PoE and Voice VLAN Design for a New Distribution Center
Northline Logistics is fitting out a 180,000 sq ft distribution center with three IDF closets. Each IDF will home-run to one 48-port PoE access switch with a 740 W PoE budget and dual power supplies. The device count per IDF is fixed: 16 Mist APs at 25.5 W each, 22 IP cameras at 12.9 W each, 40 desk phones at 6.5 W each, and 24 desktop PCs. Cameras must be on VLAN 30, phones on voice VLAN 20, and PCs on VLAN 10. Phones support LLDP-MED and have a PC pass-through port. The customer insists that a single power-supply failure must not cause phone or AP reboots, and that normal PoE utilization must stay at or below 80 percent of budget. The cabling contractor has already installed one switch per IDF and wants to know whether that is sufficient before the equipment is powered. The network team also needs a port configuration standard for the phone/PC wall outlets.
1. Calculate the PoE load per IDF and determine whether one 740 W switch per IDF is acceptable under the 80 percent rule and the single-PSU-failure requirement. Show the math and state the minimum switch count.
Read the worked answer
Per IDF: APs = 16 x 25.5 W = 408 W; cameras = 22 x 12.9 W = 283.8 W; phones = 40 x 6.5 W = 260 W. Total PoE load = 951.8 W. One 740 W switch cannot carry the normal load at all, and 80 percent of 740 W is 592 W, so it fails the utilization rule by a wide margin. Port count also rules it out: 16 + 22 + 40 + 24 = 102 access ports before uplinks, so one 48-port switch is impossible and two 48-port switches provide only 96 ports. The minimum is three 48-port PoE switches per IDF. With three switches, average PoE is about 317 W per switch, leaving room below the 592 W threshold. Distribute APs, cameras, and phones across all three switches so a single switch or PSU failure does not take out an entire device class. Confirm the switch PoE budget with one PSU failed; if that reduced budget is below the connected load, use higher-capacity power supplies or redistribute devices. Set PoE priority high for APs and phones so cameras are shed first under abnormal conditions.
2. Write the wall-outlet port standard for a phone with a PC behind it. Include VLANs, LLDP-MED, PoE, and 802.1X implications. Explain why the data VLAN should not be the voice VLAN.
Read the worked answer
Configure the wall port as an access port with a voice VLAN: data VLAN 10 untagged toward the PC and voice VLAN 20 tagged by the phone. On Junos ELS, create VLANs DATA (10) and VOICE (20), enable LLDP-MED on the interface, set the VoIP VLAN for the interface, and give the port PoE priority high. LLDP-MED tells the phone which VLAN to tag and what power it may draw; without it, the phone may send untagged voice or use the wrong VLAN. The phone passes untagged data traffic to the PC while tagging voice traffic itself, so the switch must carry both. Do not put the PC in the voice VLAN: voice should be an isolated Layer 2 segment with its own QoS and security policy, and mixing user data into it increases broadcast scope and weakens policy control. For 802.1X, plan for two identities on the port: the phone and the PC. Use multiple supplicants or MAB fallback for older phones, and have RADIUS return the voice VLAN for the phone and the data VLAN for the PC. If the PC fails authentication, it should fall to a restricted VLAN without dropping the phone's authenticated voice service.
3. Which devices should receive high PoE priority, and what operational check should be added after deployment?
Read the worked answer
Set high PoE priority on APs and phones, medium or low on cameras. APs carry the largest number of dependent users and should not be shed first; phones may carry clinical or safety calls in some environments and should also be protected. Cameras are important but can often tolerate a later shed decision, especially if local recording exists. After deployment, add a recurring check that per-switch PoE utilization stays below 80 percent and that no port reports PoE resets or power-deny events. In Mist, watch switch and port insights for PoE faults; on the CLI, check PoE interface status and power consumption. Also verify the reduced PoE budget after a single PSU failure, because a dual-PSU chassis may not present the same budget with one feed down. If the reduced budget cannot cover high-priority devices, add power or redistribute loads before go-live.