Toggle menu

Pulse

Installation, Components and Glossary

Pulse is installed as part of the regular deployment of the Standardised Site. It was first released in May 2026.

Form App Shortcut

Access to the dashboards is managed via an iCM Form App Shortcut. We recommend creating a new iCM user group called "Pulse" which will grant access to the dashboard to users in the group.

Pulse Shortcut

The shortcut should use the PULSEDASHBOARD form (found under the GOSS/Pulse form group).

iCM Objects

Pulse uses two iCM Object Types to store data:

  • form_PULSECONFIG - Stores the configuration for each monitor
  • FORM_PULSE - Stores the latest statistics for a service

History

The Platform's History service records the historical statistics logged during monitoring. These are used to build the details view of the dashboards.

Histories are automatically deleted when they fall outside of the retention period. This defaults to four days, but can be overridden as part of the "Advanced" settings when creating a monitor.

Scheduled Task

The end point goss.pulse.scheduled.pollServices is called every five minutes. The parameters are an empty object.

Glossary

The following terms are used in this documentation and throughout the Pulse service.

Live Service - A service that doesn't have a configured URL or end point to heartbeat check, so it is expected to only be used by invocations from things like form submissions, other end points, workflows etc.

Window - The "window" number is how many events are currently being used to calculate the summary stats being shown. This defaults to 100.

Fails - This denotes the number of "consecutive" failures, ie how many have happened back to back. This helps see if the service is actually down or if a single failure is marking it as such. This along with the uptime % gives you a wider picture of service health.

P95 latency - 95th percentile of latency. 95% of data points fall below this latency so you can see a more indicative picture of the general quality of a service response speed.

Healthy - A healthy service has a decent uptime, isn't showing common signs of slow responses and hasn't had any consecutive failures. 

Degraded - A degraded service is one that has a latency that is above its configured degraded setting. It is slower than it should be but its not failing.

Down - The service isn't working. It is either timing out or returning an error.

Last modified on 12 May 2026