Precisely!
There are two primary purposes for Scalability Servers:
1. Divide the load. This is a simple numbers proposition. A single Scalability Server can effectively manage around 1000 Agents, no more than 1500.
2. Reducing bandwidth utilization. Critical for sites with multiple remote locations, especially those which have more than a handful of Agents. In this case, a Scalability Server at the remote location means the software package need only be transferred across the WAN once, then it is distributed to the Agents on the Local Network. If you have 10Mb or 100Mb WAN connections and you are distributing large packages to large numbers of Agents without a Scalability Server at the remote location, you will definitely flood the network. Each Scalability Server by default can support up to 50 simultaneous package installation jobs, so that is potentially 50 simultaneous package transfers across a WAN connection. For this use case you should definitely have a Scalability Server at each location, or at the very least as Rich said, you need to configure Neighbor Aware.
Steve McCormick, ITIL
CA Technologies
Principal Services Consultant
Stephen.McCormick@ca.com
<mailto:Stephen.McCormick@ca.com>