From gh7896

Below is the update for: 7.4.5.2. JSON
— Existing —
Note:
• JSON plain text format is preferred for moderate volume data sets (option 1), as JSON has the advantage of having well-understood simple processing and being human-readable without additional decoding. Examples of moderate volume data sets include the fault alarms and performance alerts, heartbeat messages, measurements used for xNF scaling and syslogs.
• Binary format using Avro is preferred for high volume data sets (option 2) such as mobility flow measurements and other high-volume streaming events (such as mobility signaling events or SIP signaling) or bulk data, as this will significantly reduce the volume of data to be transmitted. As of the date of this document, all events are reported using plain text JSON and REST.
• Avro content is self-documented, using a JSON schema. The JSON schema is delivered along with the data content (http://avro.apache.org/docs/current/ ). This means the presence and position of data fields can be recognized automatically, as well as the data format, definition and other attributes. Avro content can be serialized as JSON tagged text or as binary. In binary format, the JSON schema is included as a separate data block, so the content is not tagged, further compressing the volume. For streaming data, Avro will read the schema when the stream is established and apply the schema to the received content.
In addition to the preferred method (JSON), content can be delivered from xNFs to ONAP can be encoded and serialized using Google Protocol Buffers (GPB).
— End - Existing —

• • • Update ***

NOTE:
• JSON plain text format is preferred for moderate volume data sets (option 1), as JSON has the advantage of having well-understood simple processing and being human-readable without additional decoding. Examples of moderate volume data sets include the fault alarms and performance alerts, heartbeat messages, measurements used for VNF or PNF scaling and syslogs.
• Binary format using Avro is preferred for high volume data sets (option 2) such as mobility flow measurements and other high-volume streaming events (such as mobility signaling events, mobility trace data or SIP signaling) or bulk data, as this will significantly reduce the volume of data to be transmitted. As of the date of this document, all events are reported using plain text JSON and REST.
• Avro content is self-documented, using a JSON schema. The JSON schema is delivered along with the data content (http://avro.apache.org/docs/current/ ). This means the presence and position of data fields can be recognized automatically, as well as the data format, definition and other attributes. Avro content can be serialized as JSON tagged text or as binary. In binary format, the JSON schema is included as a separate data block, so the content is not tagged, further compressing the volume. For streaming data, Avro will read the schema when the stream is established and apply the schema to the received content.
• In addition to the preferred delivery format (JSON), content delivered from VNFs or PNFs to ONAP can be encoded and serialized using Google Protocol Buffers (GPB). Please refer to the next section of this document for additional information.

• • • End - Update ***