Cactus Console Site A · BMS/EPMS
mock adapter
Window
Customer
Hall
System
Phase
Point type
Severity
Contracted estate ·
Halls
Contracted
Utilisation
Open alarms

Service position

live

Load trend, all contracted halls

dc/sitea/+/power/PDU_IT-*/+_PWR · summed per feed

Load by hall

current draw against 2,500 kW contracted per hall

Your halls

click a row to open the hall — live envelope, alarms and equipment

Open alarms across your estate

tenant-visible points only · click for the fault detail

Service level, trailing 90 days

availability measured against the contracted envelope
IT load · ·
Contracted
Utilisation
Cold aisle
Uptime, 90 d

Service envelope

live

Hall load by feed

dc/sitea/dh01/power/PDU_IT-*/+_PWR

Aisle temperature

dc/sitea/dh01/environment/+/_TEMP

Cooling & environment

ASHRAE A1 allowable envelope · 18–27 °C, 20–80 %RH

Alarms in this hall

tenant-visible points only · click a row for the fault detail

Monitored equipment

assets publishing into this hall's topic tree · click for the point list
Points streaming · Site A · 16 halls
Live endpoints publishing to the CACTUS feed, out of commissioned in the RFS1–RFS4 registry.
Feed availability
Site IT load
Active alarms
p95 latency

Estate coverage

registry · 12 Feb 2025

Monitored points by system

dc/sitea/+/{system}/#

Commissioning scope by delivery phase

points grouped by RFS # — each phase is a hall handover tranche

Site load, trailing window

dc/sitea/+/power/+/_PWR · 45 s latency budget

Data hall telemetry density

points per hall — click a hall to open it

Point mix

measurement · alarm · status · command

Active alarms, site-wide

dc/sitea/+/+/+/_ALM · click a row for the fault detail
Admin · Point registry

Every endpoint on site, as commissioned

45,477 BMS and EPMS points across sixteen data halls and the central energy plant. Filter it, search it, then click any row for the tag, the topic it publishes on, and what a 0 or a 1 on that point actually means.

Alarm points by system

registry count of _ALM / trip points in scope

Signal classes in scope

engineering unit families across the filtered registry

Point registry

click any row for the point detail, topic and fault definition
Admin · Alarm console

What needs attention right now

Active faults across the estate, ranked by how long they have been open, with the equipment, the asset, the hall and the customer who is affected by it.

Alarm console

every active fault across the estate — click a row for the definition and history

Open faults by hall

count of unacknowledged alarms

Open faults by system

where the estate is generating noise
Configuration & integrations

What this console is wired into

One adapter sits between the UI and everything else. Change the endpoints here and every panel follows — the broker, the API, the historian, identity, and the alarm routing that decides who gets told.

Integrations

what this console connects to

Data source

the contract behind every panel

Broker & API endpoints

swap the adapter, keep the UI — nothing else changes
mock

Topic scheme & payload

how a registry row becomes an MQTT topic

Customer → hall assignment

the subscription each tenant is granted — this is the visibility boundary
Site code sitea and tenant names are generic placeholders. Map them to the real site and accounts before the first customer login.

Publish & retention

what the broker is configured to do

Open decisions

what the backend team still has to answer