Escalation policies make sure the right people are alerted at the right time.
If an incident is not acknowledged or resolved within the escalation timeout period, it is passed on, or escalated, to the next users in line. An escalation policy lets you customize who receives the alert first, the amount of time to wait before escalating to the next users, and which users should receive the alert next.
Incidents can be assigned to multiple users or to users on-call on multiple schedules within an escalation policy. An escalation policy ties services and users together. Each service can only be assigned to one escalation policy, but an escalation policy can be assigned to multiple services.
A triggered incident will first be assigned to the on-call schedules or users in Escalation Rule 1 (the first level of the escalation policy). If nobody is on-call in the first level, then we will assign the incident to the next user who is on-call in the next subsequent escalation level. If the incident is not acknowledged within the escalation timeout, then it will be escalated to the on-call schedules or users in Escalation Rule 2, and so on down the line.
Escalation Policy Example
In the example below, if an on-call user in the Primary On-Call Schedule schedule does not acknowledge an incident in 30 minutes, then it will be escalated to Jason Collison. At that point, since Jason's escalation timeout setting is set to 30 minutes, he will have 30 minutes to acknowledge the alert before it is no longer assigned to him and the escalation policy starts over. This escalation policy is set up to repeat 9 times if no one responds.
Why did my incident follow the old escalation policy rules after I edited it?
Once an incident is triggered, it will follow the escalation policy rules as they existed at the time of incident creation. This means that if the following series of events takes place:
- Incident is created
- Escalation policy is modified
- Incident escalates
then the incident will be escalated to the level on the escalation policy that was defined prior to step 2: Escalation policy is modified.
Are there any limits on how many layers and users I can add to an escalation policy?
Here is some additional information about escalation policies:
You can add up to 20 layers/escalation rules to an escalation policy.
You have the option to repeat the set of escalation rules you define up to a maximum of 9 times. Once an escalation policy reaches the amount of loops configured, the incident will stay assigned to the last user and will not continue to alert once it has cycled through all of the contact methods for that user.
You can alert up to 10 users/schedules (25 on Standard and Enterprise plans) within a single rule of an escalation policy.
The minimum amount of time that can be set between escalation rules is 1 minute if there is one user/schedule on an escalation rule. Otherwise, the minimum is 5 minutes if there is more than one user/schedule on an escalation rule. See What is the minimum amount of time that you can have between escalations? for more information.
What is the minimum amount of time that you can have between escalations?
Within an escalation policy, you can set up escalation timeouts. These escalation timeouts will escalate a triggered incident to the next user or on-call schedule after the incident has been in the triggered state for
X number of minutes.
Escalation timeouts can be set to a minimum of 1 minute if there is one user/schedule on an escalation rule.
Escalation timeouts must be set to a minimum of 5 minutes if there is more than one user/schedule on an escalation rule (as seen in the screenshot above).
A 5-minute grace period gives the on-call users within an escalation policy a chance to review and respond to the triggered incident. In other words, setting an escalation timeout to less than 5 minutes would limit the amount of time that your first responders can attend to an incident before it is assigned to someone else.
Learn more about how to create an escalation policy in Creating an Escalation Policy.