Changelog

Checks from Northern Europe now run via Helsinki. The location sits on its own network (AS202053), so it counts as a full independent vote when several locations decide whether a service is really down. Like every location it is optional: you decide in your settings whether Helsinki should check your services — until then it receives no checks from you.

UpCloud Oy (Helsinki, Finland) runs the new checking location and is now listed in our public provider list and in Annex 2 of the data processing agreement. The location is inside the EU, so no standard contractual clauses are needed. We clarified two things in the DPA: checking locations only receive your data once you select them — and we described what the checking servers actually do: results are sent to us encrypted straight away, and if that fails for a moment they sit locally on the server until the transfer goes through.

Checks from the US East Coast now run from New York. Ashburn was the original plan; New York took its place because it sits on its own network and therefore judges independently of our German locations. You can select it in your service settings.

Both channels are now in our public provider list and in the data processing agreement. Nothing about the channels themselves changes: data is only sent once someone has added and confirmed the contact method.

It used to be frozen at install time and corrected by hand after a provider change. The agent now determines it continuously, including with dynamic addressing.

Services and status pages can be bundled into an area and granted to specific teams or roles. Every other member sees nothing of them — not in lists, not in incidents, not in notifications.

With no areas, nothing changes: everything stays visible to everyone, as before.

Open access areas

LIVI shows the overall state of every service. Colour and face change as soon as something goes down.

After the pilot it is open to every organisation: CPU, memory, disks and running services next to your other checks. Nobody has to request access any more.

Connect a server

The server overview now shows every partition, not just the fullest.

A ready-made set for the status page, instead of adding each series by hand.

Create, change and delete — the page itself and every component, plus appearance, access control, logo and favicon.

A status page no longer has to be rebuilt by hand in every environment.

Create tags, assign services, fill status page groups from a tag. New services then land in the right group on their own.

Add and verify without opening the console.

A status page's charts can be set through the API too.

Translated fields also come back with every language, so writing them back loses no language you did not request.

What the console rejects, the API rejects too.

Duty rotates daily, weekly or on a pattern of your own. Absences and stand-ins go into the same schedule. If a period ends up unstaffed, the schedule flags it on its own.

Open on-call

Every alert carries a link that works without signing in. One tap stops the escalation.

Your own shifts as a calendar subscription — in Outlook, Google Calendar or Apple Calendar.

Fires a real alert through every configured channel, so a wrong phone number turns up now rather than during a real incident.

An overview of who was on duty for how long.

Previously the organisation's language applied.

Add a provider's status page and pick the services you depend on. They then sit in your list like any other service, ready for your status page.

Atlassian Statuspage, Cachet and LIVCK-hosted pages are supported, with ready-made connections for well-known providers.

Open sync sources

Mirrored services can be renamed. LIVCK tags the provider and the connection automatically.

Mirrored services show their provider's icon.

Add, remove or rename services without recreating the connection.

Their status comes from the provider — our probes never query them directly.