Client import with profile "SAP_USER" is very long ?

[Former Member]

Hello,

In our operating mode of validation system refresh, we use an export of users+profiles (SCC8 and SAP_USER profile) made on pre-refresh step ... And the we import this "SAP_USER" export with STMS (and the post import with SCC7 transaction).

We have 8300 users ...

Import with STMS take 02h30

Post import (SCC7) take 00h30

The export of the client is very quick ...

It will be very usefull for us if the import could be faster ... Is there a problem with this duration ?

Is there some actions that can improve performance ?

Thanks by advance for any helps,

Fabrice.

[former_member185954 · Accepted solution]

Hello Daniel,

Please check notes:

Note 86535 - tp/R3trans: Splitting large exports

Note 68896 - R3trans: Performance problems and cluster

The 2nd note describes, how Import performance can be improved by preemptively deleting selected data from target client before import starts.

One key point described in SAP Note 68896 is the following:

"When importing, R3trans attempts to merge the table entries transported with the existing ones in memory and to only write the changed table entries. To do this, all entries from the logical tables must be read and expanded, requiring considerable resources. The strategy of only writing back changed records, in order to avoid database accesses is less effective for cluster tables, as a change to the logical table entry suffices to force the physical table entry to be written back to the database. A previous deletion of the physical clusters can significantly improve the speed of this."

Hope that helps.

Regards,

Siddhesh

[Former Member]

Hi Experts,

i have facing poor performance problem in sql 2000, during the client import process. log size increase very rapidly and db server not able to backup and truncate these logs in time. do to this problem log partition fill up and import process fails. i also tried to put db in simple recovery model but in this logs are increasing.

here also same situation arose, i also tried to shrink logfile but it not help too much.

Can i import the sap profiles in several time despite the importing the SAP_ALL in once without loosing the data in target client.

Thanks,

Rakesh

.

[former_member189725]

you will lose all data in target client with SAP_ALL profile. SAP_USER profile will only replace the user data and authorization profile in the target client without touching any other data in target client.

[Former Member]

Hi Daniel,

I assume that your are running on Oracle RDBMS. Please check the items, below;

1) Missing indexes: Actually, it may lead performance problems while reading the data, but in this case, you should check it,

2) Number of customer indexes: If a table has many more indexes, it may lead long runtimes for DML statements. For the instance, you may face with a performance problem, while updating indexes. Under this circumstance, you may check which table is causing long run-time, during your import process,

3) Wrong statistics on your tables,

4) Small rollback segments.

As an additioanl info, you can check "Note 68896 - R3trans: Performance problems and cluster" on SAP Marketplace.

Best regards,

Orkun Gedik

Senior SAP Netweaver and Development Consultant

SAP Technical Architect

[Former Member]

Hi Daniel,

Maybe these Note can help you.

Note 520592 - ConnTrans - Poor performanace during Import

Good Luck.

Ignacio.

[Former Member]

Hi,

Thanks for your answer, but this note concern "Mobile Client" ...

I speek on my question about client export and import on ECC6 server.

Thanks again for nay help,

Fabrice.

[Former Member]

Have you tried deleting the client before importing the transport request?

Note 552711 - FAQ: Client copy

Note 70643 - CC-TOPIC: Client Deletion (SCC5)

[martin_juen2]

I agree with Orkun Gedik.

You should check the copied tables and the select statements. Maybe you can find where-clauses where no index is defined. Additional you should create statistics for the copied tables.

If you are copying the user data very often, it can happen that the tables and indexes are fragmented. In this case you can do (online) reorgs of these tables.

regards, Martin