Front cover -- Contents -- Notices -- Trademarks -- Preface -- The team that wrote this redbook -- Become a published author -- Comments welcome -- Part 1 Technical information -- Chapter 1. Introduction -- 1.1 Historical background -- 1.1.1 HPSSDL -- 1.1.2 RS/6000® -- 1.1.3 SP1® and early switch implementation -- 1.1.4 PSSP and SP2® -- 1.1.5 The SP Switch -- 1.1.6 PCI and Enterprise Servers -- 1.1.7 The SP Switch 2 -- 1.1.8 pSeries and Cluster 1600 -- 1.1.9 Regatta -- 1.1.10 Cluster Systems Manager -- 1.2 The pSeries High Performance Switch -- 1.2.1 Naming the technology -- 1.2.2 Progression of switch technology -- 1.2.3 Major HPS components -- 1.2.4 IBM Eserver pSeries High Performance Switch (HPS) features -- Chapter 2. Technical overview of the IBM Eserver pSeries High Performance Switch (HPS) -- 2.1 Hardware components -- 2.1.1 Systems -- 2.1.2 Supported frames -- 2.1.3 Switch overview -- 2.1.4 Switch Network Interfaces -- 2.1.5 p655 Switch Network Interface -- 2.1.6 p690 Switch Network Interfaces -- 2.2 Switch connectivity -- 2.2.1 Administrative networking -- 2.2.2 Serial cables -- 2.2.3 Switch cables -- 2.3 Firmware -- 2.3.1 System firmware -- 2.3.2 Frame microcode -- 2.3.3 Hypervisor -- 2.4 Hardware Management Console (HMC) -- 2.4.1 HMC required components -- 2.4.2 WebSM (optional) remote management -- 2.4.3 LPAR definitions -- 2.4.4 Affinity LPARs -- 2.4.5 Hardware Server -- 2.5 Switch Network Manager -- 2.5.1 SNM major components -- 2.5.2 HMC SNM filesets -- 2.5.3 Switch Manager GUI -- 2.5.4 Switch Network Management CLI -- 2.5.5 Switch network manager (fnmd) daemon -- 2.6 FNM communication -- 2.6.1 Time of Day (TOD) -- 2.6.2 Switch data structures -- 2.6.3 FNM message data types -- 2.6.4 CEC states that CSP makes visible to FNM -- 2.6.5 FNM message types -- 2.6.6 FNM_Init -- 2.6.7 FNM_Route -- 2.6.8 FNM_Recovery -- 2.6.9 FNM_Diag.

2.7 Operating systems -- 2.7.1 Devices -- 2.7.2 Common issues and diagnosis -- 2.7.3 AIX tunables -- 2.7.4 Cluster Systems Manager -- 2.7.5 Additional applications -- 2.8 References -- Chapter 3. Application considerations -- 3.1 RSCT Resource Monitoring and Control -- 3.1.1 What is RSCT? -- 3.1.2 RSCT components -- 3.1.3 Resource Management and Control (RMC) subsystem -- 3.1.4 Security -- 3.1.5 Cluster environment -- 3.2 Cluster System Management -- 3.2.1 Supported platforms -- 3.2.2 Client install mechanisms -- 3.2.3 CSM 1.3.2.1 November 2003 -- 3.3 HACMP -- 3.3.1 HACMP overview -- 3.3.2 Role of HACMP -- 3.3.3 Physical components of an HACMP cluster -- 3.4 IBM Virtual Shared Disks -- 3.4.1 Overview of the Virtual Shared Disk components -- 3.4.2 Product packaging -- 3.4.3 Configuration repository -- 3.4.4 Shared external disk access -- 3.4.5 Quorum -- 3.4.6 IBM Recoverable Virtual Shared Disk overview -- 3.4.7 Recovery for concurrent Virtual Shared Disks -- 3.4.8 Disk cable and disk adapter failures -- 3.4.9 Communication adapter failures -- 3.4.10 Technical large pages overview -- 3.4.11 Restrictions for using Virtual Shared Disks -- 3.4.12 Overview of the subsystem device driver (SDD) -- 3.4.13 VSD configuration support for the SDD -- 3.4.14 HPS considerations when using the IP protocol for data transmissions -- 3.5 General Parallel File System 2.1 -- 3.5.1 GPFS overview -- 3.5.2 GPFS 2.1 (APAR IY47306) new features -- 3.5.3 Benefits of using GPFS -- 3.5.4 The basic GPFS structure -- 3.5.5 GPFS RPD cluster -- 3.5.6 GPFS HACMP cluster -- 3.5.7 Recoverability considerations -- 3.5.8 GPFS quorum considerations -- 3.6 Engineering applications -- 3.6.1 Low-level application programming interface (LAPI) -- 3.6.2 LoadLeveler -- 3.6.3 Message passing interface -- 3.6.4 Parallel Environment.

3.6.5 Parallel Engineering and Scientific Subroutine Library (ESSL) -- 3.6.6 Further information -- 3.7 Databases -- 3.8 Documentation -- Part 2 Practical implementation examples -- Chapter 4. Installation, configuration and administration of pSeries HPS -- 4.1 Planning for the High Performance Switch -- 4.1.1 Hardware planning -- 4.1.2 Software planning -- 4.1.3 A sample test environment -- 4.2 Installation and configuration of pSeries HPS -- 4.2.1 HMC installation -- 4.2.2 Firmware upgrade -- 4.2.3 Frame configuration -- 4.2.4 LPAR configuration -- 4.2.5 Starting the switch -- 4.2.6 Configuring CSM and NIM -- 4.2.7 Installing and configuring the SNIs in AIX -- 4.2.8 Verifying the switch connections -- 4.2.9 Switch tuning -- 4.3 Switch administration -- 4.3.1 Power commands -- 4.3.2 Query commands -- 4.3.3 Diagnostic operations -- 4.4 References -- Chapter 5. Clustering applications scenarios -- 5.1 Testing the switch adapter failover -- 5.1.1 Scenario description -- 5.1.2 Logical interface down and detach scenario -- 5.1.3 Logical interface detach -- 5.1.4 Cable disconnected scenario -- 5.2 VSD implementation on an HPS -- 5.2.1 Scenario description -- 5.2.2 Our environment -- 5.2.3 VSD installation -- 5.2.4 Creating the RSCT peer domain -- 5.2.5 Designating the VSD nodes -- 5.2.6 Creating the VSD -- 5.2.7 VSD verification -- 5.2.8 VSD failover -- 5.3 GPFS on VSD -- 5.3.1 Our environment -- 5.3.2 GPFS software installation and configuration -- 5.3.3 GPFS verification -- 5.3.4 GPFS failover scenario -- Part 3 Appendixes -- Appendix A. Quick references -- A.1 AIX -- A.1.1 NIM resource list -- A.1.2 Peer domain management -- A.2 HMC procedures -- A.2.1 HMC quick network configuration -- A.2.2 Service focal point failures -- A.2.3 Service Agent rebuild -- A.2.4 Upgrading the frame microcode using the instfru command.

A.2.5 Output example of /usr/local/hsctool/hps_check.pl script -- Appendix B. Troubleshooting -- B.1 Symptoms of common problems -- B.2 Typical switch issues -- B.3 Tracing -- B.3.1 Common tracing errors -- B.3.2 After fixing the problem -- B.3.3 RSCT related -- B.4 Config files -- B.4.1 Switch daemon -- B.4.2 hdwr_svr config files -- B.5 Log files -- B.6 Commands -- Abbreviations and acronyms -- Glossary -- Related publications -- IBM Redbooks -- Other publications -- Online resources -- How to get IBM Redbooks -- Help from IBM -- Index -- Back cover.