Deletion of errored update records on SAP Instance restart

[Former Member]

Dear All,

We do a quarterly reboot of our SAP Production environment. Upon SAP instances restart the below listed update sessions are automatically invoked for deletion of errored update records older than 50 days (rdisp/vbdelete) from the tables VBHDR,VBMOD,VBDATA and VBERROR.

Server No Typ PID CPU Time Cli User Report Tcod CUA rep. Scre Fcod / Job Action/Reason for waiting Table

sap2_SM2_01 16 UPD 4372 129056 000 SAPSYS SAPLTHFB 3000 Sequential read VBMOD

sap3_SM2_02 15 UPD 3720 129014 000 SAPSYS SAPLTHFB

sap4_SM2_03 17 UPD 5732 128969 000 SAPSYS Sequential read VBHDR

sap5_SM2_04 15 UPD 3644 128925 000 SAPSYS Sequential read VBHDR

sapv_SM2_00 4 UPD 7697 129094 000 SAPSYS SAPLTHFB Sequential read VBHDR

These sessions are active for almost 2 days since the SAP instance startup resulting in a high average response time (ART) on all the instances. Though it is not having an impact on SAP instance and production users but it is badly impacting the ART data.

Below is our Production enviroment information.

R/3 Release: 3.0F

R/3 Kernel: 3.1I_EXT

Database: Oracle 9.2.0.8

OS: HP-UX 11.23

Can you please confirm if this is encountered in your production environment and how to resolve it. Is it mandatory to have these update sessions run on instance startup each time.

Thanks & Regards,

Prasad

[Former Member]

Thanks Nick for your reply.

What I noticed is these update sessions are automatically started by SAP once the Instance is up and running.

I have checked and don't see any backgrould jobs that were triggered at the time of instance startup.

I have checked for SAP notes to see how these get trigerred and if they can be controlled but couldn't find any relevant information.

Appreciate if you can provide somemore information on this please.

[Former Member]

R/3 Release: 3.0F
R/3 Kernel: 3.1I_EXT
Database: Oracle 9.2.0.8
OS: HP-UX 11.23

All the releases are out of support (You are running your entire setup in a fire fighting mode)

And there must be some jobs running at the time of restart, check all the activities on your server and take a restart again.

Regards,

Nick Loy