracle best practices

Advanced cluster arbitration mechanisms and the use of robust volume managers

Helps to eliminate data corruption and preserve data integrity

Robust I/O fencing Provides data integrity protection among nodes, both inside and outside of the cluster

Supports Cluster File System (CFS), Shared Logical Volume Manager (SLVM), Cluster Volume Manager (CVM), Automated Storage Management (ASM) on SLVM and raw volumes

Flexible storage management options

Seamless integration with Serviceguard Manager, HP System Management Tools, and both Serviceguard CFS and HP Volume Managers

Improved manageability

Local area network (LAN) monitoring Delivers rapid recovery and transparent restoration of service following events like hardware network interface failures

Increased storage subsystem availability Eliminates the impact of storage link failures by providing multiple independent pathways to storage and comprehensive monitoring services

Ability to withstand multiple node failures Better node availability, service continues to run even in case of multiple node failure

Rolling upgrades Maintain application availability during hardware and software maintenance

Integration with HP-UX Virtual Server Environment Double resource utilization and flex resources in real time to maintain service levels after a failover or disaster.

Can be integrated with HP Serviceguard Disaster Recovery Solutions Protect against site outages, natural, and man-made disasters

13

CapabilitiesAdvanced cluster arbitration mechanisms and the

use of robust volume managers help remove data corruption and preserve data integrity

Robust I/O fencing provides data integrity protection among nodes, both inside and outside of the cluster

Flexible storage management options for Oracle RAC: Cluster File System (CFS), Shared Logical Volume Manager (SLVM), Cluster Volume Manager (CVM), Automated Storage Management (ASM) on SLVM, and raw volumes

Manageability with Serviceguard ManagerSeamless Integration with HP System Management

tools and both Serviceguard CFS and HP Volume managers

Increased storage subsystem availabilityLocal area network (LAN) monitoringAbility to withstand multiple node failuresRolling upgrades and DRD support help maintain

application availability during hardware and software maintenance

Integration with HP-UX Virtual Server EnvironmentExtended-distance clustering facilitates

disaster-tolerant protectionSupport for Oracle ASMJoint support from HP and Oracle

How SGeRAC can enhance the availability of an Oracle RAC environment1. Enhanced cluster availability

2. Enhanced network link availability

3. More robust data integrity

4. Faster failure detection times

5. Rich Disaster Recovery solutions integration including three data center architectures

6. Storage management choice, flexibility, and consistency

7. Improved manageability

8. Easier integration of other cluster applications

9. No dependency on the database stack for other applications

10. Value-added volume manager features with Storage Management Suite

Technical specifications HP Serviceguard Extension for RAC A.11.20

Maximum nodes 16

Cluster networks and interconnects Gigabit Ethernet, HyperFabric, FC, InfiniBand, 10GbE

Storage supported HP XP, VA, EVA, MSA1000, MSA1500; EMC Symmetrix

File systems supported Online JFS (VxFS), JFS, CFS (Cluster File System)

Cluster-wide I/O device access SLVM, CVM/CFS

Failover configurable Active to standby Active to active Rotating standby

Multi-pathing Dynamic multi-pathing and native multi-pathing in HP-UX 11i v3

Volume management SLVM, Cluster Volume Manager (CVM)

Single-view cluster management Yes, SG Manager with HP SIM

14

Serviceguard Disaster Recovery SolutionsFor some installations, the level of protection provided by a local cluster is insufficient. Consider an order processing center where power outages are common during harsh weather, or the systems running a stock market where multiple system failures, for any reason, would have a significant financial impact. For these installations, and many more like them, it is important to guard not only against single points of failure, but also against multiple points of failure or single massive failures that cause many components to fail (such as the failure of a data center, an entire site, or a small area). Configurations that protect against these types of failures are considered disaster tolerant.

Serviceguard Disaster Recovery Solutions put greater distance between the nodes so that an outage at one site does not disable the entire cluster. Nodes can be put in different rooms, on different floors of a building, or even in separate buildings or cities. The distance between the nodes is dependent on the types of disaster from which you need protection and on the technology used to replicate data. HP has helped deliver Serviceguard Disaster Recovery Solutions to its customers for more than 20 years and introduced one of the first solutions on the market. Serviceguard Disaster Recovery Solutions utilize multiple data centers and multiple copies of your companys data so that if one data center fails, a second one continues to operate. Regardless of the distance, you can survive the loss of a data center and maintain access to critical data and applications. The resultclusters

that are resistant to multiple points of failure and to singular, massive failures.

Extended Distance ClusterHP EDC is a host-based data-replication configuration that can span up to 100 kilometers. With this configuration, a single Serviceguard cluster can be implemented across two or three data centers. Since Serviceguard is the only cluster license required, it is an inexpensive and easy-to-implement solution. Additionally, you can choose and mix any Serviceguard-supported storage with this solution. Each clustered server is directly connected to all storage in both data centers.

Extended QuorumTie Breakers

Extended Node A Extended Node B

XP Disk Array XP Disk Array

Business Copy forConsistency

HP MirrorDisk/UX

Synchronous replication

15

HP Extended Distance Cluster for RAC (EDC for RAC)Extended Distance Cluster can also be tailored to Oracle RAC environments. EDC for RAC delivers failover capabilities for a single RAC database that is replicated and synchronized across two remotely located data centers (up to 100 kilometers apart) in an active/active configuration. EDC for RAC also supports HP Serviceguard Storage Management Suite and HP Cluster File System capabilities.

Table 6: HP Serviceguard Extension for SAP B.05.00 features and benefits

Features Benefits

Natural extension of Serviceguard from multi-system to multi-data center

Comprehensive, robust and unified solution for varying levels of availability

Host based data mirroring Can work with any Serviceguard supported storage

Synchronous data replication Data is available at both the sites for minimum impact on RPOs (recovery point objectives)

MetroclusterSimilar to Extended Distance Cluster, a Metrocluster isa normal Serviceguard cluster that has clustered nodes and storage devices located in different data centers separated by some distance. But unlike Extended Distance Cluster, which has a maximum range of 100km, Metrocluster operates at distances of up to 300 km and also verifies the currency and integrity of your data. Both data centers can be active, protected, and capable of providing application failover for each other.

Metrocluster can act as a single DR solution for databases as well as other applications. The distinct characteristic of Metrocluster is its integration with array-based data replication. Currently, Metrocluster implements three different solutions:Metrocluster with Continuous Access XP for

HP StorageWorks XP ArraysMetrocluster with Continuous Access EVA for

HP StorageWorks EVAMetrocluster with EMC SRDF for EMC Symmetrix

ArraysMetrocluster for RACMetrocluster includes a functionality called Metrocluster Site Aware Disaster Tolerant Architecture (SADTA) which provides support for Oracle RAC databases using novel techniques. Metrocluster for RAC makes it possible to provide disaster tolerance to RAC databases without impacting performance. The RAC databases can also be stacked, allowing flexibility. Both primary and standby sites can have active databases which enhances utilization. SADTA uses sub-clustering to provide disaster recovery for Oracle 10g and 11g RAC to provide a combination of features, namely:Automatic, robust failoverNo impact on RAC performance due to

inter-site latenciesActive/active per RAC database within a site and

active/passive across sitesBased on synchronous/asynchronous array-based

data replication that is more robust than volume manager-based mirroring

Metrocluster third location forTie Breakers

Metrocluster Site 1 Metrocluster Site 2

Disk Array Disk ArrayBusiness Copy for

ConsistencyContinuous Access

Synchronous replication

16

Technical specifications Metrocluster

Maximum distance 300 km*

Maximum sites 2**

Maximum clusters 1

Cluster size Up to 16 total nodes

Hardware support HP Integrity, HP 9000

Operating system support HP-UX 11i v1, HP-UX 11i v, HP-UX 11i v3

Oracle database support Oracle 9i, Oracle 10g, Oracle 10g RAC, Oracle 11g RAC, Oracle 11g

Network subnets Single/separate IP subnet per site

Interconnect Dense Wavelength Division Multiplexing (DWDM)

Failover type Automatic

Bi-directional failover Yes

Fast failover and failback No

Data replication Continuous Access EVA Continuous Access XP EMC SRDF

Replication mode Continuous Access EVA: Synchronous and asynchronous

Continuous Access XP: Synchronous and asynchronous

EMC SRDF: Synchronous and asynchronous

Replication link Any TCP/IP, Dark Fibre, ATM, SONET/SDH, DWDM

Cluster quorum Quorum server or arbitrator nodes

* Maximum distance listed are indicative numbers. Actual distance depends on the latency requirement of SG heart beat (200 ms) and the replication technology, whichever is smaller. Technically higher distances are possible using networks with less latency.

** Two main data centers and a third location for quorum server or arbitrator nodes

Optionally, you can now deploy with the CFS, configured at each site to make managing a RAC environment easier

The environment is managed by a new component called the Site Controller that means you can manage startup and relocation (static load-balancing) of RACdatabases

Multiple stacked RAC databases can be deployed easily using the SGeRAC integration toolkit

Capabilities Single DR solution for Oracle RAC databases as

well as other applicationsSupport for 3 data center configurations for any

type of workload

Fully integrated with the HP Serviceguard Solutions portfolio

No impact to Oracle RAC performance due to inter-site distances

A single solution that integrates high availability and disaster tolerance for Oracle RAC that provides: Local high availability

Remote disaster tolerance

Static load balancing between sites

Dynamic load balancing within a site

Storage and network flexibilityFully integrated and supported by HP, Oracle,

andSymantec

17

Table 7: HP Serviceguard Solutions 11.20 metropolitan cluster features and benefits

Features Benefits

Support for three data center configurations for all types of workloads (Unique to HP!)

Enables the configuration of disaster recovery between three separate sites, with the two primary sites configured in a Metrocluster and a third data center configured via Continentalclusters as the recovery site. This ensures the highest level of availability, even in the event of the primary zone with the two primary sites becomes unavailable.

Support for Manual Site switching Allows operator intervention for a Metrocluster failover to a remote site for greater flexibility and greater control of failovers

Integrated with Serviceguard Manager Save on the time and effort required to manage Metrocluster by using the Serviceguard Manager GUI based interface to configure, monitor, and administer packages

Support for Modular Packages Allows easier configuration of applications by reducing the number of scripts 66%.

Metrocluster with SONET/SDH site interconnect Lower cost alternative to Dense Wave Division Multiplexing (DWDM) ideal for smaller configurations

Per core licensing Makes disaster recovery feasible even for smaller environments - size your environment as you see fit, then choose either per core or per cluster licensing

Support for Thin Provisioning storage virtualization Reduce storage management time and costs, storage acquisition costs and deliver full high availability and disaster tolerance

Single disaster recovery solution for Oracle RAC databases as well as other applications

Provides a unified all-in-one solution for local high availability, remote disaster recovery, stacking of multiple databases, static load balancing between sites, dynamic load balancing within a site and storage and network flexibility

Robust, automatic failover over of entire multi-instance RAC databases between two sites without impact to Oracle RAC performance

Maintain service levels

Fully integrated and supported by HP, Oracle, and Symantec Delivers 100% confidence in the resiliency and robustness of your disaster tolerant environment

ContinentalclustersHP Continentalclusters is the most flexible Serviceguard Disaster Recovery Solution. Unlike Extended Distance Cluster and Metrocluster with their single-cluster architectures, Continentalclusters uses multiple clusters to provide application recovery. Applications run in the active/standby mode with application data replicated between data centers by either storage array-based data replication products (such as Continuous Access XP, Continuous Access EVA, or EMC SRDF), or software-based data replication (such as Oracle 8i Standby DBMS and Oracle 9i Data Guard).

Two types of connections are needed between the two Serviceguard clusters in this architecture: one for inter-cluster communication and another for datareplication. Depending on the distance between the two sites, either LAN (that is single IP subnet) or WAN connections may be used for cluster networkcommunication.

Asynchronous replicationunlimited distance

BusinessCopy

Continentalclusters Site 1 Continentalclusters Site 2

HP Continuous AccessExtension

WAN

Continentalclusters

XP Disk Array XP Disk Array

Figure 6: HP-UX 11i Serviceguard Continentalclusters

18

Technical specifications Continentalclusters A.07.01

Maximum distance Unlimited

Maximum sites 4*

Maximum clusters 4

Cluster size Up to 64 total nodes (4 sites)

Hardware support HP 9000/HP Integrity

Operating system support HP-UX 11i

Oracle database support Oracle 9i, Oracle 10g, Oracle 10g RAC, Oracle 11g, Oracle 11g RAC

Network subnets Single IP subnet for multiple clusters or separate IP subnet per cluster

Interconnect Ethernet within each data center, LAN or WAN between data centers

Failover type Push-button style automated recovery

Bi-directional failover Yes

Fast failover and failback Yes (only for local failover within a site cluster)

Data replication Continuous Access XP, EVA; EMC SRDF, Oracle Standby Database

Replication mode Synchronous and asynchronous

Replication link Any TCP/IP, Dark Fibre, ATM, SONET/SDH, DWDM

Cluster quorum Quorum server or cluster lock disk for each cluster

* Two main data centers and a third location for quorum server or arbitrator nodes

Table 8: HP Serviceguard Continentalclusters A.08.00 features and benefits

Features Benefits

Cluster to cluster failover Allows failover of an entire cluster

Support for Modular Packages All packages in a Continentalclusters recovery group can be configured as a modular package, simplifying the configuration process.

Mutual Recovery Allows for mission-critical services to run on both clusters, with each cluster configured to recover the mission-critical services of the other

Disaster Recovery over unlimited distances

Allows you to choose the location of your data center based on your priorities

Push-button failover Initiates a failover when you deem fit. Provides an automated recovery procedure that will be triggered only after a manual confirmation of your intention to recover.

Integrated with Serviceguard Extension for Real Application Cluster (SGeRAC)

Provides disaster tolerance to your Oracle RAC database over unlimited distances

CapabilitiesProvides the ability to monitor a Serviceguard cluster

and failover mission-critical services to another cluster if the monitored cluster should become unavailable.

Supports mutual recovery, which allows for mission-critical services to run on both clusters, with each cluster configured to recover the mission-critical services of the other.

Supports Serviceguard Extension for Real Application Cluster (SGeRAC) in addition to Serviceguard. In an SGeRAC configuration, Oracle RAC database

instances are simultaneously accessible by nodes in the same cluster (that is the database is only accessible to one site at a time). The Oracle database and data are replicated to the second data center. The RAC instances are configured for recoverability so that the second data center stands by, ready to begin processing in the event of a site failure at the first data center (that is across sites; this is an active/standby configuration such that the database is only accessible to one site at a time).

19

Serviceguard Storage Management SuiteHP Serviceguard Storage Management Suite (SGSMS) delivers improved availability, performance, and manageability for Oracle Database and Oracle RAC environments on HP-UX 11i. It achieves these enhancements by integrating HP Serviceguarda tested and proven UNIX high-availability productwith Veritas Storage Foundation technologies by Symantec. This suite of products also includes features that can speed data recovery from hours to minutes, improve manageability, and reduce storage cost. With SG SMS, you can enhance your existing environment while reducing the chance of human error. Additionally, it offers full integration with the HPVSE to boost your resource utilization further. SG SMS is ideal if you require the highest levels of availability and outstanding Oracle Database performanceor if you have an application that would benefit from the improved manageability and scalability offered through a clustered file system.

CapabilitiesIntegrates HP Serviceguard with Veritas Storage

Foundation Available on HP-UX 11i v3CFS simplifies application installation and ongoing

management, particularly for Oracle RAC databasesAdvanced volume management and file system

capabilities improve resource utilization

Variety of online administration options, including storage reconfiguration, volume and file system creation and resizing, dynamic storage expansion, and dynamic storage reclamation of unused capacity

Improve database performance up to 10x using Oracle Disk Manager (ODM)close to raw-volume performance

Dynamic Storage Tiering reduces storage costs dramatically by enabling administrators to set up policies that places files on different classes of devicesfor example, inexpensive or expensive devices, or fast or slow devices

FlashSnap creates a database clone or point-in-time copy, such as a duplicate database on a secondary host for off-host processing and backups

Storage Checkpoint provides an instant database backup image that can be taken more frequently and can be beneficial in reducing the effort required to backup a file systemwithout necessarily involving a Storage Administrator

Storage Rollback can be used to roll back an entire database image, tablespace or specific data files, hence providing fast restore option for user errors

Portable Data Containers promote seamless data migration between platforms

Dynamic multi-pathing feature improves performance by allowing you to distribute I/O loads across multiple data paths

Storage mapping to tune performance of disk arraysI/O fencing enables rapid application recovery in the

event of a failure by isolating your storage resources from failed nodes that are not part of the cluster

Table 9: HP Serviceguard Storage Management Suite A.03.00 features and benefits

Features Benefits

Serviceguard integration with Cluster File System (CFS) Simplifies application installation and ongoing management, particularly for Oracle RAC databases

Advanced volume management and file system capabilities Improves resource utilization

Integration with Oracle Disk Manager (ODM) Dynamic Storage Tiering

Improves database performance 10x; reduces storage costs dramatically by enabling administrators to set up policies that segment data

FlashSnap database clones or point-in-time copies Almost eliminates the need to take applications offline for storage management

Storage mapping I/O fencing Tunes performance of disk arrays; enables rapid application recovery in the event of a failure by isolating your storage resources from failed nodes that are not part of the cluster

Complete online administration capabilities Significantly reduces storage management time

Portable Data Containers Promotes seamless data migration between platforms

Dynamic multi-pathing Guards against broken links to storage, with tools to monitor availability of storage

Get connectedwww.hp.com/go/getconnected

Current HP driver, support, and security alerts delivered directly to your desktop

Copyright 20092010 Hewlett-Packard Development Company, L.P. The information contained herein is subject to change without notice. The only warranties for HP products and services are set forth in the express warranty statements accompanying such products and services. Nothing herein should be construed as constituting an additional warranty. HP shall not be liable for technical or editorial errors or omissions contained herein.

Oracle is a registered trademark of Oracle Corporation and/or its affiliates. Java is a US trademark of Sun Microsystems, Inc. UNIX is a registered trademark of The Open Group

4AA2-6268ENW, Created May 2009; Updated September 2010, Rev. 2

To learn more about how HP Serviceguard Solutions can improve availability of your mission-critical applications, please visit www.hp.com/go/serviceguardsolutions

For technical documentation: www.hp.com/go/hpux-serviceguard-docs

Share with colleagues

HP ServicesServiceguard ServicesMake the most of the Serviceguard technology with careful planning; proper execution of critical design, installation, and start-up activities; and periodic audits to help you sustain the availability levels your users expect.

Mission-critical Serviceguard Implementation Services: Utilizes processes to help deliver rapid, effective deployment

Analysis and Planning: Assesses your hardware, software, and applications, for compatibility

Implementation: Installs and customizes your Serviceguard configuration

Serviceguard Audit: Helps keep your Serviceguard environment properly configured

Complete coverage of the IT environment (server, storage, SAN, network, operating systems, physical environment, and software) reduces downtime and meets your mission-critical service levels

HP Proactive 24 Service: Integrated hardware and software support, including proactive and reactive services, improves stability and availability across your environment

HP Critical Service: This comprehensive support solution helps reduce the business impact of downtime for your mission-critical applications

Rely on our extensive track record of helping customers improve their ability to support the IT systems that run their ever-changing business needs.

ServiceguardServiceguard ManagerServiceguard ToolkitsServiceguard NFS ToolkitServiceguard Enterprise Cluster Master ToolkitNew! Serviceguard Toolkit for Oracle Data GuardServiceguard Developers Toolbox

Serviceguard ExtensionsNew! Serviceguard Extension for Oracle E-Business SuiteServiceguard Extension for SAPServiceguard Extension for RAC

Serviceguard Disaster Recovery SolutionsExtended Distance ClusterMetroclusterContinentalclusters

Serviceguard Storage Management Suite

Digg 10: Twitter 10: Facebook 10: Linked in 10: