Hello,

it seems, that there is a lacking feature from VCS preonline trigger implementation (using VCS 5.1SP1 on RHEL Linux x86_64).

How preonline trigger works in few lines:
Currently there are two possible parameter values FAULT and MANUAL.
If FAULT, then to get back control to VCS I must run:
hagrp -online -nopre $ARGV[1] -sys $ARGV[0] -checkpartial $ARGV[3]
If MANUAL, then to get back control to VCS I must run (no 4th parameter in this case):
hagrp -online -nopre $ARGV[1] -sys $ARGV[0]

Issue:
Suppose, that I have PARTIAL service group, I have preonline trigger and I want to switch it manually to another node.
hagrp -switch $SG -sys $SYS , trigger is invoked with parameter MANUAL, and suddenly PARTIAL service group tries to go ONLINE on $SYS instead of switching in PARTIAL state.

This situation doesn't occur if no preonline trigger is configured.

The fix would be to include system (as a forth parameter) where $SG is not OFFLINE currently when manually switching, so person writing preonline trigger could use hagrp -online -nopre -checkpartial $SYS_FROM_WHERE_WE_ARE_SWITCHING.

Maybe there is some way to get previous state and location in preonline script about which I don't know?

Thanks,
Paulius