Size an ARP cache across VLANs and estimate the broadcast load its refresh cycle puts on every host.
Every host in a VLAN must process every ARP broadcast in that VLAN, so the load each host sees depends on the size of its own broadcast domain rather than the whole device. The refresh rate is set by the aging timeout — a shorter timeout means fresher entries and more broadcast. Aggregation switches terminating hundreds of VLANs hit the ARP table limit long before the MAC table, and once entries are punted to software the CPU becomes the bottleneck for traffic that should be hardware-forwarded.
ARP Table
Entries = hosts per VLAN × VLANs; refreshes per second = entries ÷ aging timeout, each one a 28-byte ARP request padded to a 64-byte frame.
Broadcast bandwidth
ARP broadcast load = (entries ÷ timeout) × 84 × 8 bits per second.
Entries = hosts per VLAN × VLANs; refreshes per second = entries ÷ aging timeout, each one a 28-byte ARP request padded to a 64-byte frame. Every host in a VLAN must process every ARP broadcast in that VLAN, so the load each host sees depends on the size of its own broadcast domain rather than the whole device. The refresh rate is set by the aging timeout — a shorter timeout means fresher entries and more broadcast.
Aggregation switches terminating hundreds of VLANs hit the ARP table limit long before the MAC table, and once entries are punted to software the CPU becomes the bottleneck for traffic that should be hardware-forwarded.
This calculator takes 4 inputs: Hosts per VLAN, VLANs terminated on the device, ARP aging timeout, ARP table limit. The pre-filled defaults are a realistic starting point — replace them with figures from your own environment for a result you can act on.