RSZDELETE - quick and dirty?

[Former Member]

Hello to the community.

I'm testing the t-code RSZDELETE at the moment and facing the following problem.

When deleting a query via RSZDELETE with transport connection, the transport request only contains the UID of the query itself but none of its 'existentially required' query elements. For example local key figures etc. Not global key figures oder structures.

So the tables (RSZELTDIR, RSZELTTXT,...) further hold all entries of the query elements and only the table RSZCOMPDIR is reduced by the entry of the query.

When deleting a query via the Query Designer all these 'existentially required' elements are written to the request and also deleted.

So I wonder if this t-code is just doing a 'quick but dirty' job?

Or maybe exist different versions of it?

Can anybody confirm this or knows about a newer version?

Thanks in advance.

Regards,

Sebastian

[mike_howles4 · Accepted solution]

As you point out I'm sure that you're simply left with orphaned useless query elements in RSZELTDIR and RSZELTTXT.  There's a separate T-Code that will run an integrity check (I'll see if I can remember what it is and report back) to seek out these orphaned objects and suggest a clean-up, however maybe something only worth running once a year since I don't see a lot of harm in the entries hanging around in those tables short/mid-term.

[Former Member · Accepted solution]

Hi Sebastian,

I have also similar experience RSZDELETE will delete entries from RSZCOMPDIR table but not actually check query consistency. I prefer to run Program ANALYZE_RSZ_TABLES post running RSZDELETE and this i generally avoid to use, but if no choice then go by this way

Regards,

Ganesh

[Former Member · Accepted solution]

Thank you both for your anwsers.

Seems that the report ANALYZE_RSZ_TABLES is what I'm looking for.

Btw, the t-code for it is RSZTABLES.

Regards,

Sebastian

[Former Member · Accepted solution]

Hi Ganesh.

Thanks for your reply and that report.

I've run it and it gives a lot of messages.

@Michael: is this report that one you think of?

Regards

[mike_howles4 · Accepted solution]

Yes, I was just running on our sandbox to confirm before replying

I think you can also get to the report through some navigation beginning in RSRT or RSA1 however knowing the actual report name is always quickest!