A simple work-a-round to this is to just set OnlineTimeout at type level, rather than resource level. This will mean any other SRDF resources which do not NEED a large timeout will have a timeout longer than necesary, but this doesn't really matter. Normally the downside of setting online timeout to large is that if the resource is not going to come online, you are delaying trying another node, but since SRDF commands are run on array, not VCS node, there is little point trying another node, unless VCS node can't reach gateway device on original node.
Agree, code needs fixing.
Mike
