cancel
Showing results for 
Search instead for 
Did you mean: 

Process Chain failure

Former Member
0 Kudos

Hi All

Can anybody tell what all are the possible reason which cause failure of Process Chain in APO DP.

Also Why inconsistency happens in Time series(TSCONS) and in Live cache(OM17)

Regards

Ravi

Accepted Solutions (0)

Answers (2)

Answers (2)

former_member458819
Active Participant
0 Kudos

Reasons For Process Chain Failure

some points to remember while monitoring:

  1. 1.      In case of running local chain, right click and go to process monitor to check if there is any failure in any process step. It may happen that there might be failure in one of the parallel running process in the chain.
  2. 2.      Right click and select “Display messages” option to know more about reason of the failure in any step of process chain.
  3. 3.      Try to correct step which takes longer time to complete by comparing with other processes which are running in parallel.
  4. 4.      Check the lock entries on Targets in SM12 transaction. This will give you an idea for all the locks for the target.
  5. 5.      Perform RSRV check to analyze the error and correct it in relevant scenarios.

Monitoring - Approach and Correction

Description

Approach/Analysis

Correction

Failure in  the Delete Index or Create Index step

Go to the target and check if any load is running

Trigger the chain ahead if indexes are already deleted in  other process chain

Long running Delete Index or Create Index job

Compare the last run time of job. If it taking more time,  check system logs in SM21 and server processes in SM51

Inform BASIS team with system log and server processes  details

Check  the SM12 transaction entries for any locks on  target from some other step

Stop the delete index or create index and repeat it once  lock is released

Attribute change run failure

Check the error message

Check the lock on Master Data objects used in Infocube and  repeat the step

Roll up failure

Error message “roll up failed due to Master  Data table lock in attribute change run

Wait till attribute change run is completed and start the  roll up again

-

Open RSDDBIAMON2 transaction to check any issues in  BI Accelerator server.

Failure in Full data load

SQL error

-

Manually delete the indexes of target and start the load  again

This happens generally due insufficient memory space  available while reading data from tables.

Coordinate with BASIS team to analyze the database space  (DB02) and take necessary action to reduce the database size.

Incorrect data records in PSA table

Check the data in PSA table for any lower case letter or  special characters etc

Ask concerned R/3 person to correct the data at R/3 side  and load the data again

It is very important to check the PSA table to have  complete data in PSA before reconstruction. Otherwise all the data will not  be updated in target.

On immediate basis, delete the request from target,  correct the data in PSA table and reconstruct the request

Delta data load Failure

Delta load failed without any error message.

Go to infopackage and check for last delta load in target  if it is successful or not.

Wait till last delta load is completed or correct it then  start the infopackage again.

PSA update failed due to job termination or target lock

Make sure that all data records are available in PSA table  and reconstruct the request

Delta load job failed in R/3 system or EDWH system after  all data has come to PSA table

Delete the request from targets and reconstruct it.

Delta load job failed at R/3 side without data PSA table

Make the status of the request “Red” in process monitor  and repeat the delta load again. Make sure that data is not doubled as repeat  delta brings last successful records as well. Check the update mode of all  key figures i.e. overwrite or addition mode for this.

Delta load job failed in EDWH system without data PSA  table

Remove “data mart” tick from below source and start the  delta package again.

Or make the status of the load to “Red” in case all key  figures are in overwrite mode.

Long running delta load

Check  the R/3 job in SM37 if it is not released

Ask BASIS team to release the job from R/3 system

Compare the last successful delta load time along with R/3  connection

Ask BASIS team to repair the R/3 connection

DSO Activation

Activation failure due to incorrect value in the ODS  field.

Correct the data in PSA or check the update logic for the  corresponding source field

Activation job termination

Try to activate the data manually else check the jobs in  SM37. If most of the activation jobs are failing, contact BASIS team.

Miscellaneous

Some process types are not triggered even if above process  type is completed. Triggering line doesn’t show any status.

Chain logs takes time to update in chain log table i.e.  RSPCPROCESSLOG. Check the system performance with BASIS team as well.

Table Name

Usage

RSPCPROCESSLOG

Check the process type log

RSPCCHAIN

Check the main chain for local chain

Former Member
0 Kudos

Hi Ravi,

Process chain failure may happen because of many reason. Some of them are

- Macro issue

- Data Transfer Process (DTP) Issue

- ABAP Errors ( Space / Memory Issue) --Short Dumps

- User lock issue

- Parallel job overlap

- inconsistencies in Planning area or Transactional data

The below may answer your question why inconsistency happens in Time series

In APO data is stored in two areas

1. DB - is a permanent storage area used for Tables and DP master data (POS)

2. LC - a object oriented DB system, where the calculations happen and where the functions get data with quick access

The data for planning is held in the live Cache database. Data from the DB memory is moved to LC and back depending on the programs and operations performed. The LC uses GUIDs and maps it to the master data in the DB

This transfer of data leads to inconsistency. The consistency check reports aim to take care of these issues

For details please check the link http://scn.sap.com/thread/429219

Thanks

Diana

Former Member
0 Kudos

Hi Diana

I request you to please elaborate the points like macro issue,DTP issue, parallel job overlap and user lock issue which cause process chain failure. I mean to say please explain why these issues happens and what to be done to correct these issues related with process chain

Regards

Ravi