Concept: [ Benefits | Advantages | Advanced technologies ]

PRIMEQUEST introduces the world's first "System Mirror“. This ensures high availability using the hardware alone. Enabled by the “Dual Synchronous System Architecture” it solves the issues of performance and reliability necessary for “Open” mission critical business. It does this by applying new functions to the latest server technology that balance the cost advantages and usability characteristics of IA servers.

The "Flexible I/O system" also changes for ever the way resource configuration is optimized and managed. While the in-built redundant server "Management Boards(MMB)," provide autonomous operation and reduce the administration costs across the life of the system.

System Mirror

RAID and Clustering are typical approaches to improving reliability in IA servers. UPS and backup systems are other approaches. But all such existing technologies don’t improve the reliability of the heart of the computer system - the motherboard itself. Fujitsu’s dedication to system redundancy and improved reliability has been extended to the motherboard. Introduction of the latest dedicated Fujitsu chipset has enabled the world's first fully-synchronized fall-back operation within a single IA server.

Because the System Mirror function duplicates every CPU instruction and actions both instruction tasks in a fully synchronized mode, it can detect faults in peripherals from task output inconsistencies across both mirrors. This always ensures normal operation and as everything is performed by hardware, the operating system and applications are never affected and can operate continuously. Unlike clustering systems, this method does not need a time interval to ensure correct isolation of the fault, nor any period for system fail-over. The result is fewer faults and 50% less down-time than conventional server systems.

Flexible I/O mechanism

Flexible I/O technology means motherboards (System Board: SB) and IO Units can be flexibly connected. In general server hardware bundles SBs and IO Units, so if the application or load requirements fluctuate, SBs cannot always use the full potential of the available IO Units and end up wasting resources. To solve this issue, Fujitsu has enabled flexible combinations of SBs and IO Units that match system loads and changing partition configurations to fully optimize resource use in the system.

As shown in the following chart, the use of a floater (backup) SB together with the flexible I/O mechanism means the floater SB can be quickly connected to the same I/O as any failed SB. In fact this shortens the average recovery time by 90% compared with conventional systems. This improves system availability tremendously. Similarly IO Unit failures can be completely separated from the SB, enabling easier exchange maintenance, while operation continues.

Advanced Partitioning

PRIMEQUEST can support up to 16 mission critical class partitions (PRIMEQUEST 580A) with its advanced partitioning technologies (PPAR and XPAR). They enable effective and granular use of CPU, Memory and I/O resources for each business process, to reduce overall total cost of ownership (TCO). Ongoing development will see PRIMEQUEST supporting virtual machine (VM) capability which enables the configuration of multiple virtual servers in a single server machine with even lower management overheads.

Server Management Board

Embedded inside every PRIMEQUEST are two (for redundancy) server Management Boards (MMB) as standard, and any PC with a web browser can be used for all operational and maintenance tasks. Of course in large heterogeneous networks a dedicated server for management of the various systems, using common management software, can also be connected. A dedicated redundant LAN interface is also available for independent monitoring and management of configuration changes. This removes any issues of overloads on system LAN ports.

The MMB provides a very comprehensive range of functions to reduce administration effort. Functions include, server configuration, usage monitoring, fault monitoring, partitioning, server management, power control and scheduling, configuration monitoring, software usage, automatic emergency e-mail alerts, plus maintenance support, covers help and guidance on system extensions and faulty unit exchange.