Work
Two engagements, reported without naming them.
Client names are withheld pending written consent. Every measurement below came off a real console or a real survey on the date shown.
A multi-campus church, Oahu
Inherited a flat network on aging hardware with a controller that had already failed once. The brief was coverage. The finding was that coverage was the third problem in the queue.
What was found
- One GbE uplink carried every switch in the building. The core was uplinked into its own port, so a single cable was the whole estate
- Seven networks existed in the console and shared one physical /24. The segmentation was configured and not in effect
- 257 competing wireless networks within range, driving 45 to 68 percent airtime utilization on 2.4 GHz
- An access point had been dead for eleven days. The hardware and the cable were both fine
- Three switches had been added since the previous baseline with no record of parent switch or port
What was done
- Controller migrated off the failed appliance onto a supported host
- Wireless survey across sanctuary, offices and classrooms, then a measured channel and power plan
- Segmentation model written, with the cutover staged as a relocation event rather than a configuration change
- Asset register, topology and credential register produced and handed to the volunteer team
| Measured | Value |
|---|---|
| Access points | 13 |
| Switches | 5 plus 3 compact |
| Networks defined | 7 |
| Physical subnets in use | 1 |
| Competing networks | 257 |
| 2.4 GHz airtime | 45 to 68 percent |
| Links below rated speed | 3 |
| Facility area | 39,551 sq ft |
The access point that was dead for eleven days is written up in full under field notes. The cause was a documentation gap and the fix took ninety seconds once it was found.
A multi-site ministry, Maui
An inherited multi-vendor estate with two perimeter security products, two remote access tools, overlapping forms and telephony platforms, and no asset register. The first deliverable was knowing what existed.
- Asset and systems register built from a read-only console audit
- Ten overlapping products identified and registered for consolidation review
- An operations manual separated from the credential register, so one can be shared and the other cannot
- Licensing reviewed against actual active accounts rather than purchased seats
No savings figure is published here. Consolidation findings are blocked pending invoices and contracts, and I will not put an estimate in a public case study.
Your building probably has one of these
A single uplink carrying everything, or segmentation that exists only in the console, are the two most common findings I walk into. Both are cheap to identify and neither shows up on a speed test.