Skip to main content

Response times and escalation

Every alert IFMS raises comes with a clock. If nobody takes ownership of it before the time runs out, the alert is marked as having missed its window and is pushed back out to everyone. This page explains the two badges that record that — Breached and L1 — and how to work with them.

The response window​

The window is measured from the moment the alert was raised to the moment somebody acknowledges it. Closing the alert stops the clock just as well as acknowledging it does.

Severity of the alertYou have
Critical5 minutes
High15 minutes
Medium60 minutes
Lowno window
Infono window
(policy severity left unset)no window

The severity comes from the policy that raised the alert (Create an alert policy). Two kinds of alert set their own: legality violations are always Critical, so they carry a 5-minute window, and integration-health alerts are always High, so they carry 15 minutes.

The badge on a row does not tell you the window

Rows show three severity badges where policies use five levels. A row badged Critical may be a Critical alert (5 minutes) or a High one (15 minutes), and a row badged Info may be Low, Info, or an alert from a policy with no severity at all — none of which ever breach. To know an alert's real severity, filter by Severity on Events, which uses all five levels.

These windows are set for the whole installation, not per policy, and there is no screen in the portal for changing them. If your command centre needs different ones, that is a request to the IFMS team.

"Breached": the window was missed​

When an alert is still Open and more time has passed than its window allows, the SLA column shows a red Breached badge with a warning triangle, the row gains a red left edge and a pale red background, and the AGE figure turns red. The detail panel shows the same thing as an SLA breached badge in its heading.

An alert that has not breached shows "—" in the SLA column.

There is no warning before the window runs out

IFMS has no "due soon", "at risk" or countdown state. An alert is either Breached or it is not. With a 5-minute window on a Critical alert, that leaves very little room: the practical answer is to acknowledge Critical alerts as soon as you open them, before you start investigating, rather than after.

"Breached" disappears when you acknowledge

Only Open alerts can breach. The moment you acknowledge a late alert — or close it — the Breached badge goes away and the SLA column reverts to "—", even though the alert really was answered late. The only lasting trace of a missed window is the violet L1 badge described below, and the times in the handling timeline. If you are reporting on response times, count L1 alerts, not Breached ones.

"L1": IFMS escalated the alert​

Roughly once a minute, IFMS looks for alerts that are still Open and have passed their window, and escalates them. Escalating an alert does three things:

  1. it marks the alert with a violet L1 badge, in the STATUS column and in the detail panel's heading;
  2. it adds an Escalated step to the alert's handling timeline, showing "Level 1" and the time;
  3. it sends the alert's notifications again — a fresh item in every active colleague's Inbox, and fresh emails and text messages to everyone covered by a matching alert routing rule.

The repeat notification arrives titled just Alert, with the alert's message prefixed "ESCALATED (level 1):". That prefix is how you tell a repeat from the original.

Things worth knowing about escalation:

  • It is not a status. An escalated alert is still Open and still needs acknowledging and closing in the usual way.
  • It happens at most once per alert. L1 is the only level that exists; there is no L2, however long an alert is left.
  • The badge never goes away. It stays on the alert for the rest of its life, including after the alert is closed. That is deliberate — it is the permanent record that the alert was answered late.
  • It is automatic. There is no Escalate button anywhere in the portal and no permission for it, despite the supervisor role being described as one that "escalates alerts".

Find the alerts nobody answered in time​

  1. Open Events.
  2. Set Alert kind to All alert kinds, so legality and integration-health alerts are included.
  3. Set Escalation to Escalated only.
  4. Set Status to Open for work that still needs doing, or leave it cleared to review everything that was ever late.

This pair of filters is the most useful thing on the Events screen for a supervisor. The Escalated count card does something similar but counts closed alerts as well, so it is a historical figure rather than a queue.

To take the list away with you, use Export CSV: the exported file carries Escalation level and Escalated at columns, along with who acknowledged and closed each alert and when. See Export the alert list.

Keeping response times down​

tip
  • Acknowledge first, investigate second. The note can be as short as "Picked up — checking with the provider".
  • Work Status: Open with Severity: Critical, then High, before anything else. Those are the only two severities with a window short enough to miss during an ordinary shift.
  • Alert routing is what actually reaches someone away from a screen. If Critical alerts are being missed at night, the fix is usually a routing rule with SMS, not a change on this screen: Alert routing.
  • Legality violations are Critical and therefore carry the shortest window of all, even though they are enforcement matters rather than emergencies. Expect them to breach if the freight register is ever switched on during quiet hours, and plan who covers them.