Data, Processes, and Events: Keys to Comprehensive Analysis - High-Performance Web Databases: Design, Development, and Deployment - page 52

The system also tracks customers who have outstanding
invoices. When an invoice is paid in full, the occurrence of that invoice is
deleted from the INVOICE entity, and the CUSTOMER entity is updated to
reflect the invoice-number date, and amount of the last invoice paid in full.
Deletion in INVOICE causes an update to CUSTOMER.

Delete-Update.

There are cases, of course, when the deletion of an occur-
rence of an entity causes the deletion of one or more occurrences of
another entity. For example, if an instance of ORDER is deleted, all related
LINE ITEM instances must also be deleted.

Delete-Delete.

The Basis of the Event-Driven Approach

The nine types of slightly coupled relationships are the basis of the
event-driven approach. This approach makes use of the ELH diagram.
Unfortunately, the ELH diagram as it has been defined does not provide the
richness and vigor required of an effective analysis tool. Although it can
show changes, it cannot show what causes the change and how the change
affects other entities.

The Four Types of Events

An event triggers a change in life stage of one or more entities. Based on
this definition, there are only four categories of events.

1. Outside Influence.

Exhibit 6 shows a simple ELH diagram for the ORDER
entity of the example order entry system. The ORDER entity's first stage of
life is the creation or addition of a new occurrence of the entity. This first
stage is triggered by an outside influence. A new occurrence of ORDER
would not have been created without the outside influence of the customer
placing an order; this is modeled by Is Placed By relationship in the ERD in
Exhibit 5. Outside influence is the most easily identified type of event.

Exhibit 4-6. Entity-life history of the ORDER entity.