Compaq ProLiant 1000 Architecting and Deploying High-Availability Solutions - Page 9

Architecting and Deploying High-Availability Solutions 9 Clusters should be able to overcome the administrative failures described above. Specifically included is the fact that the application should be available while items like maintenance, hardware upgrades, software upgrades, and database updates are going on. Then there is the question of how many nodes a cluster can support as well as where the nodes physically reside. Fibre Channel, SCSI 2, UltraSCSI, and ServerNet technologies improve storage throughput and allow storage devices to be separated by greater distances while providing for the efficient, protected movement of data between clusters and storage subsystems. Channel Extenders do for CPUs what Fibre Channel and Ultra SCSI do for storage -- they provide for greater separation enhancing tolerance for disasters. Wide-Area and Global Clusters are separated by greater distances. The 500-miles-and-beyond capabilities that exist today easily qualify them for dealing with Regional events. Networks By their very nature, networks are capable of spanning enough distance. When you look at the redundancy and virtual capabilities that are available, they qualify as Regional solutions in the Availability Matrix. A network in which a two-phase commit process assures a transaction is not complete until the remote node confirms it certainly makes this high availability. The Compaq Technology Brief entitled Eliminating Single Points of Failure and Enabling Rapid Recovery in Server Subsystems (ECG0094/1197) provides more detail on increasing availability in basic components including processor, expansion, memory, primary and secondary storage, power, cooling, and management. 6. Architecting High-Availability Systems Analyzing and defining the requirements for a high-availability environment that meets your unique needs is not a trivial process. Add to that the challenges of solution deployment and the task can seem overwhelming because -- in addition to issues surrounding availability -- other factors must be considered, such as: Performance and scalability: Can the solution deliver the performance required? Can it scale appropriately to meet the size of the application? Affordability: Is the cost of purchasing, implementing, or managing the solution prohibitive or consistent with the operational loss you wish to prevent? Manageability: Does the organization have the knowledge and resources necessary to manage the solution over the long run? Service and support: What level of service and support is available? Is the infrastructure in place to support the solution reliably and affordably - wherever it is located? Turn to Compaq for the Answers Compaq unquestionably has the broadest range of availability products and solutions on the market today. We can provide better answers anywhere on the Availability Matrix with open, reliable solutions and technologies from hand-held PCs to the most scalable and available fault-tolerant and clusters systems. The professionals in the Compaq Enterprise Computing division helped invent the concepts of disaster tolerance, fault tolerance, and continuous availability almost two decades ago and have been delivering and supporting systems in the most demanding mission-critical environments ever since. The combined development, delivery, and support experience of Compaq, DIGITAL, and Tandem personnel is unsurpassed in the marketplace. ECG064/1198