Alarms

Alarms bind alerting rules, entities, and notification plans into a logical unit. Alarms are responsible for determining a state (OK, WARNING or CRITICAL) based on the result of a check, and executing a notification plan whenever that state changes. You create alerting rules by using the alarm domain-specific language (DSL).

For information about using the DSL, see Alert triggering and alarms For alarm examples by check, see Alarm examples.

The following table describes the attributes for the Alarms API resource:

Use the following alarms API operations to create, view, and manage alarm resources.

• Alarms
• Test an alarm
• List alarms
• Get alarm by ID
• Update alarm by ID
• Delete alarm by ID

Create an alarm

POST /entities/{entityId}/alarms

Create a new alarm by using a valid set of attributes from the alarms attributes table.

For more information, see Alarm language.

The following table shows the possible response codes for this operation:

Request

The following table shows the header parameters for the request:

Example Create alarm: JSON request

{
"check_id": "chAAAA",
"criteria": "if (metric[\"duration\"] >= 2) { return new AlarmStatus(OK); } return new AlarmStatus(CRITICAL);",
"notification_plan_id": "npAAAAA"
}

Response

This operation does not return a response body.

Test an alarm

POST /entities/{entityId}/test-alarm

The test_alarm endpoint is used to post alarm criteria and alarm data to test whether the alarm criteria is valid and to show how the alarm state is evaluated. You cannot specify an existing alarm ID with this endpoint; instead, you provide the alarm’s criteria within the request. The test_alarm response not only provides the final evaluated state, but also includes all state transitions noticed during the calculation.

To test an alarm, you need to provide one or more observations. An observation consists of a set of metrics from a single data center, such as DFW, LON, and IAD. A group of observations consist of a set of metrics from more than one data center.

Cloud Monitoring passes the observation(s) that you provide to the event processors, and then captures and returns an array of all state changes in the same order that they were evaluated. The state returned by the test alarm depends both upon the criteria and the metrics provided for the test alarm. For example, if you have a criteria that returns “CRITICAL” if a certain metric is above a threshold and “OK” otherwise, you should expect to receive a “CRITICAL” response if you provide a metric with a value above that threshold and an “OK” response if you provide a metric with a value below that threshold.

When you test an alarm, you provide a list of check data, which you can retrieve from a test check operation.

Note

A consecutiveCount of 1 is always used for these calls. If another value is set in your criteria it is overridden.

The following table shows the possible response codes for this operation:

Request

The following table shows the header parameters for the request:

Example Test alarms criteria

:set consecutiveCount=1
if (metric["busy_workers"] > 2) {
return new AlarmStatus(WARNING);
}
if (metric["busy_workers"] > 3) {
return new AlarmStatus(CRITICAL);
}
return new AlarmStatus(OK);

Example Test alarm: JSON request

{
"criteria": "if (metric[\"code\"] == \"404\") { return new AlarmStatus(CRITICAL, \"not found\"); } return new AlarmStatus(OK);",
"check_data": [
{
"timestamp": 1319222001982,
"monitoring_zone_id": "mzxJ4L2IU",
"available": true,
"status": "code=200,rt=0.257s,bytes=0",
"metrics": {
"bytes": {
"type": "i",
"data": "0"
},
"tt_firstbyte": {
"type": "I",
"data": "257"
},
"tt_connect": {
"type": "I",
"data": "128"
},
"code": {
"type": "s",
"data": "200"
},
"duration": {
"type": "I",
"data": "257"
}
}
}
]
}

Response

Example Test alarm: JSON response

[
{
"timestamp": 1319224500831,
"state": "OK",
"status": "Matched default return statement"
}
]

Example Test alarm, multi-state: JSON response

[
{
"state": "OK",
"status": "Ping responds as expected.",
"timestamp": "1417732774575"
},
{
"state": "WARNING",
"status": "blah.",
"timestamp": "1417732795200"
},
{
"state": "CRITICAL",
"status": "cnt=6,avail=0,min=-nan,max=-nan,avg=-nan",
"timestamp": "1417732795216"
}
]

List alarms

GET /entities/{entityId}/alarms

List the alarms on the specified entityId. Use /alarms?id=alarmOneId & id=alarmTwoId to filter the results to only include information about the specified alarms.

This operation can be paginated. For information, see Paginated collections

The following table shows the possible response codes for this operation:

Request

The following table shows the header parameters for the request:

Note

This operation does not accept a request body.

Response

Example List alarms: JSON response

{
"values": [
{
"id": "alAAAA",
"check_id": "chAAAA",
"entity_id": "enZZZZA",
"criteria": "if (metric[\"duration\"] >= 2) { return new AlarmStatus(OK); } return new AlarmStatus(CRITICAL);"
}
],
"metadata": {
"count": 1,
"limit": 50,
"marker": null,
"next_marker": null,
"next_href": null
}
}

Get alarm by ID

GET /entities/{entityId}/alarms/{alarmId}

Get information for a single alarm.

The following table shows the possible response codes for this operation:

Request

The following table shows the header parameters for the request:

Note

This operation does not accept a request body.

Response

Example Get alarms by ID: JSON response

{
"id": "alAAAA",
"check_id": "chAAAA",
"entity_id": "enZZZZA",
"criteria": "if (metric[\"duration\"] >= 2) { return new AlarmStatus(OK); } return new AlarmStatus(CRITICAL);"
}

Update alarm by ID

PUT /entities/{entityId}/alarms/{alarmId}

Update an alarm by using a valid set of attributes from the Alarms attributes table.

The following table shows the possible response codes for this operation:

Request

The following table shows the header parameters for the request:

Example Update alarm by ID: JSON request

{
"criteria": "if (metric[\"average\"] < 100) { return new AlarmStatus(OK); } return new AlarmStatus(WARNING);"
}

Response

This operation does not return a response body.

Delete alarm by ID

DELETE /entities/{entityId}/alarms/{alarmId}

Delete an alarm from your account.

The following table shows the possible response codes for this operation:

Request

The following table shows the header parameters for the request:

Note

This operation does not accept a request body.

Response

This operation does not return a response body.