X hits on this document

263 views

0 shares

0 downloads

0 comments

25 / 43

  • I/O error recovery and system resilience to I/O card failures

  • I/O cards with fully isolated buses

  • Prevention of silent corruption of data going to I/O

  • Recovery of a single I/O-to-cell controller link failure

  • Recovery of a single crossbar wire failure

  • Localization of crossbar failures to the partitions using the link

  • Automatic de-allocation of bad crossbar link upon boot

  • ASIC full burn-in and high-quality production process, resulting in 10X improvement in ASIC failure

rates

  • Full “test to failure” and accelerated life testing on all critical HP Integrity Superdome assemblies

  • Strong emphasis on quality for multiple nPartition single points of failure (SPOFs)

  • System resilience to Guardian Service Processor (GSP)

  • Isolation of nPartition failures—failure of one nPartition will not bring down another partition

  • Protection of nPartitions against spurious interrupts or memory corruption

Supportability

Supportability is another key feature of the Integrity Superdome product line. Listed here are some of the supportability features of these servers:

  • Event Monitoring Service (EMS) (only available on HP-UX)

  • HP Systems Insight Manager support for Windows Server 2003

  • Support Management Station (SMS)

  • Instant Support Enterprise Edition (ISEE)

Event Monitoring Service

For the Integrity Superdome, HP Event Monitoring Service (EMS) on HP-UX aids system management by keeping track of the system’s vital signs. EMS is installed and activated by default. It allows monitoring of virtually all of the hardware on the system, including:

  • Mass storage

  • Memory

  • Fibre Channel components (multiplexer, switch, card, fabric, etc.)

  • I/O cards

  • ECC errors on the main system bus

  • ECC errors in the CPU cache

  • System temperature

  • Support processor problems

  • Processor-dependent hardware problems

  • Processor-dependent hardware low-battery conditions

  • Cabinet fans

  • Cabinet power supplies

  • Chassis code logging failure

  • Processor-dependent code (firmware) problems

The system hardware configuration and selected kernel parameters are also monitored.

25

Document info
Document views263
Page views263
Page last viewedFri Dec 09 03:17:51 UTC 2016
Pages43
Paragraphs1082
Words15511

Comments