Different Point Consumption for Bound Comparison Tasks
First, we have to talk about the process of issuing binding comparison tasks. First, we will require unified monitoring frequency and Probe Group, as well as validity period and browser kernel. Otherwise, tasks cannot be bound to each other. When the client retrieves the Number of Requests data, our delivery scheduling system will deliver the two bound tasks to the client at once. In this way, the client will perform two tasks in succession.
At this point, I haven’t said why the data volume is different. The reason for the problem lies in the postback of client data. During the return process, not all data may be "safely" included in the database. There are several possibilities that may lead to failure in the database entry.
1.
The missing amount is a numerical value, just like the literal meaning. The scheduler delivered the task normally but did not collect the data returned by the client within the task time. This part Probe will cause the data amount of the comparison task to be different.
2.
If the collected data is noisy data, it cannot be stored in the database. Noisy data, for example: DNS does not match (because we only verify the primary DNS, if this task uses a secondary DNS and does not match, it will cause the database to be killed).