Skip to content
plugboard.appRequest a demo

Running the desk

Major incidents

Needs module.tickets. Declaring, adding tickets and ending need ticket.create and ticket.update.

A major incident is a fault a lot of people report at once: the Wi-Fi in one building, the student information system, printing across the school. One incident ticket owns it. The reports hang under that ticket, each requester is told once, and the school sees one notice on the portal and kiosk until it is over.

On the desk it reads Major incident: a chip and a callout on its ticket, a badge on its row in the queue, and a line in the banner above the queue. The portal and kiosk call the notice a known issue.

  1. Open the incident’s ticket.
  2. Open More actions (the button with three dots) and choose Declare major incident.
  3. Give it a Title people will recognise, at least three characters, and optionally a Message: what is affected, and what people should do meanwhile. Keep other people’s details out of both.
  4. Choose the Service it is about, if you can see services. The ticket’s own service is chosen to begin with, and None names none. See naming a service.
  5. Choose Declare.

People at the ticket’s campus see the title and message on the portal and kiosk until it ends. A ticket with no campus puts the notice up for the whole school.

Declare major incident is offered on an open ticket of type Incident that is not part of another ticket and is not already a major incident. A ticket whose major incident has ended can be declared again while it is open.

A major incident can name the service it is about. While the major incident is open, the service reads Down: on its page, on Services, on Home and, for a service on the status page, there too. The callout says so, such as Wi-Fi reads Down until this ends, and the service’s page links back to the ticket. When the major incident ends, or its ticket is resolved, the service goes back to its monitors.

On an open major incident, Name a service… on the callout, or Change service… once one is named, asks Which service is this about?. Choose a service, or None to clear it, and Save. Only services at your campuses are offered. It needs ticket.create, ticket.update and service.view.

A service going Down this way starts the flows that wait for a service to go down, as any other outage does.

  1. On Tickets, select the reports.
  2. Choose Add to major incident in the bar that appears.
  3. Choose the major incident. Each open one is listed with its ticket number and its campus, or “Whole school”.
  4. Leave Tell each requester on to send each requester one update, or turn it off.
  5. Choose Add.

Each ticket hangs under the major incident’s ticket. The update is a fixed template, never text from a model: that the desk knows about the fault and is working on it, the title and message as typed, and that the requester’s ticket is part of the incident so there is no need to report it again. Telling requesters needs ticket.comment.

Add up to 200 tickets at a time. Every ticket is read with your own desk and campus access, and a selection that includes a ticket you cannot open is refused. The confirmation says how many were added and how many people were told, and names any ticket that could not be added.

  • Major incidents in the Filter menu on Tickets lists the tickets that own one, with a chip you can remove.
  • The banner above the queue lists each major incident that is up: its title, its campus, how long it has been on the portal and kiosk, its ticket and how many reports are under it.

There are two ways.

Resolve its ticket. The notice comes down, and every open ticket under it is resolved with the last public reply the desk wrote on the major incident’s ticket since it began. Without one, each requester gets a line saying the incident has been resolved. Write the reply that explains the fix before you resolve, and every requester gets it.

End major incident… on the callout. The notice comes off the portal and kiosk. The tickets under it stay attached, and resolving the ticket afterwards leaves them as they are.

Once a major incident has ended and its ticket is resolved, the callout offers Raise a problem, so the cause gets a problem record.

A major incident’s ticket lists Changed in the last 72 hours: the changes in progress, or implemented with a window in the 72 hours before the incident began, that touch it. A change touches it when it links the ticket’s device, or a monitored service the ticket names that is down, or when it is at the ticket’s campus. A change with no campus is the whole school’s, so it is listed too. Each change says why, the most specific first.

The same list shows on an ordinary incident that names a monitored service that went down. It needs change.view, and says nothing when no change is near. It is worked out by rules from the records, with no AI.

On the change itself, the major incident shows under Incidents since this change.

Plugboard watches the tickets coming in for an outage nobody has declared. When at least five people raise tickets in one category at one campus within an hour, well above what that hour has brought over the last eight weeks, it proposes an incident. Three people are enough when equipment in that area, or a monitored service the tickets name, is down.

The proposal appears above the ticket queue, saying how many tickets came in over how many minutes and that nobody has been told, with Why we think this and the equipment behind it. Spotted on Home lists it too, as similar tickets that may be one outage. A proposal stays for two hours.

  • Create incident… makes an incident with the open reports as its children. Give it a title and an optional message, and leave the switch to tell the people who reported it on to send each of them the same one update. The incident is a major incident from then on, with its notice on the portal and kiosk.
  • Dismiss puts the proposal away.

Detection counts tickets. It is statistics, not a model, so it runs whether or not the school’s AI is on, and nothing reaches a requester until a person creates the incident. With the school’s AI on, proposals follow the mode of Propose incidents in AI settings, under Keep tickets tidy; switching that feature off stops them either way.

Creating an incident needs ticket.create and ticket.update, dismissing a proposal needs ticket.update, and telling the reporters needs ticket.comment.

Declaring, with the service it names, adding tickets, naming or clearing its service (Service named on a major incident) and ending, by hand or by resolving the ticket, are recorded in the audit log. Flows never declare, add to, name a service on or end a major incident: each of those puts a notice in front of the school, marks a service Down for everybody or messages its requesters, so a person decides.