Classification and application of DSO

1, the DSO classification, the standard DSO (generate the main data identification, the same key field values are merged, can be issued directly report), write-optimized DSO (do not generate master data identification, not merge the same key field values, speed can be used to store large-capacity data), directly write DSO, these three kinds of DSO are useful
2, the standard DSO should pay attention to the problem is, if it is to activate the business content of the DSO, and then copied out as their own DSO use, generally do not need to consider the key fields of the problem, the system settings will generally ensure that the DSO data and PSA data exactly the same, but if you are directly input information objects created by the DSO , you need to consider the value of the merger, because the standard DSO for the same key field value system will automatically merge, this time need to pay attention to the conversion of the rule details are overwritten or total, if there is a merged field, that is, the DSO updated data less than the PSA data, it is recommended that the conversion of the value of the field of the rules of To ensure the correct value, in short, the use of standard DSO to consider the merger of the value of the problem, and consider the conversion of the rule details, but the use of write-optimized DSO does not exist the problem
3, recommended if the use of three-tier data Model architecture, you can consider the first layer using write-optimized DSO, to ensure that the layer DSO data and PSA exactly the same, the second layer DSO using standard DSO, the third layer using cube
4, after the discovery in a fully updated data source case, write optimization DSO in the new data, will be repeated calculation of the previously extracted data, it is recommended that the DSO should be built carefully consider the type of DSO, Data Update method, DSO key field and data field settings, and the conversion of the collection in the detail rule set to overwrite or total

Standard: 3 tables in the background, NEW table/log table/active table, that is, the change LOG data saved with delta increments. The data is typically updated with DTP. This DSO is also the most we use the type.
Direct update : Generally used to do certain applications, such as APD. is through the BW business platform, the data through the API way, directly written. The subsequent table only corresponds to the active table.
Write-Optimized : This type of DSO is used to process a large amount of data, but the backend only corresponds to the active TABLE. Use DTP to update data.

Generate master Data ID after activation: When the data is activated, the SID of the master data is generated, which can increase the speed of the report query. This option is generally not selected if the DSO is not directly used for reporting requirements, or if it is used less often.
Unique data record: This option is only available when activate generate master data ID after activation. Identifies that the uploaded data is a primary key unique. In doing so, the speed of activation can be increased, the uploaded data is only done by the insert operation, regardless of the sort and pre-mirroring situation. If we are going to do a delta extraction of the data, and the data may involve modifications, this option will cause an error to occur. It is generally recommended that if the standard DSO selects "Unique Data records", the write-optimized type is used instead.
Automatically set the status to ' good ': After activating the data, if no error occurs, the state is automatically set to OK, that is: green light, generally we select this option.
Automatically activates data/automatically updates data: identifies automatic activation of data and updates the data to targets such as cube. In order to handle anomalies in the processing chain, we do not generally choose here.

Classification and application of DSO