Harry, Another way (and one that would use less bandwidth) would be to have both downstream LDM-s make identical requests of the upstream LDM and from each other. The upstream LDM would honor only one of the downstream requests at a time and if one of the downstream LDM-s failed, then the other would take over. I can't say which way is better for you so I guess I'd try both and look at the rtstats(1) results in order to determine which way to go. Regards, Steve Emmerson Ticket Details =================== Ticket ID: GSH-399207 Department: Support LDM Priority: Normal Status: Closed