Create an escalation policy
curl "{{API_URL}}/api/v1/escalation_policies/" \
--request POST \
--header "Authorization: meowmeowmeow" \
--header "Content-Type: application/json" \
--data '{
"escalation_chain_id": "F5JU6KJET33FE",
"type": "wait",
"duration": 60
}'
The above command returns JSON structured in the following way:
{
"id": "E3GA6SJETWWJS",
"escalation_chain_id": "F5JU6KJET33FE",
"position": 0,
"type": "wait",
"duration": 60
}
Parameter | Required | Description |
---|---|---|
escalation_chain_id |
Yes | Each escalation policy is assigned to a specific escalation chain. |
position |
Optional | Escalation policies execute one after another starting from position=0 . Position=-1 will put the escalation policy to the end of the list. A new escalation policy created with a position of an existing escalation policy will move the old one (and all following) down in the list. |
type |
Yes | One of: wait , notify_persons , notify_person_next_each_time , notify_on_call_from_schedule , notify_user_group , trigger_action , resolve , notify_whole_channel , notify_if_time_from_to . |
duration |
Optional | The duration, in seconds, when type wait is chosen. |
important |
Optional | Default is false . Will assign “important” to personal notification rules if true . This can be used to distinguish alerts on which you want to be notified immediately by phone. Applicable for types notify_persons , notify_on_call_from_schedule , and notify_user_group . |
action_to_trigger |
If type = trigger_action |
ID of an action, or webhook. |
group_to_notify |
If type = notify_user_group |
ID of a User Group . |
persons_to_notify |
If type = notify_persons |
List of user IDs. |
persons_to_notify_next_each_time |
If type = notify_person_next_each_time |
List of user IDs. |
notify_on_call _from_schedule |
If type = notify_on_call_from_schedule |
ID of a Schedule. |
notify_if_time_from |
If type = notify_if_time_from_to |
UTC time represents the beginning of the time period, for example 09:00:00Z . |
notify_if_time_to |
If type = notify_if_time_from_to |
UTC time represents the end of the time period, for example 18:00:00Z . |
HTTP request
POST {{API_URL}}/api/v1/escalation_policies/
Get an escalation policy
curl "{{API_URL}}/api/v1/escalation_policies/E3GA6SJETWWJS/" \
--request GET \
--header "Authorization: meowmeowmeow" \
--header "Content-Type: application/json" \
The above command returns JSON structured in the following way:
{
"id": "E3GA6SJETWWJS",
"escalation_chain_id": "F5JU6KJET33FE",
"position": 0,
"type": "wait",
"duration": 60
}
HTTP request
GET {{API_URL}}/api/v1/escalation_policies/<ESCALATION_POLICY_ID>/
List escalation policies
curl "{{API_URL}}/api/v1/escalation_policies/" \
--request GET \
--header "Authorization: meowmeowmeow" \
--header "Content-Type: application/json"
The above command returns JSON structured in the following way:
{
"count": 2,
"next": null,
"previous": null,
"results": [
{
"id": "E3GA6SJETWWJS",
"escalation_chain_id": "F5JU6KJET33FE",
"position": 0,
"type": "wait",
"duration": 60
},
{
"id": "E5JJTU52M5YM4",
"escalation_chain_id": "F5JU6KJET33FE",
"position": 1,
"type": "notify_person_next_each_time",
"persons_to_notify_next_each_time": [
"U4DNY931HHJS5"
]
}
]
}
The following available filter parameter should be provided as a GET
argument:
escalation_chain_id
HTTP request
GET {{API_URL}}/api/v1/escalation_policies/
Delete an escalation policy
curl "{{API_URL}}/api/v1/escalation_policies/E3GA6SJETWWJS/" \
--request DELETE \
--header "Authorization: meowmeowmeow" \
--header "Content-Type: application/json"
HTTP request
DELETE {{API_URL}}/api/v1/escalation_policies/<ESCALATION_POLICY_ID>/
Related Grafana Cloud resources
Intro to Prometheus and Grafana Cloud
Prometheus is taking over the monitoring world! In this webinar, we will start with a quick introduction to the open source project that’s the de facto standard for monitoring modern, cloud native systems.
How to set up and visualize synthetic monitoring at scale with Grafana Cloud
Learn how to use Kubernetes, Grafana Loki, and Grafana Cloud’s synthetic monitoring feature to set up your infrastructure's checks in this GrafanaCONline session.
Using Grafana Cloud to drive manufacturing plant efficiency
This GrafanaCONline session tells how Grafana helps a 75-year-old manufacturing company with product quality and equipment maintenance.