Reliability, Availability, And Serviceability Features; Major Components Of The Compute Node - IBM NeXtScale nx360 M4 Installation And Service Manual

Reliability, availability, and serviceability features

Three of the most important features in compute node design are reliability,
availability, and serviceability (RAS). These RAS features help to ensure the
integrity of the data that is stored in the compute node, the availability of the
compute node when you need it, and the ease with which you can diagnose and
correct problems.
The compute node has the following RAS features:
v Advanced Configuration and Power Interface (ACPI)
v Automatic server restart (ASR)
v Built-in diagnostics using DSA Preboot
v Built-in monitoring for temperature, voltage, and hard disk drives
v Customer support center 24 hours per day, 7 days a week
v Customer upgrade of flash ROM-resident code and diagnostics
v Customer-upgradeable Unified Extensible Firmware Interface (UEFI) code and
v ECC protected DDR3 DIMMs
v ECC protection on the L2 cache
v Error codes and messages
v Integrated management module II (IMM2)
v Light path diagnostics
v Memory parity testing
v Microprocessor built-in self-test (BIST) during power-on self-test (POST)
v Microprocessor serial number access
v Processor presence detection
v ROM-resident diagnostics
v System-error logging
v Vital product data (VPD) on memory
v Wake on LAN capability
v Wake on PCI (PME) capability

Major components of the compute node

Use this information to locate the major components on the compute node.
The following illustration shows the major components of the compute node.
1. Service availability varies by country. Response time varies depending on the number and nature of incoming calls.
10
IBM NeXtScale nx360 M4 Type 5455: Installation and Service Guide
Enclosure. This policy is enforced when the initial power is applied to the IBM
NeXtScale n1200 Enclosure or when a compute node is inserted into the IBM
NeXtScale n1200 Enclosure.
The following settings for this policy are available:
– Basic power management
– Power module redundancy
– Power module redundancy with compute node throttling allowed
diagnostics
1