Metric binding to resources

  • Release version: Xanadu
  • Updated August 1, 2024
  • 3 minutes to read
    • Summarize
    Summarized using AI
    This content was generated using new OpenAI-powered functionality. Results are provided on an as is basis and are not guaranteed to be accurate or complete.

    Summary of Metric binding to resources

    Metric binding to resources in ServiceNow Metric Intelligence enables more granular and efficient monitoring by associating metric data not only with Configuration Items (CIs) but also with specific resources within those CIs, such as disks, web pages, or network interfaces. This approach enhances metric event binding by allowing detailed modeling, storage, aggregation, and querying of metrics at the resource level, providing deeper insights especially when multiple similar resources exist within a CI.

    Show full answer Show less

    Key Features

    • Two Binding Methods:
      • CI/Metric format: Binds metrics directly to CIs (e.g., winserver001/responsetimemean).
      • CI/Resource/Metric format (Resource Binding): Binds metrics to specific resources within a CI (e.g., winserver001/Disk — C/diskusage), allowing finer granularity.
    • Resource Binding Process: The MID Server generates metric binding events, which the instance processes by identifying the CI and then attempting to bind the metric to a resource. If the resource does not exist, it is created automatically.
    • Configuration: Enable resource binding by setting the sa.metric.use.resource.binding system property to true. Populate the resourcepath attribute either via event rules or as part of the CI identifier sent to the MID Server.
    • Resource and CI Class Mappings: Uses the CI Type To Resource Class [sacitypetoresourceclass] table to map CI classes to corresponding resource classes. This mapping is critical for correct resource binding and can be adjusted by users with the evtmgmtadmin role.
    • Resource Tables: Metric Intelligence provides a hierarchy of resource tables extending from the parent CI Resource [ciresource] table, including hardware, application, service, VM object, and database resource tables.

    Key Outcomes

    • Enables detailed monitoring and analysis of metrics for individual resources within CIs, improving accuracy and relevance of metric aggregation.
    • Supports use cases where monitored entities are not present in the CMDB, such as individual web pages or network interfaces, allowing meaningful metric aggregation across similar resource types.
    • Improves metric data modeling and querying capabilities at a more granular level, facilitating better operational insights and decision-making.

    Bind metrics to resources to simplify metric events binding by enabling binding to resources such as specific disks or web pages, in addition to binding to CIs.

    Metric Intelligence models metric series in either of the following methods:
    • Binding a metric series to a CI and to the metric being monitored for that CI using a 'CI/Metric' format. For example, 'win_server_001/response_time_mean'.
    • Binding a metric series to a CI, a resource within that CI, and the metric being monitored for that resource, using a 'CI/Resource/Metric' format. For example, 'win_server_001/Disk — C/disk_usage'. This method is referred to as resource binding.
    The first method enables modeling, storing, aggregating, and querying data at the CI level. However, resource binding creates resource records for specific monitored entities, such as an individual web page or a disk. Therefore, resource binding is more efficient when there are multiple resources of similar types within a CI, and metrics applicable to a category of these resources are mapped to these resources.
    Use cases:
    • Common examples are disks, processors, and network interfaces. In cases in which these entities are being monitored but do not exist in the CMDB, using resources for metric binding is useful.
    • Some monitoring solutions capture metric data within services such as KPIs for individual web pages in an application. In such cases where the entity being monitored is not a configuration item, metric binding to resources can be helpful.
    In those situations, using resource binding results in more meaningful aggregations across similar metrics (such as: avg disk_usage for a host across all disks).

    Resource binding process

    The MID Server generates metric binding events that are processed by the instance. When processing a metric binding event, there is an attempt to identify the CI to which this metric binding event belongs. If this attempt is successful, then the metric binding event is bound to the identified CI and binding to a resource is attempted. The following steps describe the attempt to bind a metric binding event to a resource:
    1. Identify the CI class of the CI that was bound to the metric binding event.
    2. Locate the resource class which is mapped to that CI class (using the CI Type To Resource Class [sa_ci_type_to_resource_class] table).
    3. Read the resource_path attribute value in the additional_information field in the metric binding event.
    4. Check if a resource record exists in the resource class table, in which name is equal to resource_path and cmdb_ci is equal to the CI that was bound to the metric binding event.
    5. If such resource record exists, then the metric binding event is bound to that resource. Otherwise, a new resource record is created with the preceding values and the metric binding event is bound to the newly created resource.

    Configure resource binding

    • Ensure that the sa.metric.use.resource.binding system property is set to true (default).
    • For series intended to be bound to resources, ensure that the resource_path attribute is populated by doing either step:
      • Use an event rule to add the resource_path attribute to the Additional information field in events. For more information, see Create an event rule to map metrics to specific CIs.
      • Populate the resource_path attribute as a part of the respective CI identifier when data is sent to the MID Server for processing.
    • Review the default mappings in the CI Type To Resource Class [sa_ci_type_to_resource_class] table and adjust as needed. It is critical that mappings are set as desired prior to data processing.

      Managing the CI Type To Resource Class table requires the evt_mgmt_admin role.

    Resource tables

    Resource binding uses an underlying hierarchy structure of resources which is a subset of the [cmdb_ci] hierarchy. Metric Intelligence installs the following resource tables:
    • CI Resource [ci_resource] (parent table):
    • Tables that extend CI Resource [ci_resource]:
      • ci_resource_hardware
      • ci_resource_appl
      • ci_resource_service
      • ci_resource_vm_object
      • ci_resource_database

    Mapping CIs to resources

    Mappings of CI classes to resource classes are stored in the CI Type To Resource Class [sa_ci_type_to_resource_class] table. This table is installed with Metric Intelligence, and is used during metric binding to resources.

    Table 1. Mappings in the CI Type To Resource Class table (default)
    CI class Resource class
    cmdb_ci_hardware ci_resource_hardware
    cmdb_ci_appl ci_resource_appl
    cmdb_ci_service ci_resource_service
    cmdb_ci_database ci_resource_database
    cmdb_ci_vm_object ci_resource_vm_object