Adobe Analytics Classification import via FTP

[nico9]

Hi all, 

got a question on the managing of a FTP connection.

Since the data files we have to import with the Classification Importer (.tab and .fin files) should be removed after the import is processed, i wanted to ask if there is some kind of implementation to do, or it is something Adobe system does automatically.

Thanks!

[Pablo_Childe]

I am unaware of a time of life the ftp files exist. You would typically need to delete after data is processed. In some case can take days if large data sets.

For regular ftps we perform we use same file name then overwrite with new one. this is a sort of auto cleans for us to ensure we don`t leave large data files for long times.

GLTU

[nico9]

Thank you for sharing, it is a good workaround 

[yuhuisg]

You can usually leave them alone. The server belongs to Adobe, so they're paying for the disk space! 

The only thing you should be really concerned about is the relationship between your .tab and .fin files. Let's say you imported File1.tab. After Adobe Analytics finishes processing it, it saves File1.fin at the same folder location, indicating that File1.tab has been processed. Now, if you were to delete File1.fin but leave File1.tab alone, Adobe Analytics could re-process File1.tab and then save File1.fin again. There's no harm done if there have been no changes made to File1.tab, since the same classifications that are specified there would be applied again. But this also means that if you upload a new File1.tab and delete any File1.fin, Adobe Analytics will process your new File1.tab and save a new File1.fin, even though the filename "File1.tab" is the same.

TL;DR use unique filenames.

See @Hyder_Ziaee's reply for the correct relationship between the .tab and .fin files.

[nico9]

Thank you, it is clear

[Hyder_Ziaee · Accepted solution]

File deletion is automatic for classification import via FTP/SFTP server. 

You just need to follow the order, i.e., upload the .tab file first and then upload the .fin file. Once the server uploads the data, it will automatically be deleted from that file server.

Unique names is always preferrable (you can append date and/or time of upload) since it helps when debugging if something goes wrong.