Additional documents of interest
- Successful Business Continuity - Part 1 - Users and Groups
-
This article was published in the April 2005 issue of
AIX Update magazine
and discusses system administration needs and requirements oriented
around users and groups. The overall emphasis of this series of
articles is for implementation of enterprise wide unique identifiers
for a variety of parameters, such as user names, group names, UID and
GID numbers.
- Successful Business Continuity - Part 2 - Machine and Host Names
-
This article was published in the May 2005 issue of
AIX Update magazine
and discusses naming structures for machines, systems, adapters, and
aliases. The overall emphasis of this series of articles is for
implementation of enterprise wide unique identifiers for a variety of
parameters.
- Successful Business Continuity - Part 3 - Volume Names
-
This article was published in the December 2005 issue of
AIX Update magazine
and discusses naming structures for volume groups, logical volumes, log
logical volumes, directory mount points, etc. The overall emphasis of
this series of articles is for implementation of enterprise wide unique
identifiers for a variety of parameters.
- Successful Business Continuity - Part 4 - MQ Series, Startup/Shutdown Scripts, Error Processing
-
This article was published in the April 2006 issue of
AIX Update magazine
and discusses how to implement AIX in an environment dedicated to
business continuity. The topic of this article is the assignment of MQ
Series queue names and aliases, resource group startup and shutdown
script names (Application startup/shutdown script names), error logging,
and error notification.
- Successful Business Continuity - Part 5 - Miscellaneous topics
-
This article was published in the August 2006 issue of
AIX Update magazine
and discusses how to implement AIX in an environment dedicated to
business continuity. A variety of topics is discussed in this article
including automated documentation generation and management.
- Automated Microcode Management System
-
One of the most difficult administration tasks in an AIX environment is
attempting to keep the firmware and microcode up-to-date. Mt Xia has
devised an automated method of gathering the Microcode information,
determining which microcode needs to be updated, generating reports, and
uploading the required microcode updates to each individual system.
- Calculating the size of a Virtual Processor
-
This document describes the algorithms used to calculate the size of a
virtual processor when using shared processors in an LPAR. The IBM
documentation describes how to calculate CPU utilization, NOT how to
size for configuration, this document clarifies this process. A
description of the HMC input fields for the processor tab is included.
- Basics of Partition Load Manager Setup
-
This presentation was provided by Ron Barker from IBM regarding the PLM Basic
setup.
- ppt
- pdf
|
Logical Partition Standards
LPAR, short for logical partitioning, is a mechanism of taking a
computer's total resources - processors, memory and storage -- and
splitting them into smaller units that each can be run with its own
instance of the operating system and applications. Each partition can
communicate with the other partitions as if the other partition is in a
separate machine.
In Mt Xia's environment, the ability to obtain outages for the purpose
of maintenance and upgrades will be difficult. Furthermore, systems
supporting multiple business functions will be even more difficult to
obtain outage windows. Therefore it is desirable to create LPAR's to
support each business function, thus reducing the impact of an outage
upon the overall environment. So rather than creating large LPAR's
supporting multiple business functions for a client, it is preferable to
create multiple LPARs to support each business function.
When creating an LPAR, the following standards will be applied:
- The LPAR name and profile name will be the same as the short
machine name assigned to the LPAR.
- Minimum Memory: 512 MB
- Desired Memory: 512 MB
- Maximum Memory: All available memory
For LPARs that will require 3 physical processors or less during
normal operations:
- Processor Mode: shared
- Minimum processing units: 0.20
- Desired processing units: 0.20
- Maximum processing units: All available CPUs
- Minimum Virtual Processors: 2
- Desired Virtual Processors: 2
- Maximum Virtual Processors: 30
For LPARs that will require more than 3 physical processors during
normal operation:
- Processor Mode: shared
- Minimum processing units: 1.00
- Desired processing units: 1.00
- Maximum processing units: All available CPUs
- Minimum Virtual Processors: 2
- Desired Virtual Processors: 2
- Maximum Virtual Processors: 64
Physical I/O will be assigned as required by the business functions
supported by by each LPAR, Virtual I/O will be assigned as required by
each LPAR and in accordance with Mt Xia's VIO
standards. Connection monitoring will be enabled for each LPAR and
no LPARs will be started automatically when the frame is powered on.
|
|
|
