First, give conclusions that can be used for decision-making
For each business event, a stable unique identifier is assigned and pre-writing or record-processing status is obtained. Limited numbers and indices of errors can be avoided for connections overtime, temporary restricted streams, etc. The missing fields, inadequate privileges and business rules are rejected, not automatically repeated, but into readable abnormal queues. Cross-system processes are required to record every step of external numbering, summary requests and results, with partial success continued, cancelled, manually confirmed or completed on the basis of business decisions. Compensation is not necessarily a technical reverse operation, such as the inability to actually recover mail already sent, which may require correction and notification to the responsible person.
What conditions need to be identified before judgement is made?
The same question may have different answers under different business, data and project phases. It is suggested that the following conditions be checked and that the common findings on the web be incorporated into their own projects.
Suggested order of advance
First, we'll be clear about the target and the border.
Creates the only key for events, business objects and each written action.
Validation Key Dependence
Retry, stop, compensate or queue manually according to the wrong type of configuration.
Development of assessable outcomes
Saves the step status, external numbering, input summary and cause of error.
Make sure you decide the next step with the real results.
Final consistency verified through failure injections and time reconciliations.
How do you understand it in the actual business?
The system cannot be created immediately, but the client order number should be used to query whether the ERP has been recorded; if it already exists, follow up and if it is confirmed that it does not exist, try again. If the ERP succeeds, but the CRM update fails, the ERP order number should be retained and the CRM should be completed instead of rolling back or repeating the entire process. The examples do not represent the performance of a particular client, and the actual conclusions need to be verified in conjunction with the enterprise’s own business volume, sample, system and liability boundaries.
The easiest pit to step on.
Unlimited retry for all nodes
Only technical errors recorded, no business objects and no external numbers
Partly successful, running back from head to head.
How should we end up receiving and confirming?
The tests should proactively create repeat events, time overruns, restricted flow, inadequate authority, field errors and partial success, and validate that duplication of business records does not occur; anomalies can enter the correct queue, the alarms contain information that can be processed and the compensation and reconciliation results are supported by audit evidence.
When preparing to communicate with suppliers or internal teams, it is recommended that current processes, representative samples, existing systems, planning time and budget levels be brought. First, the unknown items are clearly marked, and then the decision is made to use diagnostics, PoC, fixed-range projects or ongoing research and development, which is usually more reliable than a direct demand for a price and duration without borders.