HP Serviceguard manual Planning and Documenting an HA Cluster

Adding and Deleting Relocatable IP Addresses	68
Load Sharing	69
Monitoring LAN Interfaces and Detecting Failure: Link Level	69
Local Switching	70
Switching Back to Primary LAN Interfaces after Local Switching	72
Remote Switching	73
Address Resolution Messages after Switching on the Same Subnet	73
Monitoring LAN Interfaces and Detecting Failure: IP Level	73
Reasons To Use IP Monitoring	74
How the IP Monitor Works	74
Failure and Recovery Detection Times	75
Constraints and Limitations	76
Reporting Link-Level and IP-Level Failures	76
Example 1: If Local Switching is Configured	76
Example 2: If There Is No Local Switching	77
Automatic Port Aggregation	77
VLAN Configurations	78
What is VLAN?	78
Support for HP-UX VLAN	78
Configuration Restrictions	79
Additional Heartbeat Requirements	79
Volume Managers for Data Storage	79
Types of Redundant Storage	79
About Device File Names (Device Special Files)	80
Examples of Mirrored Storage	80
Examples of Storage on Disk Arrays	82
Types of Volume Manager	84
HP-UX Logical Volume Manager (LVM)	84
Veritas Volume Manager (VxVM)	84
Propagation of Disk Groups in VxVM	84
Package Startup Time with VxVM	84
Veritas Cluster Volume Manager (CVM)	85
Cluster Startup Time with CVM	85
Propagation of Disk Groups with CVM	85
Redundant Heartbeat Subnets	86
Comparison of Volume Managers	86
Responses to Failures	88
System Reset When a Node Fails	88
What Happens when a Node Times Out	88
Example	88
Responses to Hardware Failures	89
Responses to Package and Service Failures	90
Responses to Package and Generic Resources Failures	90
Service Restarts	91
Network Communication Failure	91
4 Planning and Documenting an HA Cluster	92
General Planning	92
Serviceguard Memory Requirements	92
Planning for Expansion	92
Hardware Planning	93
SPU Information	94
Network Information	94
LAN Information	94
Setting SCSI Addresses for the Largest Expected Cluster Size	95

Contents 5