NOC2 · away from the desk
Operations · anywhere

The gateway does not wait
for you to reach a desk.

NOC2 gives an operator three surfaces onto the same live network: a phone console built for triage, the full console on tablet and desktop, and a wall-board for the operations room. One login, the same data, nothing to install.

Most faults arrive at the least convenient moment. What decides how bad they get is not the tooling in your office — it is what you can see and do from wherever you happen to be standing.

3
surfaces, one system
13
alert conditions watched
1 tap
to acknowledge from a phone
0
apps to install

One system, three surfaces

One system, three surfaces A phone console at /m for triage, the full console on tablet and desktop, and a wall-board at /wall for a war-room screen. All three read the same live data and use the same login. PHONE alerts first, ack in a tap /m TABLET & DESKTOP the full console / WALL war-room screen /wall one login · the same live data · nothing to install

These are not three products. They are three doors onto the same running system, and which one you use is a matter of what is in your hand at the time.

Phone

Built for triage

Opens on alerts, because that is why you reached for your phone. Bottom tabs for alerts, gateways and subscribers — nothing else competing for the screen.

Tablet & desktop

The whole console

Everything: per-gateway metrics, sessions, configuration, capacity, subscriber history. The same pages, laid out for the room you have.

Wall-board

For the room

A full-screen view with no sidebar and no chrome. Pin a browser to it on a television and the operations room has a shared picture.

What the phone actually gives you

The phone view is not the desktop console squeezed into a narrow column. It is a separate, purpose-built surface, and it holds the three things you need when you are not at a desk.

Phone console showing an alert list: each card carries the alert type, the gateway, a plain-English cause such as a process being down or subscribers dropping, and an acknowledge button.
What is wrong, and why. Every alert carries its reason — a process down, subscribers dropped with the figures that triggered it, a gateway not answering ping or TCP. Active, acknowledged and all, kept apart.
Phone console showing the gateway list with online, offline and maintenance counts across the top and per-gateway cards with address, CPU and memory.
Every gateway, at a glance. Online, offline and in-maintenance counted at the top; each gateway with its address, CPU and memory. Search by hostname or address when you know what you are looking for.
Phone console showing the subscriber search surface for looking up an individual customer session.
The customer on the phone. Look up a subscriber while they are still talking to you, from the car park, without asking anyone to remote in to a terminal.

Acknowledge from the card. You do not need to get to a laptop to stop a page from repeating and let the rest of the team know it is being handled. One tap, on the alert itself.

How an alert reaches you

The path an alert takes A threshold is crossed on the gateway, the agent reports it, NOC2 raises an alert and sends an email, and the engineer opens the phone console to see the cause and acknowledge it. 1 ON THE GATEWAY a threshold is crossed 2 NOC2 RAISES IT with the reason attached 3 EMAIL GOES OUT to the people on the rule 4 YOU OPEN /m cause, gateway, and Ack WORTH KNOWING Acknowledging silences that alert. It is a decision, not a dismissal — so the console shows you what is acked, separately, rather than hiding it.

NOC2 watches thirteen distinct conditions across your estate, and each one carries its own threshold, its own occurrence count before it fires, and its own list of who to tell:

server_offlineprocess_downno_data cpu_usagememory_usagedisk_usage temperatureuser_droppool_exhaustion pppoe_completionaccounting_lossinterface_down bgp_down

Note what is in that list beyond the usual. user_drop catches subscribers leaving a gateway that is otherwise perfectly healthy. pppoe_completion catches sessions that start and never finish. accounting_loss catches a gateway that is forwarding traffic while its records quietly stop. None of those show up as down, and all of them are somebody’s Monday morning.

Monitoring that pages you

A message that says a threshold was crossed, and leaves you to find out what and where.

Requires a laptop, a VPN and a terminal before you can form an opinion.

Silence means either nothing is wrong or nothing is watching. You cannot tell which.

NOC2

The alert names the gateway and the cause, with the numbers that triggered it.

Opens in the browser you already have, on the device already in your hand.

The console states how much it could actually read, so an unread gateway is never mistaken for a healthy one.

On a tablet, and on the wall

The same console on a tablet-width screen, centred with comfortable margins and larger touch targets.
The same surface, given more room. On a tablet the console keeps a comfortable reading width rather than stretching the content thin — the layout follows the screen instead of fighting it.
The full-screen wall-board view: no sidebar or header, dark theme, showing fleet state for a shared operations-room display.
The war-room view. Gateways online against the total, subscribers live, throughput both directions, sessions flapping over the last day, attack packets dropped, and the alerts still open — with a sixty-minute throughput trace under it. No sidebar, no header, dark by default. Pin a browser to it on a television and the room shares one picture without anyone driving.

What this does not do

It does not push notifications to the phone. Alerts are delivered by email, and the phone console is where you go to act on them. We would rather say that plainly than imply a notification you would not receive.

Acknowledging silences an alert. That is deliberate — it stops a known fault from paging the team repeatedly — but it means acknowledgement is a decision, not a way to clear a list. The console keeps acknowledged alerts on their own tab rather than hiding them.

It is a browser, not an installed app. On iOS and Android you can add it to the home screen and it opens like an app, full-screen with its own icon. There is nothing to approve in an app store, and nothing to update on a handset.

The test is simple. A gateway stops answering while you are on a train. Can you find out which one, learn why, tell the team you have it, and know how many subscribers are affected — before you get off? That is the whole design brief.

About the screenshots. Every screen shown is a real production console, captured at true phone, tablet and display sizes. Gateway hostnames are replaced with placeholders and every IP address is replaced with an address from the RFC 5737 documentation range before the image is taken; no operator name, site, region or subscriber identifier appears in any image.