In this part we will discuss the loading of data using the NLS Writer. This part is connected only with loading the data into InfoCube. There are 2 ways how to load the data, Ioading via DTP or via InfoPackage
InfoCube loading using DTP
Here is a description of the NLS Writer process for the loading of data into InfoCube using DTP. As the DTP load is done in packages, the data from locked period is checked separately in each load package. If the package contains locked period data and writer is turned on for InfoCubes, this process will pick out the locked records from the package, for the corresponding archiving request a new package is created, in which new records are inserted. Subsequently the rest of the records, which are not in locked period, will be processed using standard way of loading data into InfoCubes.
This process also works for SPO InfoCubes.
InfoCube loading using InfoPackage
Here the Writer process for loading of data using InfoPackage is described. The InfoPackage can be connected to several InfoCubes, therefore the Writer process is connected with transfer rules. This process is similar to loading the data using DTPs. As the loading happens in packages, each data package will be checked for records from locked period. If there are no locked records the process will continue as the standard process. When there is locked data, then the process will generate report similar to the transfer rule, but with Writer functionality included. This process will create a new package in the corresponding NLS Request and records that are in a locked period will be moved into this package. The rest of the data will be forwarded in standard way into the InfoCube.
InfoCube Request Deletion
When loading with Writer is enabled there may be some cases where the deletion of requests may be necessary. Data stored in DB table is updated with this information from loading request. If the user decides to delete some requests, all the necessary information is used to find out which packages were created through the loading of these requests. Afterwards, the process will delete the corresponding packages and the process continues to standard deletion, meaning the data that is not in the archive will be deleted. When deleting data archived with the Writer, the loading request can only be deleted only one by one and this starts with the newest request. If it were possible to delete older requests first, it can lead to the problem that packages from archive request would be inconsistent. Therefore causing a problem during the reloading of the archive.