Jump to content United States-English
HP.com Home Products and Services Support and Drivers Solutions How to Buy
» Contact HP
More options
HP.com home
Using the Event Monitoring Service

Chapter 5 Monitoring ServiceGuard Package Dependencies

» 

Technical documentation

Complete book in PDF
» Feedback
Content starts here

 » Table of Contents

 » Glossary

 » Index

This chapter describes how to use SAM to define package dependencies on EMS resources. ServiceGuard by itself automatically monitors specific resources. Using ServiceGuard with EMS adds to the list of resources that can be monitored. These resources need to be configured and identified to ServiceGuard as package resource dependencies.

You create a monitoring request to observe the EMS resource and to notify ServiceGuard when that resource reaches a critical user-defined level. At that time ServiceGuard will fail over the package. The following are some examples of how EMS might be used:

  • In a cluster where one copy of data is shared between two nodes (both configured with EMS), you may want to fail a package over when, for example, the LAN or SCSI host adapter fails on the node running the package. ServiceGuard compares the resource UP values on other configured nodes, and fails the package over to the node that has the correct resources available.

  • In a cluster where each node has its own copy of data, you may want to failover a package to another node for any number of reasons:

    • host adapter, bus, controller, or disk failure

    • unprotected data (the number of copies is reduced to one)

    • degraded performance because one of the PV links has failed

This information for creating requests is also valid for EMS sold with other products (ATM, OTS, HyperFabric, or STM hardware monitors, for example) and for user-written monitors written according to developer specifications in Writing Monitors for the Event Monitoring Service (EMS) (HP Part Number B7611-90016).

NOTE: Create the same requests on all nodes configured for an ServiceGuard package.

A package can depend on any resource whose monitor is registered with EMS. To create package dependencies:

  1. Halt the cluster. Include a force option to stop all packages, by typing:

    cmhaltcl -f

    To add an EMS resource, the ServiceGuard cluster must be down. You can modify existing EMS resources through ServiceGuard while the cluster is running.

  2. From your command line, start SAM, by typing:

    sam

  3. Double-click the Clusters icon.

  4. Double-click the High Availability Clusters icon.

  5. Double-click on the Package Configuration icon.

    The High Availability Clusters screen shows all requests currently configured on that system. If there are no requests currently configured, the field area is empty. Figure 5-1 “High Availability Clusters Screen” shows the High Availability Clusters screen with the Actions menu displayed.

    Figure 5-1 High Availability Clusters Screen

    High Availability Clusters Screen
  6. From the Actions menu, select either the Create/Add a Package or Modify Package Configuration option.

    If you select Create/Add a Package, a screen similar to Figure 5-2 “Create/Add a Package Screen”, displays.

    Figure 5-2 Create/Add a Package Screen

    Create/Add a Package Screen
  7. If you have not previously done so, click Specify Package Name and Node and Specify Package SUBNET Address. Then click on Specify Package Resource Dependencies... to add EMS resources as package dependencies. A screen similar to Figure 5-3 “Package Resource Dependencies Screen” displays.

    Figure 5-3 Package Resource Dependencies Screen

    Package Resource Dependencies Screen
  8. The Resources: field lists all the installed resources discovered by ServiceGuard. To make a package dependent on an EMS resource, select it from the list, then click Add Resource.... An Add Resources screen, similar to Figure 5-4 “Add Resources Screen” displays.

    Figure 5-4 Add Resources Screen

    Add Resources Screen
  9. Click through the Resource Classes and Resource Names to select the entity you wish to monitor. Click OK. A Resource Parameters screen, similar to Figure 5-5 “Resource Parameters Screen” displays.

    The example in Figure 5-5 “Resource Parameters Screen” shows the possible values for pv_summary. Different resources show different available UP values.

    Figure 5-5 Resource Parameters Screen

    Resource Parameters Screen
    NOTE: Make sure you always select UP in the Resources Up Values field. ServiceGuard creates an EMS request that sends an event if the Resources Up Value field is not equal to the UP value.

    If you select only UP, the package fails over if the value is anything but UP. In this example, if you select UP and PVG-UP, the package fails over if the pv_summary value is not equal to UP or PVG_UP; in other words, if pv_summary is SUSPECT or DOWN.

    The polling interval determines the maximum amount of elapsed time before the monitor knows about a change in resource status. For critical resources, you may want to set a short polling interval, such as 30 seconds, but this could adversely affect system performance. With longer polling intervals you gain system performance, but you risk not detecting problems soon enough.

    You can also add resources as package dependencies by modifying the package configuration file. The default filename is /etc/cmcluster/pkg_name.ascii. See Managing MC/ServiceGuard for details on how to modify this file. For example, the syntax might be:

    RESOURCE_NAME               /vg/vg01/pv_summary
    RESOURCE_POLLING_INTERVAL   60
    RESOURCE_UP_VALUE           = UP
    RESOURCE_UP_VALUE           = PVG_UP    
    

    ServiceGuard automatically distributes this information to every node in the cluster when the package configuration file is applied to the cluster configuration (see the cmapplyconf command).

Printable version
Privacy statement Using this site means you accept its terms Feedback to webmaster
© 1999 Hewlett-Packard Development Company, L.P.