Skip to Content
avatar image
Former Member

Why segment SXMSCLUR and SXMSCLUP increase rapidly?

Hello expert,

Regarding table SXMSCLUR and SXMSCLUP, there are a lot of blog and notes on it.

Most of them are talking on using a program to delete surplus data,or archive,deletion job.

The notes give  solution to delete table content,but it didn't explain the reason.

Deletion and archive jobs iwill delete message,but it seems no jobs will delete content of SXMSCLUR and SXMSCLUP.

In recent days,we met a issue related to these two tables.

DBA found one of our table-space increased quickly in recent days,in DB02 we found the remarkable increase were in table  SXMSCLUR and SXMSCLUP and related SYS_LOB0000$$ table.

We checked our regular deletion program:

RSXMB_DELETE_MESSAGES

RSXMB_TABLE_SWITCH

RSXMB_DELETE_HISTORY

They ran normally everyday.


We don't understand why the content of SXMSCLUR and SXMSCLUP increase so quickly after a certain date?

Actually,we didn't find remarkable increase of SXMSCLUR and SXMSCLUP before that date.


Is there any auto housekeeping mechanism to clean the table content, and it is easy to break?


Pls help.

Thanks.



Add comment
10|10000 characters needed characters exceeded

  • Get RSS Feed

3 Answers

  • Jan 17, 2014 at 03:21 PM

    Hi,

    Please check the below discussion

    check the below in the discussion

    You should run report "RSXMB_DEL_CLUST_TABLE" to delete entries from this two tables.

    Yo need activate Archiving and Deletion Jobs.

    To find the relevant jobs, call transaction SXMB_ADM and choose: Administration > Schedule Archiving Jobs

    If you using BPM see this notes:

    SAP Note 836092: Archiving/deleting work items in the XI

    SAP Note 874708: BPE HT: Deleting message persistence data in SWFRXI*

    you can also check the retention period configuration in SXMB_ADM (retention period should not be too large).

    to check why the size is increasing, check the reply from Daniel in below discussion

    regards,

    Harish

    Add comment
    10|10000 characters needed characters exceeded

  • Jan 17, 2014 at 03:46 PM

    The above suggested report (RSXMB_DEL_CLUST_TABLE) is applicable only if you are under XI 3.0 SP 31.

    However please check if you have scheduled multiple archive/deltion jobs.

    go to SXMB_MONI->background job monitor -> It'll give an overview and schedule details.

    You must not setup multiple archival/deletion jobs.

    Also please go through the troubleshooting guide :

    872388 - Troubleshooting Archiving and Deletion in PI

    We had the same issue. In our case unfortunatly there were multiple jobs scheduled and not all the messages were getting deleted.

    When you are in the process of troubleshooting, i would also suggest you to create an OSS message.

    Add comment
    10|10000 characters needed characters exceeded

  • Jan 17, 2014 at 08:15 PM

    Hello

    The delete jobs in PI will do just that i.e. they delete the PI messages from the DB. But they do not release the DB tablespace. This is up to the DB Administrator. Depending on your DB, it will be necessary to run

    DB reorganization jobs to release the space. See note 1791627 PI DB tables relevant for Reorg jobs, for more details.

    Some possible reasons why these tables would increase greatly in size:

    1. a big increase in the number of messages being processed
    2. if the Retention period of the PI messages has been increased
    3. if there are Communication Channels in error status continually reprocessing error messages (these messages should be cancelled)

    Regards

    Mark

    Add comment
    10|10000 characters needed characters exceeded

    • Former Member

      very helpful.

      A strange thing was. when found the entries number was decreased by the house-keeping job,but, when we checking the table size in DB02,we found the size increased.