Status
System status
How we watch the platform, how you find out when something is wrong, and what to do if you think something is wrong right now.
Being straight about it
What this page does not do
It does not show a live green tick. That is deliberate, and it is worth explaining rather than quietly omitting.
A status indicator written into a static page says “all systems operational” during an outage just as confidently as it does at any other time — because nothing is checking it. A reassuring green light that cannot go red is worse than no light at all: it actively delays people from reporting a real problem.
A live status feed is on our list. Until it exists and is driven by real measurements, this page tells you how monitoring works and how to reach us instead of pretending to be a dashboard.
How you will actually hear about an incident
- We contact you. For anything affecting your service, we reach out to your nominated contacts rather than waiting for you to check a web page.
- Planned work is scheduled with you in advance, outside your business hours wherever it is possible.
- After anything significant you get a plain‑language account of what happened and what we changed, not a sentence about an “intermittent degradation”.
Monitoring
How the platform is watched
Most of this is invisible to you, which is the point. Two parts are worth describing because they are the difference between finding out quickly and finding out from a customer.
Watched from outside itself
Monitoring that runs on the same equipment as the thing it monitors reports nothing at all when that equipment fails — exactly when you needed it. The platform sends a regular heartbeat to a service running somewhere else entirely. If the heartbeat stops, that external service raises the alarm.
Alerts that do not get missed
Alerts reach a person over two independent channels at once, so one failing delivery route does not swallow the warning. A critical alert repeats until somebody acknowledges it rather than scrolling past it at 3am.
Recorded before it is sent
Every alert is written down before we try to deliver it. A delivery that fails leaves a record and is retried, instead of vanishing. We built it this way after finding a mail configuration that had been silently discarding notifications, and nothing was watching the counter that would have shown it.
Checks that fail loudly
The platform’s own configuration is verified automatically against the running system. When something drifts from what we believe is true, a check fails rather than a document quietly becoming wrong.
If calls are not working
Before reporting, two quick checks that decide where the problem is:
- Is it one phone or all of them? One phone is usually that phone or its cable. All of them, at one site, is usually that site’s internet connection.
- Is anything else on the network working? If nothing can reach the internet, the phones are a symptom rather than the cause.
Then tell us, with whichever of these you have:
- The number that was called or calling
- Roughly when — within a few minutes is plenty
- Which extension was involved
- What happened — no ring, no audio one way, cut off, or poor quality
With the number and the time we can find the call itself and see what occurred, which is far faster than asking you to make it happen again.
Emergency calls
If you have any reason to believe 911 is not working from a phone, treat it as urgent and contact us immediately by phone rather than by email.
Please do not test by dialing 911. Occupying a dispatcher to check a configuration is discourteous and, in many places, an offence. We can verify emergency routing from our side without placing a call to a real dispatcher, and we will show you the result.
Need us now?
Existing customers should use the support route from handover. If you cannot find it, this form reaches us during business hours.