Dataflux 2.7 and Encrypted SAS Datasets

[Lenvdb]

Hi all

We are now starting to use the Dataflux Data Quality Management Suite => DM Server + DM Studio.

We have a Metadata server connection and want to set up a few SAS Base Dataset Libraries as Data Sources.

The bulk of our core datasets are Metadata-bound and encrypted. 

I am able to set up a connection to datasets that are not encrypted, but seem to have a problem with encrypted datasets in our Libraries.

I realise that this may not be something everyone is doing, but I will appreciate any help I can get to find out

a) Is it possible for DMStudio to work with Encrypted SAS Datasets (That are also Metadata-bound)

b) And if so, How would I do this?

[LinusH]

I haven't done this from DM suite.
But from a meta data perspective you need to log on to a SAS session with user registered in SAS meta data and with appropriate authorization for the data set.

[Lenvdb]

Thank you Linus

Yes - we have no problems in SAS EG to work with our encrypted / Metadata-bound Libraries.

But the problem is in Dataflux Data Management Server + Studio.

Connections seem only possible to unencrypted datasets.

I was wondering if there is anyone on the planet that may have had success with this?

(We are very security conscious, hence the encryption and Metadata-bound state of our core data)

[nar_sas · Accepted solution]

There are a couple of ways to do this using DM Studio and DM Server.  The BASE driver in DM Studio/Server cannot read them directly, but there are a few options available to get this to work. 

One way is to use SAS Federation Server to access the data and setup a connection to federation server from DM Studio/Server.  Here are instructions on how to configure SAS Federation server connections to metadata bound libraries.  http://support.sas.com/documentation/cdl/en/fedsrvag/70395/HTML/default/viewer.htm#p19idcme77fiomn11...

Another way is to use SAS to reach through to the dataset.  You can add a SAS code node to your Data jobs and in the code use SAS to reach through to the library.  You would then have to process and land the data for DM Studio/Server to read it, but it could only be the required subset that you need for your data management jobs.  

A third way is that  you can call many of the data quality functions available directly from SAS such as dqparse, dqstandardize, and others.  You can also call services in dm server directly from your SAS jobs that perform many data quality functions, which would not require you to move the data to a Data Management instance.  For example, DI Studio has a transform available that will processrows of data in the DM Server. 

Hope this helps.  

[LinusH]

@nar_sas can you comment on the specific situation here about meta bound libraries, and how that would interact with each of your scenarios?