Network Watch

Your show runs on the network.
Nobody is watching it.

Interface errors climb for hours before anything is audible. Fiber fades for weeks before it fails. A clock master changes and the system limps until someone power-cycles it. It's all already in your switches.

DANTE  ·  AES67  ·  sACN  ·  ART-NET
NDI  ·  SMPTE ST 2110  ·  NMOS  ·  PTP
reynolds-hall · core-sw-01 MONITORING
P3 · --:--:-- · AUTOMATED TRIAGE
What actually happens

Nothing fails suddenly.
It fails visibly, suddenly.

A dropout during a show is the last event in a chain that started hours earlier. Every link in that chain is measurable, and almost nobody measures it.

T − 6:00
Errors begin
A connector on a fiber run seats slightly loose during load-in. Input errors start accumulating at a rate nobody would notice.
T − 4:15
Threshold trips
Error rate crosses the tuned baseline for this site. Alert fires. A named engineer acknowledges within the response window.
T − 3:40
Fixed during focus
Port identified, connector cleaned and reseated while the room is still empty. Total cost: fifteen minutes of a tech's afternoon.
T − 0:00
The show that didn't drop
Nobody in the audience learns anything. Nobody in the building knows it nearly happened. That's the product.
The console

Two views of the same network.

Your team gets a room they can read at a glance. We get the instrumentation. Same data, different jobs — nobody should have to learn PromQL to find out whether tonight is going to be fine.

Venue console — what your technical director opens on a phone at 6pm.
Reynolds Hall
12 DEVICES · 214 PORTS · LAST SWEEP 14s AGO
ALL SYSTEMS NOMINAL
Clock
LOCKED
Optical margin
+4.1 dB
PoE headroom
38%
Open issues
1
Live transport
Attention
P3 · OPTICAL
core-sw-01 Te1/0/23 has drifted 1.4 dB below its 7-day median. Not urgent. Clean and reseat at the next dark day — we'll book it if you want.
LAST 30 DAYS
ISSUES CAUGHT ······ 6
SHOW-TIME FAULTS ··· 0
MEAN ACK ·········· 4m 12s
Eighty32 — Operations
8 SITES · 71 DEVICES · 2 SHOWS LIVE
1 SITE NEEDS EYES
Triage queue
P2 · PTP · CANYON AMPHITHEATRE
Grandmaster changed at 18:42 — domain 0 moved from the Yamaha to a laptop that joined switch 3 port 14. Correlated: STP topology change on the same switch, same minute. Someone plugged in. Recommend isolating the port before doors.
P3 · MULTICAST · GRAND BALLROOM
sACN universes 12–19 flooding the audio VLAN. IGMP snooping disabled on the switch added last Tuesday. Presents as an audio fault; it is a lighting one.
The one nobody watches

ST 2022-7 is designed to hide a failure.

That's the whole point of seamless protection switching — and it's exactly the problem. A leg drops, the receiver reconstructs from the survivor, and the picture never flinches. Nothing tells anyone. You run unprotected until the second leg goes, and the second leg always goes on air.

SOURCE RECEIVER LEG A · RED NETWORK LEG B · BLUE NETWORK
What the receiver reports
NO ERRORS · OUTPUT CLEAN
Stream-level monitoring sees a healthy picture and shows green. It is telling the truth. It is also useless here.
What Network Watch reports
BOTH LEGS NOMINAL
We poll each path independently, so the loss of redundancy is an event in its own right — not a silence.
Coverage

Every payload on the show network.

It's one network and one clock underneath all of it. A fault in the transport doesn't care which payload it ruins — so we don't scope ourselves to one.

AUDIO
Dante
AES67 · AVB
Grandmaster identity, clock offset excursions, per-port error and discard rates.
LIGHTING
sACN / E1.31
Art-Net · DMX
Multicast flooding, IGMP snooping and querier state, universe loss, broadcast storms.
VIDEO
NDI · NMOS
SMPTE ST 2110
Per-flow bandwidth against provisioned capacity, ST 2059-2 lock, 2022-7 leg integrity.
TRANSPORT
Fiber · PoE
Switching · PTP
Optical Tx/Rx drift, topology changes, PoE headroom, configuration drift from baseline.
How it works

Three layers. Only one of them is clever.

A fanless appliance goes in your rack and polls read-only over SNMP. It carries its own cellular connection, so when your network is the thing that failed, our visibility doesn't fail with it.

LAYER 01
Deterministic alerting

Tuned thresholds on your actual traffic. No model anywhere in this path. This is what fires at 3am and what your service level is measured against.

Never skipped
LAYER 02
Correlation

Every fifteen minutes, anything that tripped gets correlated across devices and protocols, then written up in a sentence a tech can act on at 2am.

Adds context
LAYER 03
A person

Named lead engineer on Support and Critical tiers. Escalation goes to someone who has your network documented, not to a queue.

The actual product
The rule the whole system is built around: automated correlation can raise the severity of a finding. It can never lower one, and it can never suppress a page. A missed P1 during a show ends a contract — so nothing clever is allowed anywhere near the decision to wake someone up.
Pricing

Per site, per month.

Each tier covers one site with up to eight managed switches. Twelve-month agreement, cancel for convenience after month six.

MONITOR
$650
PER SITE / MONTH
  • Live dashboard access
  • Email and SMS alerting
  • Monthly health report
  • Business-hours email support
  • Next business day response
Start here
SUPPORT
$1,600
PER SITE / MONTH
  • Everything in Monitor
  • Named lead engineer
  • 4-hour response, show hours
  • Remote remediation
  • Config backup and restore
  • On-site health check, twice yearly
Talk to us
CRITICAL
$3,200
PER SITE / MONTH
  • Everything in Support
  • 24/7/365, 1-hour response
  • Guaranteed remote-hands escalation
  • Critical spares staged on site
  • Quarterly on-site health check
  • 4 event-standby days per year
Talk to us
ADD-ONS  ·  EVENT-DAY STANDBY $1,200/DAY  ·  ON-SITE LAS VEGAS $225/HR  ·  RF SPECTRUM SURVEY $2,400  ·  ADDITIONAL SWITCHES $85/MO
Before you ask

The questions we always get.

Are you going to try to take over our network?

No. We monitor and advise on the network you have. We don't assume ownership or design liability for systems installed by others. Where we find a design fault we document it and quote the remedy separately — and you're never obligated to use us for the fix.

Does this see our audio or video content?

No. We collect device telemetry, configuration data and topology information. We do not capture, record or inspect production traffic traversing your network. That's written into the services agreement, not just promised here.

What if the fault is in our internet or our consoles?

Then it's outside scope and we'll say so quickly, which is still useful. WAN, carrier services, end-user devices, DSPs, media servers and application software are excluded — along with anything not on the covered device list. The exclusions are enumerated in the agreement so there's no argument about it later.

You're guaranteeing our network won't go down?

No, and be suspicious of anyone who does. The service is early detection and notification. Some failure modes have no observable precursor. What we commit to is a response time from alert acknowledgement — measured, credited if missed, and never dependent on anything automated deciding it wasn't important.

We already have a monitoring tool.

Most venues have something that watches whether devices are up. That's a different question from whether they're healthy — an interface passing traffic with climbing error counters is up. And an ST 2022-7 stream missing a leg is up. Bring us the tool you have and we'll tell you honestly whether this adds anything.

What does onboarding actually involve?

An afternoon on site. We install the appliance, discover and document the network, tune thresholds against your real traffic, and hand you a written baseline report of what we found. The report is yours whether or not you continue. It usually surfaces two or three real problems on day one.

Get started

Start with the baseline.

Book the onboarding audit on its own — $3,500, one afternoon. We document your network, tune the thresholds, and hand you a written report of what we found. If you don't continue, you keep the report.

Email us
hi@eighty32.com
702-907-4762

3725 W Teco Ave #8
Las Vegas, NV 89118

FULLY INSURED · COI ON REQUEST