Database - Grows unexpectedly and space issue occurs

[former_member199632]

Hi,

In the DEV system Database grows unexpectedly and now we are running out of space for /oracle

Monthly Growth rate is as follows; in (MB)

01.11.2015    PSAPSR3	21,200.00
01.10.2015    PSAPSR3	19,440.00
01.09.2015    PSAPSR3	9,360.00
01.08.2015    PSAPSR3	80.00
01.07.2015    PSAPSR3	160.00
01.06.2015    PSAPSR3	80.00
01.05.2015    PSAPSR3	20.00
01.04.2015    PSAPSR3	20.00
01.03.2015    PSAPSR3	20.00
01.02.2015    PSAPSR3	80.00

What would be the reason for this and how to locate what are the files grows abnormally?

(Any sql queries?)

Growing partition is /oracle

Oracle 10.2.0.4

SLES 10 SP2

Note- In this server currently there is a block corruption issue as well. Will that causing some abnormal act?

Ref:  http://scn.sap.com/thread/3806478

Looking forward for the soonest of response

regards,

zerandib

[Brindavan_M]

Hi ,

Check the growth history in DB02 at SAP level. Go to  DB02 -> Segments -> Overview -> Top growth in days. check any  BG job is causing issue and os level just check how frequently orrach log occurring. AP Check the SAP standard jobs are running  like  reorg  any dumps, spools log is filled... Below link may help full

How to monitor table growth - SAP on Oracle - SCN Wiki

Thanks,

Brindavan

[former_member199632]

Hi Brindavan,

I have found out this table grows abnormally -> GVD_LATCHCHILDS

(Which is affected by data block corruption issue. This block corruption issue is still existing in the server)

SAPSR3    GVD_LATCHCHILDS

TABLE    PSAPSR3    40,956.000    377.600    830    5,242,368

SAPSR3    GVD_LATCHCHILDS~0    INDEX    PSAPSR3    17,709.000    189.867    461    2,266,752

Any idea how to overcome this abnormal growth?

regards,

zerandib.

[JamesZ]

Hi Zerandib,

You can check SAP note

706478  for GVD_LATCHCHILDS

Best regards,
James

[Brindavan_M]

Hi,

The Oracle corruption is most frequently caused by a bad disk, although there are rare cases of "logical" corruption within the Oracle data blocks. You may use this command to list the block corruption.  By daily monitoring and health check the system you can identified or might some script need to be set like CCMS and OS level threashold alerts. So that you can take action before the issue started critical.

SELECT * FROM V$DATABASE_BLOCK_CORRUPTION;

Thanks,

Brindavan

[former_member207186]

Hi,

To handle corruption issue, you can proceed as per SAP Notes 23345 - "Consistency check of ORACLE database" and 365481 - "Block corruptions". For more information, please also refer to SAP KBA 1559652 - "How to deal with block corruptions on Oracle".

Regarding GVD_* tables, you can also check carefully the SAP Notes below

706478 - "Preventing Basis tables from increasing considerably (point 44.)"

1411834 - "DBA Cockpit: GVD_* tables, RSORAHCL, history"

1080813 - "DBA Cockpit: Changing history update (Oracle database)"

Regards,
Bíborka