Output: CDF-I topics

Agent Events

An Agent Event contains the ID, properties, and time the event was generated.

The ID is decomposed as follows:

• Agent Id
  The identity of the agent that generated the event.

• Item UID
  The identity of the logical item that the event relates to:

  • Flow
    The flow the item is in.

  • Correlation Id
    The unique correlation Id of the item this event relates to or null/unset if this event could not be correlated. Note that correlation ID is only unique within each flow, but the combination of flow and correlation id uniquely identifies an 'Item' across the system.

Timestamps used in the protocol are expressed as nanoseconds since UTC Epoch.

Interval

An interval event is purely a timing between two Agent Events. See below for a description of the JSON schema.

Associations

An Association links two items together.  It is uniquely identified by a combination of:

• Association ID
  The name of the association link: quoteToPriceAssociation for example

• This Item ID
  The ID of the item at ‘this’ end of the association

• Other item ID
  The ID of the item at the ‘other’ end of the association

Alerts

Alerts are identified as follows:

• Definition Id
  The definition ID for this category of alerts: “systemAlert-1” for example.

• Label
  The human readable name of the alert: “System Alert” for example.

• Id
  The unique ID of this alert instance.

• Type
  The type of alert event.

• Status
  An optional traffic-light indicator.

• Timestamp
  When the alert was generated.

• Message
  Describing what happened.

• Properties
  Underlying properties about the alert.

• Formatted Properties
  The above properties after formatting.

• Group
  Optional alert group.

• Config Type
  Optional configuration type.

Timeseries

A Tick Summary is a time-series recording event of a 'double' value over a given period of time.  It is the summary of the recording of a number of physical ticks generated by an underlying source:

• Stream ID
  The unique identity of the stream/source being recorded.

• Timestamp
  The UTC time for the event.

• Period
  The interval over which the recording was made in nanoseconds.

• The following values are recorded during the period:

  • Tick Count
    The number of times the underlying source ticked.

  • Open
    The first value.

  • Close
    The last value.

  • High
    The highest value.

  • Low
    The lowest value.

  • Mean
    The mean of all values recorded during the period.

  • Time-weighted Average
    The time-weighted average of all values recorded during the period.

A Tick value is a time-series recording event of a specific value. This complements the Tick-Summary (above) by storing other values alongside ticking 'double' values.  The types of value currently supported are STRING and URL, but these may be extended at a later date:

• Stream ID
  The unique identity of the stream/source being recorded.

• Timestamp
  The UTC time for the event.

• Period
  The interval over which the recording was made in nanoseconds.

• Type
  The type of the value recorded: STRING or URL. 

• String-Value
  The value held in a String.  Other types may be added in the future.

Configuration

Configuration of the main data types being sent.  The configuration is sent every 60 seconds and can be used to map raw configuration Ids to labels and descriptions.  Please note that data may flow in advance of the configuration being sent.

• timestamp
  The timestamp associated with the configuration, changing when any Velocimetrics configuration has changed. 

• agent_configuration
  A list of agent ids, their flows and labels.

• interval_configuration
  A list of interval ids, their agents and labels.

• flow_configuration
  A list of flows ids and their labels.

• association_configuration
  A list of associations ids, labels and links to flows and agents.

• alert_configuration
  A list of alerts ids, their labels and descriptions.

• timeseries_configuration
  A list of time-series source configurations.  Note that the details here are for the source of the time series than than for the time-series configuration used to capture as multiple time-series configurations can be used to capture the same source:

  • source_id
    This is the first part of the VmxTickSummary.stream_id when splitting the string using ':' as the separator.  For example: "myAggregator:a/b/c:a_column" → "myAggregator".

  • source_label
    The label for the source (usually aggregator name).

  • source_description
    The source description (usually aggregator description).

  • source_column
    A list of time-series columns:

    • value_id
      This is the column id portion of VmxTickSummary.stream_id when splitting the string ':' and taking the last part. For example: "myAggregator:a/b/c:a_column" → "a_column".

    • value_label
      The column label.