Linux clustering with CSM and GPFS / [Stephen Hochstetler, Bob Beringer].

Format:

Book

Contributor:

Hochstetler, Stephen.

Beringer, Bob.

International Business Machines Corporation. International Technical Support Organization.

Series:

IBM redbooks.

IBM redbooks

Language:

English

Subjects (All):

Parallel computers.

Computer network architectures.

File organization (Computer science).

Linux.

IBM Cluster systems management.

Physical Description:

xxii, 316 p. : ill.

Edition:

3rd ed.

Place of Publication:

Austin, TX : IBM, International Technical Support Organization, 2004.

Language Note:

English

Contents:

Front cover

Contents

Figures

Tables

Notices

Trademarks

Preface

The team that wrote this redbook

Become a published author

Comments welcome

Summary of changes

December 2003, Third Edition

Part 1 Fundamentals

Chapter 1. Clustering concepts and general overview

1.1 What is a cluster

1.2 Cluster types

1.2.1 High availability

1.2.2 High performance computing

1.2.3 Horizontal scaling

1.3 Beowulf clusters

1.4 Linux, open source, and clusters

1.5 IBM Linux clusters

1.5.1 xSeries custom-order clusters

1.5.2 The IBM eServer Cluster 1350

1.6 Cluster logical structure

1.6.1 Cluster node types and xSeries offerings

1.7 Other cluster hardware components

1.7.1 Networks

1.7.2 Storage

1.7.3 Terminal servers

1.7.4 Keyboard, video, and mouse switches

1.8 Cluster software

Chapter 2. New Linux cluster offering from IBM: Cluster 1350

2.1 Product overview

2.2 Hardware

2.2.1 Racks

2.2.2 Cluster nodes

2.2.3 Remote Supervisor Adapters

2.2.4 External storage

2.2.5 Networking

2.2.6 Terminal servers

2.2.7 Hardware console: Keyboard, video, and mouse (KVM)

2.3 Software

2.3.1 Linux operating system

2.3.2 IBM Cluster Systems Management (CSM) for Linux

2.3.3 General Parallel File System for Linux

2.3.4 Other software considerations

2.4 Services

2.4.1 Installation planning services

2.4.2 On-site installation of the IBM eServer Cluster 1350

2.4.3 Warranty service and support

2.4.4 Project support services

2.4.5 Installation and customization

2.4.6 Continuing support services

2.5 Summary

Chapter 3. Introducing Cluster Systems Management for Linux

3.1 IBM Cluster Systems Management overview

3.2 CSM architecture

3.2.1 Resource Monitoring and Control subsystem

3.2.2 CSM components

3.2.3 Security in CSM.

3.3 CSM monitoring

3.3.1 How CSM monitors a system

3.3.2 Resource Managers

3.3.3 Predefined conditions

3.3.4 Responses

3.3.5 Associating conditions and responses

3.3.6 Creating new conditions and responses

3.4 CSM management components

3.4.1 Node and group management commands

3.4.2 Controlling the hardware

3.4.3 Using DSH to run commands remotely

3.4.4 Configuration File Manager (CFM)

3.5 CSM hardware requirements

3.5.1 Minimum hardware requirements

3.6 Software requirements to run CSM

3.6.1 IBM CSM software packages

3.6.2 Third party software components

3.7 Quick installation process overview

3.8 CSM futures

3.9 Summary

Chapter 4. Introducing General Parallel File System for Linux

4.1 Introduction to GPFS

4.1.1 GPFS terms and definitions

4.1.2 What is new in GPFS for Linux Version 1.3

4.1.3 GPFS advantages

4.2 GPFS architecture

4.2.1 GPFS components

4.2.2 GPFS Network Shared Disk considerations

4.2.3 GPFS global management functions

4.2.4 Disk storage used in GPFS

4.2.5 Data and metadata replication capability

4.2.6 GPFS and applications

4.2.7 Scenario and operation example

4.3 GPFS requirements

4.3.1 Hardware requirements

4.3.2 Software requirements

4.4 Summary

Part 2 Implementation and administration

Chapter 5. Cluster installation and configuration with CSM

5.1 Planning the installation

5.1.1 Before you begin

5.1.2 Develop a network plan

5.1.3 Develop a hardware resources plan

5.1.4 Develop a plan to update your hardware

5.1.5 Develop your security plan

5.1.6 Installation media

5.1.7 Documenting the cluster configuration

5.2 Configuring the management server

5.2.1 Red Hat Linux 7.3 installation

5.2.2 Install additional Red Hat Linux 7.3 packages

5.2.3 Install Red Hat Linux 7.3 updates.

5.2.4 NTP configuration

5.2.5 Fix syslogd

5.2.6 Domain Name System (DNS) configuration

5.2.7 Install Terminal Server

5.2.8 System Management hardware configuration

5.2.9 Configuring environment variables

5.2.10 Deciding which remote shell protocol to use

5.2.11 Installing the CSM core package

5.2.12 Running the CSM installms script

5.2.13 Install the license

5.2.14 Verify the CSM installation on the management node

5.3 CSM installation on compute and storage nodes

5.3.1 BIOS settings for compute and storage nodes

5.3.2 Preparing to run the definenode command

5.3.3 Running the definenode script

5.3.4 Verify that rpower works

5.3.5 Customize the KickStart template (optional)

5.3.6 Running the csmsetupks script

5.3.7 Running the installnode script

5.3.8 Verifying compute and storage node installation

5.3.9 Configuring NTP on your compute and storage nodes

5.4 Special considerations for storage node installation

5.5 Summary

Chapter 6. Cluster management with CSM

6.1 Changing the nodes in your cluster

6.1.1 Replacing nodes

6.1.2 Adding new nodes using the full installation process

6.1.3 Adding new nodes using the CSM only installation process

6.1.4 Removing nodes

6.1.5 Changing host names of nodes

6.2 Remote controlling nodes

6.2.1 Power control

6.2.2 Console access

6.2.3 Node availability monitor

6.2.4 Hardware status and management

6.3 Node groups

6.4 Running commands on the nodes

6.4.1 Distributed shell (dsh)

6.4.2 Distributed command execution manager (DCEM)

6.5 Configuration File Manager (CFM)

6.6 Software maintenance system (SMS)

6.7 Event monitoring

6.7.1 RMC components

6.7.2 Activating condition responses

6.7.3 Deactivating condition responses

6.7.4 Creating your own conditions and responses

6.7.5 RMC audit log.

6.8 Backing up CSM

6.9 Uninstalling CSM

Chapter 7. GPFS installation and configuration

7.1 Basic steps to install GPFS

7.2 GPFS planning

7.2.1 Network implementation

7.2.2 Documentation

7.3 Preparing the environment

7.3.1 Nodes preparation

7.3.2 Prerequisite software

7.3.3 Prepare kernel source file for GPFS and Myrinet adapter

7.3.4 Time synchronization

7.3.5 Setting the remote command environment

7.3.6 Myrinet adapter installation

7.3.7 Prepare external storage for GPFS

7.3.8 Setting PATH for the GPFS command

7.4 GPFS installation

7.4.1 Installing the source files

7.4.2 Building the GPFS open source portability layer

7.5 Creating the GPFS cluster

7.5.1 Creating the GPFS nodes descriptor file

7.5.2 Defining the GPFS cluster

7.6 Creating the GPFS nodeset

7.7 Starting GPFS

7.8 Disk definitions

7.8.1 GPFS nodeset with NSD network attached servers

7.8.2 GPFS nodeset with direct attached disks

7.9 Exporting a GPFS file system using NFS

7.10 GPFS shutdown

7.11 Summary

Chapter 8. Managing the GPFS cluster

8.1 Adding and removing disks from GPFS

8.1.1 Adding a new disk to an existing GPFS file system

8.1.2 Deleting a disk in an active GPFS file system

8.1.3 Replacing a failing disk in an existing GPFS file system

8.2 Removing all GPFS file systems and configuration

8.3 Access Control Lists (ACLs)

8.4 GPFS logs and traces

8.4.1 GPFS logs

8.4.2 Trace facility

8.5 Troubleshooting: Some possible GPFS problems

8.5.1 Authorization problems

8.5.2 Connectivity problems

8.5.3 NSD disk problems

8.6 Gather information before contacting Support Center

Chapter 9. Migrating xCat clusters to CSM

9.1 xCAT overview

9.2 Migrating xCAT clusters to CSM

9.2.1 Using xcat2csm

9.2.2 Edit the generated files.

9.2.3 Importing the files into CSM

9.3 xCAT and CSM co-existence

Part 3 Appendixes

Appendix A. SRC and RSCT

SRC and RSCT components overview

System Resource Controller (SRC)

Subsystem components

Reliable Scalable Cluster Technology (RSCT)

Topology Services subsystem

Group Services (GS) subsystem

Appendix B. Common facilities

DNS server

Package description

DNS installation

DNS configuration

Starting the DNS server

Testing the DNS server

BIND logging

Other features

OpenSSH

OpenSSH authentication methods

Update the file /etc/hosts

Key generation for the root user

Generation of authorized_keys file

Distribution of the authorized_keys file to the other nodes

Ensuring all nodes know each other

Verification of the SSH configuration

Additional information and trouble shooting

Appendix C. Migrating to GPFS 1.3 from earlier versions

Migration steps

Appendix D. Planning worksheets

CSM planning worksheets

Management node TCP/IP attributes worksheets

Compute node TCP/IP attributes worksheet

Node attributes worksheets

GPFS Planning worksheets

File system descriptions

Network File Shared descriptions

Glossary

Abbreviations and acronyms

Related publications

IBM Redbooks

Other publications

Online resources

How to get IBM Redbooks

Help from IBM

Index

Back cover.

Notes:

"January 2004."

"SG24-6601-01."

Includes bibliographical references (p. 307-310) and index.

OCLC:

939263508