Add supplementary material as documentation for distributed data synchronization in ONE Record

[ddoeppner]

Supplementary Material: Data Synchronisation in ONE Record

With the ongoing implementation of ONE Record, which provides a robust framework for data sharing within the air cargo industry, building on the idea of decentralized data that stays at the data source and in control of the data holder (see Single Source of Truth and Linked Data design principles), a key challenge has emerged: data synchronization in distributed systems/databases with event-like update mechanisms.

Currently, we manage updates through mechanisms such as ChangeRequests, asynchronous approval and updating, and subsequent notification of applied/discarded changes. However, comprehensive documentation and agreed-upon practices for handling these updates in the context of ONE Record would be highly beneficial.

Example Scenario: When the goods are physically received, a discrepancy is found in the data, i.e., the consignment has 3 pieces instead of 2 and weighs 15kg instead of 10kg. The Ground Handling Agent (GHA) updates the properties of the shipment in its handling information system and requests the change to the shipper's or forwarder's logistics objects via ChangeRequests. The shipper/forwarder then:

• Agrees to the ChangeRequest, resulting in the data being updated in their system.
• Rejects the ChangeRequest, requiring the data in the GHA system to be rolled back to the previous status.

While similar scenarios can occur in other contexts with distributed databases, the challenge is not new. Nevertheless, documenting and establishing a general agreement on handling this within ONE Record is crucial.

Topics and Concepts for Discussion and Exploration:

The list below is not exhaustive but highlights key areas that require attention:

Design Principles:

• Eventual consistency / optimistic replication
• Event-driven architecture

Implementation Alternatives:

• Command-Query Responsibility Segregation (CQRS)
• Saga pattern
• Event sourcing
• Idempotency
• Retry and backoff mechanisms

Automation:

• Automation of ActionRequests (rule engines, workflow systems)

Benefits:

Addressing these topics in the ONE Record API specification will provide clear guidance on handling data synchronization challenges. This, in turn, will lead to more efficient and reliable operations within the digital logistics ecosystem, fostering improved collaboration and data integrity across the air cargo industry.

---

Please review and let us know if any further adjustments are needed.

[aloccid-iata]

Hi Daniel, I am fine with adding it as supporting material. Do you want to enter more in details in the implementation? Best, Davide

[ddoeppner]

Thank you, Davide.

Let's share the information and see if others - who are also currently dealing with these topics - have additional requirements and information demands for such a documentation extension; before we further elaborate and structure the solution.

[aloccid-iata]

I can add it to the next API call