I've finished planning the main distribution frame and would appreciate a second set of eyes on the rack layout. I've worked around racks and IT closets for years, but this is my first time designing one from scratch. Most of the equipment has already been ordered; the main unknowns are the form factors of the primary fiber ISP modem and the failover connection, which will probably be Starlink. I may also add another 2U of equipment later.
The planned equipment includes four 48-port PoE switches, two UDM Pro Max firewalls, a Dell server, a UniFi Enterprise NVR, and two 3000VA UPS units. One UPS will have an additional battery bank and a 120V transformer. There will also be two managed PDUs for the 208V power equipment.
I'm considering adding a redundant power supply for the switches and firewalls, along with an 8-port aggregation switch. The two firewalls would use two ports, and there are six switches total across the building, with two located in other IDF closets. I've attached the rack diagram and would mainly like feedback on the physical layout, airflow, cable management, serviceability, power distribution, and any issues I may have overlooked.
3 Answers
For the physical arrangement, I’d put the firewalls and other edge devices near the top, followed by the core or aggregation switching, then the access switches and servers. That roughly follows the traffic path from the WAN edge toward the rest of the network, although it’s partly a matter of preference.
Make sure the layout leaves room for rear-mounted PDUs, power plugs, cable bend radius, and service access. Horizontal PDUs can consume valuable rack space, while vertical PDUs usually avoid that but still need clearance behind the equipment. Also leave some usable space for the additional 2U you may add later.
The biggest concern is that the Dell server appears to be a single point of failure. Redundant UPS units and a backup WAN connection help with power and connectivity, but neither protects you if that server fails. Depending on what it runs, consider a second server, virtualization host, replication target, or at least a tested recovery plan.
It would also be worth confirming the expected number of users and VLANs, whether you need static public addresses for vendor allowlisting, how 802.1X will be handled, and how printers, cameras, and other unmanaged devices will be placed on the network.
The equipment choice may be perfectly reasonable for a small or moderately complex organization, especially if the environment has already been running reliably and the cost savings are significant. The more important question is whether the design has an adequate support and recovery plan when something fails.
For the rack itself, focus on separating high-heat equipment where possible, keeping the UPS units and batteries serviceable, maintaining clear airflow, labeling both ends of every cable, and leaving expansion space. Document the power circuits and make sure the UPS and PDU ratings match the actual loads, especially with 208V equipment.
That’s helpful. I’m mainly looking for layout feedback rather than debating the hardware vendor, but the points about service clearance, power ratings, and documenting the circuits are exactly the kind of things I wanted to catch before installation.