Is there a general rule of thumb to apply when capacity planning a remote site?

Is there a general rule of thumb to apply when capacity planning a remote site?

I am looking for a general rule of thumb to apply when capacity planning a remote site. For example, if I have 10 workstations at a site, how much bandwidth in kb do I roughly need per workstation? Are we looking at 50 kb per workstation therefore at least a 512 kb link or 100 kb per workstation therefore at least a 1m link? This is assuming all services are remote: web, e-mail, file sharing, telnet apps etc..

    Requires Free Membership to View

    Login

    By submitting your registration information to SearchNetworking.com you agree to receive email communications from TechTarget and TechTarget partners. We encourage you to read our Privacy Policy which contains important disclosures about how we collect and use your registration and other information. If you reside outside of the United States, by submitting this registration information you consent to having your personal data transferred to and processed in the United States. Your use of SearchNetworking.com is governed by our Terms of Use. You may contact us at webmaster@TechTarget.com.

It is difficult to point to any specific "rule of thumb" as capacity planning is very dependent upon the type of work being done at the remote site. For example, a retail location might require small Point of Sale (PoS) transactions whereas an office may require the transfer of bandwidth intensive graphics, real-time traffic (like VoIP) or other applications with unique handling requirements.

I would recommend profiling an existing remote location to get a baseline of the type of traffic you are seeing, noting the volumes of data that are sent during different periods of the day. If no such site exists, look at the traffic generated by a subset of users in the headquarters location and make an "educated guess."

It is often useful to use a WAN emulator to simulate different link sizes, latencies, and packet loss. This will help you predict performance under varying conditions. In addition, WAN optimization appliances can help save bandwidth and mitigate latency, which can save on costs by increasing WAN capacity (and improve application performance in the process). As these devices are being designed into WAN networks more and more, it is probably a good idea to include this into the evaluation process.

This was first published in January 2007

Back to top