Data Refresh in Data Manager

[former_member480924]

Hi Experts,

In Data Manager there are 1000 records currenlty. In the Source file we have 600 Records which will be imported though "Access" Type. Once we import only those 600 records should exist in Data Manager. The data earlier which was exisitng i.e. 1000 should be deleted without any manual intervention. Is this Scenario can be Built?

The data will get refreshed every day. Everyday the Count Varies.

Could you please guide regarding this.

Thanks & Regards,

U G

[Former Member]

Hi U G,

Yes this can be built if you dont use Key Mapping for the table where you are using this data.

But I wonder why would you do that in a MDM system which is supposed to store less changing,noun like Master data.Rather it would be better to model your data in a way that Main data is not replaced but lookup or Tuple level data is refreshed from some other leading system.

Can you please provide your scenario related details.

Thanks,
Ravi

[former_member480924]

Hi Ravi,

This is for the SRM MDM Catalog. MDM Version is 5.5 SP6. Weekly the data will be refreshed through Import with the "Access" Type. The data existing in the Source file when loaded should exist in Data Manager. Remaining data should be deleted automatically (No Manual Intervention - Deleting records directly in Data Manager which is not relevant).

Can this scenario be built? Weekly the data count will be keep changing as explained earlier. Source file data should be in Sync with Data Manager.

Thanks & Regards,

U G

[Former Member]

HI U G,

There can be multiple ways to achieve this -

1. You can use APIs to first delete the matched records and then create new records.Since you want this process to be automated APIs can be useful here.

2. Maintain your table with no remote keys enabled and replace every time new data comes in.This would only replace the matched records and not entire records.Consider this if it suits your purpose.

3.Maintain Flag = True for latest records and updated others as False.So every record before import which has to exist would have status as True.Set up a job at non-business hours which deletes these records with Flag = False.

Thanks,

Ravi

[Former Member]

Hi UG,

Just a curious question,why don't you use named searches to your advantage here.

Thanks,

Ravi

[Former Member]

Hi,

As far as i know you might not be able to delete existing records automatically unless manually deleted. I am sure you would already be aware that these records after initial update can be segregated and deleted from repository. It would be more easier approach to delete the left out records(400 nos.) once initial update is done. Please let us know if any issues!

Regards,

Ali