[
        {
            "description": "OperatingSystemCPUSystemUtilization",
            "driver": "PrometheusDatabaseDriver",
            "labelNames": [
                "BaremetalVMUuid",
                "CPUNum"
            ],
            "name": "OperatingSystemCPUSystemUtilization",
            "namespace": "ZStack/BaremetalVM"
        },
        {
            "description": "OperatingSystemCPUUserUtilization",
            "driver": "PrometheusDatabaseDriver",
            "labelNames": [
                "BaremetalVMUuid",
                "CPUNum"
            ],
            "name": "OperatingSystemCPUUserUtilization",
            "namespace": "ZStack/BaremetalVM"
        }
]

• namespace: The namespace of a metric. A namespace is used to classify metrics of the same type.
• name: The metric name.
• labelNames: The label name of a metric. A label is used to filter metric data. For example, the DataVolumeCount label indicates the number of volumes on a primary storage. You can specify the PrimaryStorageUuid label to filter data so as to obtain the number of volumes on a primary storage.

{
    "labels": [
        {
            "CPUNum": "0",
            "HostUuid": "a133cbcfdc1d462c9b51d7bce6ac6b11"
        },
        {
            "CPUNum": "1",
            "HostUuid": "a133cbcfdc1d462c9b51d7bce6ac6b11"
        },
        {
            "CPUNum": "2",
            "HostUuid": "a133cbcfdc1d462c9b51d7bce6ac6b11"
        },
        {
            "CPUNum": "3",
            "HostUuid": "a133cbcfdc1d462c9b51d7bce6ac6b11"
        },
        {
            "CPUNum": "4",
            "HostUuid": "a133cbcfdc1d462c9b51d7bce6ac6b11"
        },
        {
            "CPUNum": "5",
            "HostUuid": "a133cbcfdc1d462c9b51d7bce6ac6b11"
        },
        {
            "CPUNum": "6",
            "HostUuid": "a133cbcfdc1d462c9b51d7bce6ac6b11"
        },
        {
            "CPUNum": "7",
            "HostUuid": "a133cbcfdc1d462c9b51d7bce6ac6b11"
        }
    ],
    "success": true
}

{
    "data": [
        {
            "labels": {
                "HostUuid": "a133cbcfdc1d462c9b51d7bce6ac6b11"
            },
            "time": 1516363957,
            "value": 1678.9
        },
        {
            "labels": {
                "HostUuid": "a133cbcfdc1d462c9b51d7bce6ac6b11"
            },
            "time": 1516364017,
            "value": 1920.1000000000001
        }
    ],
    "success": true
}

• In the preceding sample, value is a data value of the double type, time is an epoch time (in seconds), and labels is a label group that can be used to filter response values. The following is an example of obtaining data of the host with the UUID a133cbcfdc1d462c9b51d7bce6ac6b11:

  GetMetricData namespace=ZStack/Host metricName=NetworkAllOutByte labels=HostUuid=a133cbcfdc1d462c9b51d7bce6ac6b11

• A label supports two filter methods: = (equal to) and =~ (regular expressions). You can use regular expressions to filter multiple label values. The following is an example of returning the metric values for the hosts with UUIDs a133cbcfdc1d462c9b51d7bce6ac6b11 and 506bd3b46a6543718488b5ba5850aaaa, respectively.

  GetMetricData namespace=ZStack/Host metricName=NetworkAllOutByte labels=HostUuid=~a133cbcfdc1d462c9b51d7bce6ac6b11|506bd3b46a6543718488b5ba5850aaaa

• You can transfer multiple labels as filter conditions. For example,

  GetMetricData namespace=ZStack/PrimaryStorage metricName=SnapshotCount labels=PrimaryStorageUuid=b632652cc16044cdb6b4f516ed93a118,PrimaryStorageType=Ceph

• You can use startTime and endTime to specify the time range for query. For example,

  GetMetricData namespace=ZStack/PrimaryStorage metricName=SnapshotCount labels=PrimaryStorageUuid=b632652cc16044cdb6b4f516ed93a118,PrimaryStorageType=Ceph startTime=1516364017 endTime=1516365000

  Note: Both startTime and endTime are epoch time. If you leave these two fields blank, data of the latest 1 minute will be returned by default. If you leave the endTime field blank, the current time will be used as the end time by default.
• You can use period to specify the interval for returning data.
• You can use functions to transfer a pair of functions for data post-processing. For example,

  GetMetricData namespace=ZStack/PrimaryStorage metricName=SnapshotCount labels=PrimaryStorageUuid=b632652cc16044cdb6b4f516ed93a118,PrimaryStorageType=Ceph startTime=1516364017 endTime=1516365000 functions=sort(by=\"value\"\\,direction=\"desc\")

• Here, the sort function is used to sort the returned value in descending order. If you transfer multiple functions, these functions will be executed in ascending order. For example,

  GetMetricData namespace=ZStack/PrimaryStorage metricName=SnapshotCount labels=PrimaryStorageUuid=b632652cc16044cdb6b4f516ed93a118,PrimaryStorageType=Ceph startTime=1516364017 endTime=1516365000 functions=sort(by=\"value\"\\,direction=\"desc\"),limit(limit=1)

  Note: In zstack-cli, you need to translate the double quotation marks (") in a function by using backward slashes (\), and translate the commas (,) for delimiting function parameters by using double backward slashes (\\).
• The system firstly sorts the data, and then returns the first data.
  Note: For more information about functions, see the Appendices.

[{
    "metricName": "MySQLMaxConnections",
    "value": 1000,
    "labels": {
        "ip": "10.0.0.10"
    }
}]

• metricName: The metric name.
• value: The data value, which can be an integer or a floating-point number.
• labels: The custom labels.

[{
    "metricName": "MySQLMaxConnections",
    "value": 1000,
    "labels": {
        "ip": "10.0.0.10"
    }
}, {
    "metricName": "MySQLMemory",
    "value": 1999999999999,
    "labels": {
        "ip": "10.0.0.10"
    }
}]

Note: The namespace and metric are automatically created when data is stored. No API is available to delete namespace and metric. The related data will be automatically deleted after its Time To Live (TTL) is exceeded.

{
    "events": [{
            "description": "VMStateChangedOnHost",
            "labelNames": [
                "OldState",
                "NewState",
                "SourceHostUuid",
                "DestinationHostUuid"
            ],
            "name": "VMStateChangedOnHost",
            "namespace": "ZStack/VM"
        },
        {
            "description": "VRouterDisconnected",
            "labelNames": [
                "Error"
            ],
            "name": "VRouterDisconnected",
            "namespace": "ZStack/VRouter"
        }
    ],
    "success": true
}

• namespace: the namespace of the event
• name: the event name
• labelNames: the names of event labels

{
    "events": [
        {
            "emergencyLevel": "Normal",
            "labels": {
                "NewStatus": "Connecting",
                "OldStatus": "Connected"
            },
            "name": "HostStatusChanged",
            "namespace": "ZStack/Host",
            "resourceId": "184eae9058f9477a91975f549b54f59f",
            "resourceName": "Host-1",
            "time": 1510984704006
        },
        {
            "emergencyLevel": "Normal",
            "labels": {
                "NewStatus": "Disconnected",
                "OldStatus": "Connecting"
            },
            "name": "HostStatusChanged",
            "namespace": "ZStack/Host",
            "resourceId": "184eae9058f9477a91975f549b54f59f",
            "resourceName": "Host-1",
            "time": 1510984704633
        }
    ],
    "success": true
}

• emergencyLevel: the emergency level, including Major, Info, and Emergent
• name: the event name
• namespace: the namespace of the event
• resourceId: the ID of the resource where the event occurred. If the resource belongs to ZStack Cloud, the ID is the UUID of the resource.
• resourceName: the name of the resource where the event occurred
• time: the time when the event occurred

{
  "inventory": {
    "uuid": "a6bed2b7c5154371a384f0f90dcd227e",
    "name": "VM CPU Alarm",
    "comparisonOperator": "LessThanOrEqualTo",
    "period": 60.0,
    "namespace": "ZStack/VM",
    "metricName": "org.zstack.zwatch.datatype.Metric@22c0dd63",
    "threshold": 30.0,
    "repeatInterval": 1800.0,
    "repeatCount": -1.0,
    "status": "Alarm",
    "state": "Enabled",
    "createDate": "Jan 8, 2019 4:08:43 PM",
    "lastOpDate": "Jan 8, 2019 4:08:43 PM",
    "labels": [
      {
        "uuid": "c74cb057092243ea8f838a1b4caec128",
        "key": "VMUuid",
        "operator": "\u003d",
        "value": "b5defee7d82f44c692d31e9a9d9e8e7f"
      }
    ],
    "actions": [
      {
        "alarmUuid": "433a94c9da494565a28dc4736e159297",
        "actionType": "sns",
        "actionUuid": "6d5583e8210445e3963c3784d550b8c2"
      }
    ]
  }
}

• namespace: the namespace of the metric
• metricName: the metric name
• comparisonOperator: the threshold comparison operator, including GreaterThanOrEqualTo (greater than or equal to), GreaterThan (greater than), LessThan (less than), and LessThanOrEqualTo (less than or equal to)
• period: the threshold duration, in seconds.
• repeatInterval: the alarm repeat interval, in seconds. This field is used to repeat alarms in a specified time duration. For example, if the alarm condition is to trigger alarms when the CPU utilization exceeds 90% in 1 minute, but the CPU utilization of the VM instance continuously exceeds 90% for 10 minutes, you will receive 10 alarm messages. If you set repeatInterval to 10 minutes (600 seconds), you will receive only one alarm message. If you set repeatInterval to 5 minutes (300 seconds), you will receive two alarm messages. This helps you avoid alarm flooding. If you do not set the repeatInterval field, the default value is 30 minutes (1800 seconds). You can modify the default value by modifying zwatch.alarm.repeatInterval in global settings.
• threshold: the threshold, which is a floating-point number
• actions: the alarm actions, that is, the actions performed by the system when the alarm is triggered. In the current version, ZWatch can only send alarm messages to SNS topics. The structure of an alarm action is as follows:

  {
  "actionUuid": "2447db65cb46495e87b53d9afbad6936",
  "actionType": "sns"
  }
  

  In the sample above, actionUuid is the UUID of the SNS topic. The value of actionType is sns, which is a fixed value. You can specify multiple alarm actions as needed.

• labels: the labels for filtering metrics. You can use labels to match specific metrics. For example, to create alarms for a specific VM instance, you need to set the VMUuid label.

  {
      "key": "VMUuid",
      "operator": "Equal",
      "value": "1a1d7395cf74474ca52deb80c41214a1"
  }

  In the sample above, key is the label name. operator is the label comparison operator, which can be either Equal (equal to) or Regex(regular expressions). value is the label value. You can specify multiple labels as needed. You can also use GetAllMetricMetadata to obtain labels supported by each metric.

• OK: the normal status, indicating that the alarm is not triggered by the metric data
• Alarm: the alert status, indicating that the alarm is triggered by the metric data
• InsufficientData: the data is insufficient, indicating that the metric data is insufficient

{
    "inventories": [
        {
            "actions": [
                {
                    "actionType": "sns",
                    "actionUuid": "e7d6f5e23bb74e99a2777126078b551c",
                    "alarmUuid": "b632652cc16044cdb6b4f516ed93a118"
                }
            ],
            "comparisonOperator": "GreaterThanOrEqualTo",
            "createDate": "Jun 20, 2018 10:45:45 AM",
            "labels": [
                {
                    "key": "DirPath",
                    "operator": "Equal",
                    "uuid": "dfed592b822f46439195d7ce7728cdeb",
                    "value": "/var/lib/zstack/"
                }
            ],
            "lastOpDate": "Jun 20, 2018 10:45:45 AM",
            "metricName": "ManagementServerDirUsedCapacityInPercent",
            "name": "ZStack Data Directory Capacity Alarm",
            "namespace": "ZStack/System",
            "period": 60,
            "repeatInterval": 1800,
            "state": "Enabled",
            "status": "OK",
            "threshold": 70.0,
            "uuid": "b632652cc16044cdb6b4f516ed93a118"
        }
    ],
    "success": true
}

• Alarm variables
  • ALARM_NAME: the alarm name
  • ALARM_UUID: the alarm UUID
  • ALARMCOMPARISONOPERATOR: the alarm operator
  • ALARM_METRIC: the alarm metric
  • ALARM_NAMESPACE: the alarm namespace
  • ALARM_THRESHOLD: the alarm threshold
  • ALARMLABELS: the alarm label array. You can use *${ALARMLABELS.join(",")}* to combine the labels into a comma-separated string.
  • ALARM_DURATION: the alarm duration when a metric value exceeds the threshold
  • ALARMPREVIOUSSTATUS: the alarm status before the alarm is triggered
  • ALARMCURRENTSTATUS: the current status of the alarm
  • ALARMCURRENTVALUE: the metric value when the alarm is triggered
• Event subscription variables
  • EVENT_NAMESPACE: the event namespace
  • EVENT_NAME: the event name
  • EVENTLABELS: the event label array. You can use *${EVENTLABELS.join(",")}* to combine the labels into a comma-separated string.
  • EVENTEMERGENCYLEVEL: the event emergency
  • EVENTRESOURCEID: the ID of the resource where the event occurred
  • EVENTRESOURCENAME: the name of the resource where the event occurred
  • EVENT_ERROR: If the event is related to an error, such as the network error that resulted in a host disconnection event, this field contains a string that describes the error. If no error exists, the field is an empty string expressed as “”.
  • EVENT_TIME: the time when the event occurred. This variable is a string.
  • PARAMEVENTSUBSCRIPTION_UUID: the UUID of the subscribed event.