Thresholds

Thresholds is a standard configuration interface for adding, editing, and removing thresholds in Oracle Communications Unified Assurance.

For information about interacting with the grid and form, see Standard Configuration Interface in Unified Assurance Concepts.

This user interface calls REST methods from api/metric/thresholds. See REST API for Unified Assurance Core for details.

To open this interface, from the main navigation menu, select Configuration, then Metrics, then Thresholds, and then Thresholds.

Form Fields

The following fields appear in the form for this UI:

• Name: The name of the threshold.

• Type: The threshold type. Options are:

  • Abnormal: Uses the Abnormal Thresholding Engine.

  • Missing Data: Uses the Missing Data Threshold Engine.

  • Standard: Uses the Standard Thresholding Engine.

  • Trend % Change: Trend Threshold Engine.

  • Trend Prediction: Trend Threshold Engine.

• Abnormal Profile: Appears when Type is set to Abnormal. The abnormal profile to use for the confidence band calculation and threshold.

• Measurement: The metric type that the threshold checks.

• Metric Field: The data type that the threshold checks. Must be set to one of the following:

  • availability

  • utilization

  • value

  Note:

  You can also optionally use a Telegraf metric fieldKey. In this case, the Value in the notification is assumed to be a percentage. To see the keys, make the following queries against the Metric schema using the Database Query Tools UI:

  show field keys from telegraf_mem;
  show field keys from telegraf_swap;
  

  When Type is set to Abnormal, this field is ignored and value is always used.

• Time Range: Unless Type is set to Abnormal, the period of time in which data points are retrieved and evaluated. For example, 15m.

  Must be a combination of a numerical value and one of the following duration values:

  • m: Minutes

  • h: Hours

  • d: Days

  • w: Weeks

  • M: Months

  • Y: Years

  When Type is set to Abnormal, this field is ignored and 14d is always used.

• Forecast: Appears when Type is set to Trend Prediction. The relative time, starting now, to use when predicting when the metric value will breach the threshold.

  The format is a combination of a numerical value and one of the following duration values:

  • m: Minutes

  • h: Hours

  • d: Days

  • w: Weeks

  • M: Months

  • Y: Years

  For example, use 6M to notify when the threshold is expected to be breached within six months.

• Frequency: Appears when Type is set to Standard and Check Location is set to Threshold Engine. How often the threshold is checked.

  The format is a combination of a numerical value and one of the following duration values:

  • m: Minutes

  • h: Hours

  • d: Days

  • w: Weeks

  • M: Months

  • Y: Years

  For example, use 5m to check every five minutes.

  Note:

  Use the following guidelines when setting the threshold frequency:

  • Because thresholds can only be checked as frequently as the Standard Thresholding Engine polls, you must set this to a value greater than or equal to the PollTime of the Standard Thresholding Engine (the default is one minute).

  • Because metric values are only updated when the metric is polled, you must set this to a value greater than or equal to the PollTime of the metric the threshold is for (the default is five minutes).

  See Coordinating Poll Times in Unified Assurance Implementation Guide for more information.

• Offset: Appears when Type is set to Standard and Check Location is set to Threshold Engine. The offset to apply to the time range.

  The format is a combination of a numerical value and one of the following duration values:

  • m: Minutes

  • h: Hours

  • d: Days

  • w: Weeks

  • M: Months

  • Y: Years

  For example:

  • Use 5m to go back five minutes when checking the threshold to ensure most recent poll data is considered.

  • Use 0m for no offset.

• Function: Appears when Type is set to Standard. The function to apply when calculating if a threshold is breached. Expands from previous Mean. The following options are available:

  • 95th Percentile: Use when you do not want to count infrequently-occurring high outliers and odd values, when most data points will be under a certain value.

  • Count: Use to ensure there is no missing data, when you expect data points frequently. For example, if you expected a data point every minute, checking a TimeRange of X minutes every Y minutes shows if data is missing after the last Y minutes.

  • Maximum: Use to ensure all data is below a certain value.

  • Mean: Use to consider the average of the values.

  • Minimum: Use to ensure the data never falls below a certain value.

  • Standard Deviation: Use when the variation of data points should be below a certain value.

• Warning: When to send a threshold warning notification. When Type is set to Abnormal, only one notification level is supported. If you set both, the values for Critical will be used.

  • Operator: The threshold comparison operator. Although the UI always requires an operator, thresholds with the Missing Data, Trend Prediction, or Abnormal type do not use it.

  • Value: The value to compare against. Although the UI always requires a value, thresholds with the Missing Data or Abnormal type do not use it.

    The type of value depends on Metric Field:

    • When Metric Field is set to value, this is a raw value.

    • When Metric Field is set to anything else, including a Telegraf field, this is a percent value from 0 to 100.

  • Severity: The severity for the event on the event list.

• Critical: When to send a threshold critical notification. When Type is set to Abnormal, only one notification level is supported. If you set both, the values for Critical will be used.

  • Operator: The threshold comparison operator. Although the UI always requires an operator, thresholds with the Missing Data or Trend Prediction type do not use it.

  • Value: The value to compare against. Although the UI always requires a value, thresholds with the Missing Data type do not use it.

    The type of value depends on Metric Field:

    • When Metric Field is set to value, this is a raw value.

    • When Metric Field is set to anything else, including a Telegraf field, this is a percent value from 0 to 100.

  • Severity: Unless Type is set to Abnormal, the severity for the event on the event list.

• Message: The string to use in events generated for threshold violations. Populates the EventType and Summary fields for the event as follows:

  • EventType: Populated with the string following a colon in the message. If there is no colon, or nothing after the colon, the threshold Name value is used instead.

    For example, if the Message is Performance threshold violation: Ping Jitter High, EventType will be Ping Jitter High.

  • Summary: Populated with the entire message, and the value or percentage.

  The rules that set these fields are in the standard_threshold.rules file in the Rules UI.

• Check Location: Used for standard thresholds only. What checks the threshold, an application or the Standard Thresholding Engine. Although you can change the value for other threshold types, the UI validation fails when you click Submit.

• Status: Whether the threshold is enabled or disabled.

---

Title and Copyright Information

User’s Guide

G36447-01

Copyright © 2025, Oracle and/or its affiliates.