SAS RRM symbolic link

[jbkm]

Hi all,

As you know when using SAS RRM (on SAS IRM) datasets in libraries (e.g. ead) are symbolic link pointing to a pool folder.

I have modified a node to generate an extra dataset but that one does not appear in the specified library but I noticed that the other datasets created in that process are symbolic linked.

How can I make my extra dataset symbolic linked too ? Including a param[out] in the sas code didn't help.

Any idea ?

Regards,

JBKM

[zina_stewart]

Hi JBKM -

I do not know the answer to your question - but I am trying to find the answer for you.... 

Thanks!

Zina

[sinpad]

Hi JBKM,

You mentioned you modified one of the nodes to create an extra data set. Are you using the task in a flow? If yes, could you send the flow bpmn and the specific task? You may want to contact Eric Yang for this issue.

Thanks!

[jbkm]

Hi Sinpad,

See in the attachment the bpmn file and the related task. The only additional part in the sas code is:

   /* ML 20180705 Mod 007 - CRE/RRE treatment - debt to service coverage ratio */
   /*- Hash table for debt to service coverage ratio -*/
   data RD_HASH.map_regulatory_rm_dsc_pct (keep = HASH_KEY HASH_VALUE);
  
      length HASH_KEY $10;
      set RD_STAT.mutual_recognition end = lastobs;
      HASH_KEY = kcompress(COUNTRY_CD);
      HASH_VALUE = X_REGULATORY_RM_DSC_PCT;
      output;

      if lastobs then do;
         HASH_KEY = '**OTHER**'; /*i18NOK:LINE */
         HASH_VALUE = 40;
         output;
      end;
   run;  
   /* ML 20180705 Mod 007 */

I create a new mapping table using same principle as for the other mapping tables but somehow my new mapping table (map_regulatory_rm_dsc_pct) is not assigned to library rd_hash due perhaps to symbolic link.

Regards,

[sinpad · Accepted solution]

Hello JBKM,

Thanks for sharing the bpmn and the task code. While you have changed the task code to create the new output, the bpmn does not know about it. So, naturally, you will not see the new output the way you see the others that the bpmn is aware of.

Your modification of an existing task in an already deployed Federated Area (FA) violates the IRM principle that once deployed an FA is immutable except for the landing area where data can be updated through Live ETL.

Please note that once you modify a task signature (in this case, creating a new output), you have essentially created a new task. The correct way of doing this change will be to have the new task in your Personal Federated Area (PFA), use the Scripting Client to create job-flow script, execute the job-flow script (say, in SAS Studio), inspect the created instance that everything looks correct and then have this new task and the created bpmn (generated when you execute the job-flow script) copied over into a new FA. Why a new FA? Because once deployed an FA is immutable as mentioned above.

If you do want to make this change in the existing FA, you will have to first make sure that you have deleted ALL instances of the job flow, replace the current bpmn with the new bpmn and then create an instance of the job flow.

Hope this helps!

[zina_stewart]

Hi JBKM --

Following up to see if you received an answer to your question?  If not, and you need immediate assistance, please contact technical support.

Kind regards,

Zina