Slow update - RSM13000 (SAPMV45A/SAPMV60A/SAPMV75A/SAPMV56A...)

[Mark17]

Hello guys,

in our system like 30% of V* transactions have slow update.
Example: VA01, VA02, VA22, VT02n, VT03n, VF11…

During this slow update, users are waiting and complaining.

In ST03n I can see, 
Update V1 = slow
Update V2 = fast

There is no bottleneck in Workprocesses (UPD*) or CPU usage or anything else.
There are also several instances where update request gets distributed.
Also no problem with hanging HTTP RFC connections (SAPLHTTP_RUNTIME...).
As far as I can see there is no locking problem in SM12.

Most of time is always used under "Processing Time".

The only thing I can see is high "RFC idle time"... can this be a reason (?)
Because "calling time" and "data send time" are pretty low.

ARCHIV_CREATE_TABLE_MULT

The UPD WP is remaining in program SAPLCEI3 without more information, only time counts...
and after 5-20 seconds it will finish.

This happens with ~30% of update requests in V* transactions.

Maybe you have some other ideas what to check?

Thank you.

[Daniel_AC]

Hello,

SAPLCEI3 is tied to CDI and Change Pointers, that are often triggered during sales document changes.
If you're using MDG, CRM, EWM, PI/PO... with a tRFC/BG RFC destination configured, the UPD task could be waiting for RFC response...

- SM59: RFC destinations, any RFC still active and not properly responding?
- SMQR and SMQS: qRFC queues for bottlenecks or outbound ALE queues
- SM58: Any tRFC stuck or delayed?
- BD64 and SCC4: There's distribution model configured for Sales Docs? Can be triggering outbound processing...
- ST12: Debug one slow updates and you will see SAPLCEI3 call to MASTER_IDOC_DISTRIBUTE, ALE_CHANGE_POINTER_WRITE...

Best regards,
Daniel Alves Costa