Filter
Top Products
Cluster Customization 123
For example, a file system cannot be unmounted, because of a process
running on it. Then, you might want to kill that process first, before
unmounting the file system, in order to get the event script done. Now, since
the event script didn’t succeed in its first run, the
Retry
feature enables
HACMP for AIX to retry it until it finally succeeds, or the retry count is
reached.
5.1.5 Notes on Customizing Event Processing
You must declare a shell (for example #!/bin/sh) at the beginning of each
script executed by the notify, recovery, and pre- or post-event processing
commands.
Notify, recovery, and pre- and post-event processing do not occur when the
force option of the
node_down event is specified.
Synchronizing the cluster configuration does not propagate the actual new or
changed scripts; you must add these to each node manually. Also, it is
allowed to have different contents in these scripts on different nodes in order
to be able to act upon different environments. However, the name of these
scripts, their location in the file system, and their permission bits have to be
identical.
5.1.6 Event Emulator
To test the effect of running an event on your cluster, HACMP for AIX
provides a utility to run an emulation of an event. This emulation lets you
predict a cluster's reaction to an event as though the event actually occurred.
The emulation runs on all active nodes in your cluster, and the output is
stored in an output file. You can select the path and name of this output file
using the
EMU_OUTPUT environment variable, or, use the default
/tmp/emuhacmp.out file on the node that invoked the Event Emulator.
For more information on event emulation, see these chapters: “Administrative
Facilities” in the
HACMP for AIX, Version 4.3: Concepts and Facilities,
SC23-4276,
and “Monitoring an HACMP Cluster” in the
HACMP for AIX,
Version 4.3: Administration Guide,
SC23-4279.
5.2 Error Notification
The AIX Error Notification facility detects errors matching predefined
selection criteria and responds in a programmed way. The facility provides a
wide range of criteria that you can use to define an error condition. These
errors are called
notification objects.