Application Development Discussions
Join the discussions or start your own on all things application development, including tools and APIs, programming models, and keeping your skills sharp.
cancel
Showing results for 
Search instead for 
Did you mean: 

Advice on when to archive data that is never really "retired"

Former Member
0 Kudos

My customer has a requirement to change and reprocess old data. I would appreciate some advice on archiving strategy.

One example is agreements in SAP CRM, where up to 10 years old agreements may be changed and this should generate new invoices. On one hand, there is a very large number of agreements (~500.000 agreements created each year), but only a fraction of the agreements need to be accessed and reprocessed more than a few (say 5) years back --> I.e. archiving and deletion of 5+ years old data from the SAP online database seems logical.

On the other hand, the small fraction of 5+ years old data that needs reprocessing still translates to quite high numbers in absolute terms, say in the order of a few thousand agreements per year. This number is high enough for the business to require fully automatic reprocessing capabilities, i.e. changing the old agreements should trigger the same process flows as changing current agreements, invoices should be automatically generated etc...

As a matter of fact, there are also requirements to retain other data (e.g. business partner data) for indefinitely (50+ years) and allow allow changing this data while these changes will/should trigger business processes (that are implmemented in the SAP system) for performing recalculations etc...

I know that with the SAP ADK/ArchiveLink arhiving solution, SAP strongly recommends against reloading archived data into the database for reprocessing, unless the reload takes place immediately after the archiving step.

My question is: With these kind of customer requirements in mind, is it worth to even consider SAPs archiving solution ADK/ArchiveLink and implement some kind of reload strategy? Or is it a definite no-no to do this, and instead fight back all requirements to reprocess old/archived data automatically (i.e. only consider to READ archived data and e.g. manually re-create archived data as NEW objects in the SAP system)?

Are there any other "archiving" solutions that should be considered, other than using ADK/ArchiveLink, that would better fulfill these kind of requirements?

I feel this should be a relatively common requirement these days to be forced to retain data for a long time and also be obligated to change this data, at the same time as performance and offloading the online database is necessary. But I can't find any documentation or advice from SAP about how to normally handle this situation.

Can someone advise on this please?

Thanks,

Lars

1 ACCEPTED SOLUTION

christian_wohlfahrt
Active Contributor
0 Kudos

Hi Lars!

From time to time there will be some changes in customizing. As long as your data (customers) are still in the system, you might update them via mass changes - but surely not when they are 'outside' in an archive.

So after a view weeks / month in archive you can't be sure, if your data still matches the current customizing and can be inserted with a consistent result.

That's a problem of the procedure, so also 3rd party archiving won't help in this case. You have to go for a new creation of your data - and you have to think about a reference data strategy (to add some new additional default values).

If you read / create the data out of SAP's archive or somewhere else - in the end the creation part might anyway the more complex part.

I'm not so sure, if this scenario is so common - but at least easy to understand.

Regards,

Christian

3 REPLIES 3

christian_wohlfahrt
Active Contributor
0 Kudos

Hi Lars!

From time to time there will be some changes in customizing. As long as your data (customers) are still in the system, you might update them via mass changes - but surely not when they are 'outside' in an archive.

So after a view weeks / month in archive you can't be sure, if your data still matches the current customizing and can be inserted with a consistent result.

That's a problem of the procedure, so also 3rd party archiving won't help in this case. You have to go for a new creation of your data - and you have to think about a reference data strategy (to add some new additional default values).

If you read / create the data out of SAP's archive or somewhere else - in the end the creation part might anyway the more complex part.

I'm not so sure, if this scenario is so common - but at least easy to understand.

Regards,

Christian

0 Kudos

Thanks Christian,

I understand the complexities involved with using the "reload" function in SAP archiving, with database inconsistency and database configuration changes over time.

Are you proposing/suggestion to utilize the archive "read" functions, but to build custom programs to recreate (new) objects based on what is read from the archive? Can the read data be accessed and passed on to this custom program?

You said "not so sure, if this scenario is so common", but have you heard of any customers who have done this?

Thanks,

Lars

0 Kudos

Hi Lars!

> Are you proposing/suggestion to utilize the archive

> "read" functions, but to build custom programs to

> recreate (new) objects based on what is read from the

> archive?

Yes, try to use as much existing functions as possible - you don't have to invent the wheel a second time.

> Can the read data be accessed and passed on to this custom program?

Existing function not in the meaning of transaction, more like function module (or class method) -> then it's just the first step in your own program; passing values isn't an issue.

> You said "not so sure, if this scenario is so

> common", but have you heard of any customers who have

> done this?

No, but I'm only working in implementations - never had a customer with such a long SAP history.

Regards,

Christian