Cluster Edition fails to start

A number of problems can cause the Cluster Edition to not start.

The Cluster Edition creates a resource fault if VCS cannot start the Cluster Edition or detects that it is already running. Use the hagrp -clear command to clear the resource fault. This example clears the resource faults for an VCS group named “asegrp1”:

hagrp -clear asegrp1

VCS uses the ps command to locate the dataserver process: do not rename the dataserver binary.

Validate the Home attribute for the Cluster Edition resource using hares -display. Home must point to $SYBASE and the dataserver binary must be located in $SYBASE/ASE-15_0/bin. If Home does not point to the correct value, change the value using the hares -modify command.

VCS uses the RUN_server_name file to start the instances. If the RUN_server_name files do not exist, VCS uses the parameters defined in the Home, Server, and Quorum_dev variables to build the execution string that starts the instance:

HOME/ASE-15_0/bin/dataserver -sserver_name -QQuorum_dev

If you receive a resource fault, you should verify that: