cancel
Showing results for 
Search instead for 
Did you mean: 

How does the buffer trace identify changes on tables?

Former Member
0 Kudos

Hello all!

If I switch on a buffer trace, I can see the executed operations. There are READ SINGLE's, FETCHE's and OPEN's.

But if I make a change (Insert, Delete...) on a buffered table, I can only see Read single, fetch and open. These stmts correspond to queries. But what is about changes? How are they called in the trace?

Thanks,

Holger

Message was edited by: Holger Schmidt

View Entire Topic
Former Member
0 Kudos

Hi Holger,

Your understanding is correct on rdisp/bufreftime and 'Last modified' . On the F1 help on buffer state for 'Pending' I can only find the OSS which is also not clear. But I tried to read a table in "Pending" for a about 6 -7 times in SE16 . After a while ( say 10 mins ) I ran ST02 again and the table status changed to valid . So the documentation is correct that buffer will be loaded after certain read access but its still not confirmed as to how many read access would reload the buffer in "pending" state and after how much time.

That is it on this topic.

Cheers

Former Member
0 Kudos

Hi!

You have to test it!

An invalidated table resides in the pending state till a defined number of read accesses were executed on it!

The helpdata says 'a certain number' because this may change from release to release...

So you have to test how many read accesses it takes till the table is in the vaild state!

That is the solution. I am totally sure about this, because I talked to sb. who wrote a book concerning that.

Holger