Engineering inputs
Set the engineering assumptions and design parameters.
Planning inputs
Engineering results
Review calculated capacity, limits and design checks.
UPLINK CAPACITY TOOL
Size switch uplinks from access ports, utilization, concurrency and redundancy.
Set the engineering assumptions and design parameters.
Review calculated capacity, limits and design checks.
Access line rate is active ports multiplied by port speed. Estimated busy-hour demand applies average utilization, concurrent active-port percentage and protocol or burst allowance. Target usable capacity is the sum of active forwarding uplinks multiplied by the target utilization.
The line-rate oversubscription ratio compares all entered access ports at full speed with active uplink capacity. Estimated uplink utilization compares modeled busy-hour demand with that capacity. A high line-rate ratio can still be workable when measured demand is low, but bursts and correlated traffic can create short congestion.
For office, school, Wi-Fi AP and IoT access, use measured busy-hour traffic and allow for growth. Server, storage, backup, imaging and east-west workloads may be more synchronized and bursty, so they often need a lower sharing ratio, larger safety margin and closer queue monitoring.
Count only uplinks that actively forward in the scenario. First model normal operation, then repeat with the links that survive a cable, optic, member switch or line-card failure. Confirm that the remaining topology still meets the target utilization and that traffic can actually use those links.
IEEE 802.1AX defines link aggregation so parallel links can act as one logical, resilient load-sharing interconnect. RFC 7424 explains that hash-based distribution depends on flow mix and may not balance component links evenly.
The Cisco Campus LAN and Wireless LAN Design Guide calculates oversubscription from actual access-port speeds and active uplinks, and recommends checking both normal operation and switch or link failure scenarios.
There is no universal ratio. Select it from measured busy-hour demand, application sensitivity, growth, buffer and queue behavior, and the failure scenario. Campus user access often tolerates more sharing than storage, backup or server traffic.
Usually not. A link aggregation group can provide 20 Gbps aggregate capacity, but hash-based distribution normally keeps one flow on one component link. Flow mix and hash balance determine how evenly the bundle is used.
No. Enter only links that forward traffic in the scenario being evaluated. If a standby link activates only after failure, calculate normal operation and the surviving failure topology separately.
The line-rate ratio compares every access port at full speed with total active uplink capacity. Busy-hour demand also applies measured utilization, concurrency and burst allowance, so the two metrics answer different planning questions.
No. Validate the estimate with interface counters, percentile and peak data, packet drops, queue statistics and application requirements. QoS, buffers, flow hashing, hardware limits and growth remain separate design checks.
ENGINEERING WORKFLOW
Plan addressing, VLAN capacity, packet size, uplinks, optics, access policy and DNS behavior as one LAN design sequence.