Skip to Content

Loopback Process – Follow up

One of articles on “Loopback Process” was published in the SDN Technical articles section. The link to the article is: I just thought I would give an insight to the article and also add some more information which I had missed out in the article through this Weblog.   In the Loop back process what we essentially do is to create an export data source on the ODS/Cube in which the data is to be updated. Now use the same data source to load data back into the Cube/ODS.   Apart from the advantages mentioned in the article the loopback process would also be beneficial in these cases:  1. Re-init not needed – When any characteristic/Key figure is added and if the loopback process serves our purpose then we can avoid the hassle of doing a initialization load again. This way you could continue with your delta loads.  2. Correct duplicate/incorrect records – Lets say one of the request in a cube is doubled and we are unable to delete the duplicate request (possibly because the requests are compressed). Re: How to delete duplicate data from compressed requests? Typically in that scenario we create a flatfile with negative key figure values to correct the data. The creation of the flatfile is a cumbersome process as we need to create the file manually.   However we can use the loopback process to negate the key figure values of the duplicate request and thereby diminish the duplicate request effect. This is faster and more efficinet as there is no manual process.  The loopback process can also be used to correct incorrect data records in the same way as stated above. The only difference would be that we would need to give the selection parameters in the infopackage for the incorrect records.  Now comes the tricky part. Can the BW experts review the article and give their feedback. Awaiting your response.
You must be Logged on to comment or reply to a post.