Table of Contents
Advertisement
Recovery component
The recovery component gathers network hardware events and frame-level events. This
component also logs each event in the CNM_ERRLOG and sends most events to TEAL.
The recovery component also performs some event consolidation to avoid flooding the TEAL
with too many messages in the event of a CEC power up or power down.
Performance counter data management
This data management periodically collects ISR and HFI aggregate performance counters
from the hardware and stores the counters in the cluster database. The collection interval and
amount of data to keep are configurable.
Command handler
This handler is a socket listener to the command module. This handler manages ISNM
commands, such as requests for hardware status, configuration for LNMC, and link
diagnostics.
IBM High Performance Computing Hardware Server
In addition to the CNM software components, the HPC Hardware Server (HWS) handles the
connections to the service network. Its primary function is to manage connections to service
processors and provide an API for clients to communicate with the service processors. HWS
assigns every service processor connection a unique handle that is called a
number
hardware.
In a Power 775 cluster, HPC HWS runs on the EMS, and on each xCAT service node.
1.9.2 DB2
IBM DB2 Workgroup Server Edition 9.7 for High Performance Computing (HPC) V1.1 is a
scalable, relational database that is designed for use in a local area network (LAN)
environment and provides support for both local and remote DB2 clients. DB2 Workgroup
Server Edition is a multi-user version of DB2 packed with features that are designed to reduce
the overall costs of owning a database. DB2 includes data warehouse capabilities, high
availability function, and is administered remotely from a satellite control database.
The IBM Power 775 Supercomputer cluster solution requires a database to store all of the
configuration and monitoring data. DB2 Workgroup Server Edition 9.7 for HPC V1.1 is
licensed for use only on the executive management server (EMS) of the Power 775
high-performance computing cluster.
The EMS serves as a single point of control for cluster management of the Power 775 cluster.
The Power 775 cluster also includes a backup EMS, service nodes, compute nodes, I/O
nodes, and login nodes. DB2 Workgroup Server Edition 9.7 for HPC V1.1 must be installed
on the EMS and backup EMS.
1.9.3 Extreme Cluster Administration Toolkit
Extreme Cloud Administration Toolkit (xCAT) is an open source, scalable distributed
computing management, and provisioning tool that provides a unified interface for hardware
control, discovery, and operating system stateful and stateless deployment. This robust toolkit
is used for the deployment and administration of AIX or Linux clusters, as shown in
Figure 1-56 on page 73.
72
IBM Power Systems 775 for AIX and Linux HPC Solution
(vport). This handle is used by clients to send synchronous commands to the
virtual port
Advertisement
Table of Contents
