First, give conclusions that can be used for decision-making
Technical monitoring is about time outages, error codes, delays and calls, operational monitoring is about whether records are stored, consistent in their state, and whether the amount is equal to the amount. Retesting is about matching the tatter, etc., to avoid creating duplicate orders for the same request. Unable to process anomalies require a clear demonstration of the cause, impact and operation suggestions.
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.
Design a unified tracking number and structured log for interfaces and business objects.
Validation Key Dependence
Set targets for availability, delay, error rate, backlog and variance.
Development of assessable outcomes
(d) The realization of the entropy, the limited re-test, the death letter and the manual compensation.
Make sure you decide the next step with the real results.
Regular reconciliation and validation of whistle-blowers and recovery processes through exercises.
How do you understand it in the actual business?
The payment platform is sent over time and may be duplicated if the order service is not sorely in place. The difference between the slip and the amount is detected using the payment order number and the daily reconciliation of the platform transaction with the local order.
The easiest pit to step on.
Only HTTP 200, no business status check.
Unlimited automatic retesting leads to duplicate data or avalanches
The alarms have no business context and the personnel cannot handle it quickly
How should we end up receiving and confirming?
Receiving and inspection is performed by injection of time overtime, duplication, disorder, non-availability of targets and data discrepancies, confirming that tracking, alarm, retesting, compensation, reconciliation and manual processing are valid and can be done for statistically long periods of time.
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.