HP-UX and Linux Serviceguard Cluster Manageability, June 2007

Figure 10. HP Operations Manager with Serviceguard Manager
Example
A two-node cluster in a main data center provides failover services for critical applications that are
accessed from various locations and for customer access through the Internet.
The current management model uses a consultant who is remote and connects to the systems over a
secure Internet connection. In the event of a failover, it is important to get notification to the consultant
so that time can be scheduled to correct the initial failure and restore to full state of redundancy.
The consultant will have remote access to monitor and manage systems and will schedule on-site
services within 72 hours of a failure, providing services are still available.
Management solution and objectives
Consultant can be on-site within two hours of complete system failure.
HP SIMmust be installed at main data center.
Serviceguard Manager must be installed on the same system as HP SIM (integration into HP SIM is
automatic).
SNMP will be enabled on all cluster member nodes.
EMS HA monitors will be configured for SNMP and e-mail.
EMS HA monitors will be configured to send e-mail event notification to the consultant’s e-mail
address.
Trap destinations will be set on each cluster member node to include the IP address.
To make HP SIM highly available, a Serviceguard package will be created for HP SIM.
14