Hi gurus. I need to archive the table CDHDR from my production any valuable suggestions from your side

[former_member432274]

I got these sap notes 1257133, 1591251, 1291342. I couldn't understand from where I need to start. I know we can archive data from SARA and taking the object as CHANGEDOCU. But these tables are too big. It almost contains 6billion of data.

I got permission to do it in sandbox environment before doing it in production. Thanks in advance.

[former_member432274]

This message was moderated.

[paul_power]

Hi Sowjanya,

If you archive with Objectclas = * no index is used and the search for

the data to be archived taeks very long.

The best way to do it is, first determine the existing objectclas in the

table CDHDR using the transaction TAANA and more important how many

entries has each of them. You should also create an index over

OBJECTCLAS and UDATE in CDHDR.

In the case you have many entries with few positions, then it make sense

to adopt the strategy described in the note 1639120, but this note

does not work when OBJECTCLAS = *.

Some further proposals for improving the
performance of the archiving with CHANGEDOCU:

. Create a secondary index on table CDHDR for the fields
  MANDANT, OBJECTCLAS and UDATE (see also note 573864) if one does not already exist

. Delete the change documents from the data base tables after they
  have been archived.

. Deactivate single data processing with report
  RSSCD_CUST_OBJECTS_ARCH for change document object you want to archive
  This takes effect only in archiving runs if documents for exactly
  one document object are archived

Aside from this, as there are a huge amount of records to be archived, there is not much else that can be done unfortunately. They should be archived as soon as possible, so that the entries

in the changedocument tables don't increase so much and build to such an amount.

Hope this helps.

Regards,

Paul

[former_member432274]

Hi Paul,

Thank you for your advice, i just run the transaction TAANA for CDHDR i got the no. of entries for the change doc. object.

They are many entries in change doc. object. like MATERIAL, HANDL_UNIT, LIEFERUNG and etc. These having no. of entries more.

So can i select these and can start my archiving stuff.

As you said to create an index for OBJECTCLAS and UDATE in CDHDR, if i create an index it should come from development system  to sandbox system as a transport.

Please, advice. Thanks in advance  

[former_member432274]

For CHANGEDOCU how archiving process done? By objects? How can we determine which functional area is affected by an object?

We néed to archive the related objects also. How can I find the object is related to other object