HSG80 ACS Solution Software V8.6 for IBM AIM Installation and Configuration Guide
Table Of Contents
- HSG80 ACS Solution Software Version 8.6 for IBM AIX
- Contents
- Figures
- Tables
- About this Guide
- Chapter 1
- Planning a Subsystem
- Chapter 2
- Planning Storage
- Where to Start
- Configuration Rules
- Device PTL Addressing Convention
- Determining Storage Requirements
- Choosing a Container Type
- Creating a Storageset Profile
- Storageset Planning Considerations
- Storageset Expansion Considerations
- Partition Planning Considerations
- Changing Characteristics through Switches
- Storageset and Partition Switches
- Initialization Switches
- Unit Switches
- Storage Maps
- Chapter 3
- Preparing the Host System
- Chapter 4
- Installing and Configuring the HS-Series Agent
- Why Use StorageWorks Command Console (SWCC)?
- Installation and Configuration Overview
- About the Network Connection for the Agent
- Before Installing the Agent
- Installing and Configuring the Agent
- Running the Agent
- Reconfiguring the Agent
- Uninstall Agent, Storageworks Software, and Cambex Driver.
- High Availability Cluster Multi-Processor (HACMP)
- Enabler Software
- General Note on Software Installation
- SWCC Clients
- Running the Agent
- HACMP Implementation for the Agent
- Prerequisites
- Implementation
- Installing the Agent on an HACMP Cluster
- Installing and Configuring the Client on an HACMP cluster
- Operation of the Agent on an HACMP Cluster
- Normal Operation of the Agent on an HACMP Cluster
- Starting, Stopping and Checking for the Agent
- Password and Notification
- Client Functions and Storage Subsystem Options
- Failover with Agent as Part of a Resource Group
- Client IP Address
- Event Scripts
- Notification within HACMP
- Node Isolation Recovery
- Failure of the Agent
- Polling all nodes in cluster for presence of a running Agent
- Restarting the Agent on surviving node
- Notification
- Supporting Tools
- HACMP for AIX Primary Event Scripts
- HACMP for AIX Secondary Event Scripts
- Chapter 5
- Configuration Procedures
- Establishing a Local Connection
- Setting Up a Single Controller
- Setting Up a Controller Pair
- Configuring Devices
- Configuring a Stripeset
- Configuring a Mirrorset
- Configuring a RAIDset
- Configuring a Striped Mirrorset
- Configuring a Single-Disk Unit (JBOD)
- Configuring a Partition
- Assigning Unit Numbers and Unit Qualifiers
- Configuration Options
- Chapter 6
- Verifying Storage Configuration from the Host
- Chapter 7
- Configuration Example Using CLI
- Chapter 8
- Backing Up the Subsystem, Cloning Data for Backup, and Moving Storagesets
- Appendix A
- Subsystem Profile Templates
- Storageset Profile
- Storage Map Template 1 for the BA370 Enclosure
- Storage Map Template 2 for the second BA370 Enclosure
- Storage Map Template 3 for the third BA370 Enclosure
- Storage Map Template 4 for the Model 4214R Disk Enclosure
- Storage Map Template 5 for the Model 4254 Disk Enclosure
- Storage Map Template 6 for the Model 4310R Disk Enclosure
- Storage Map Template 7 for the Model 4350R Disk Enclosure
- Storage Map Template 8 for the Model 4314R Disk Enclosure
- Storage Map Template 9 for the Model 4354R Disk Enclosure
- Appendix B
- Installing, Configuring, and Removing the Client
- Appendix C
- Exploitation of the CLONE and SNAPSHOT Functions of the HSG80 in an AIX Environment
- Glossary
- Index

Planning Storage 2–27
Large chunk sizes also tend to increase the performance of random reads and writes. It is
recommended that you use a chunk size of 10 to 20 times the average request size,
rounded to the closest prime number.
To calculate the chunk size that should be used for your subsystem, you must first analyze
the types of requests that are being made to the subsystem:
■ Many parallel I/Os that use a small area of disk should use a chunk size of 10 times the
average transfer request rate.
■ Random I/Os that are scattered over all the areas of the disks should use a chunk size
of 20 times the average transfer request rate.
■ If you do not know, then you should use a chunk size of 15 times the average transfer
request rate.
■ If you have mostly sequential reads or writes (like those needed to work with large
graphic files), make the chunk size for RAID 0 and RAID 0+1 a small number (for
example: 67 sectors). For RAID 5, make the chunk size a relatively large number (for
example: 253 sectors).
Table 2–2 shows a few examples of chunk size selection.
e
Increasing Sequential Data Transfer Performance
RAID 0 and RAID 0+1 sets intended for high data transfer rates should use a relatively
low chunk size (for example: 67 sectors). RAID 5 sets intended for high data rate
performance should use a relatively large number (for example: 253 sectors).
Table 2–2 Example Chunk Sizes
Transfer Size (KB) Small Area of I/O Transfers Unknown Random Areas of I/O Transfers
241 5979
4 79 113 163
8 157 239 317