| GATHER_STATS_JOB: Stopped by Scheduler [message #541977] |
Fri, 03 February 2012 09:50  |
 |
BeefStu
Messages: 208
Registered: October 2011
|
Senior Member |
|
|
I am getting the following error in a trace file.
ORA-00600: internal error code, arguments:
[kqldxbr_30], [1403], [], [], [], [],
ORA-00600: internal error code, arguments: [kqldxbr_30],
[1403
DBMS_STATS: GATHER_STATS_JOB: Stopped by Scheduler.
DBMS_STATS: Consider increasing the maintenance window duration if this happens
frequently.
DBMS_STATS: The following objects/segments were not analyzed due to timeout:
DBMS_STATS: TABLE: "OPS$ORACLE"."EXT_GROWTH_DATAPUMP_LOG".""
DBMS_STATS: TABLE: "IFD_OWN"."EXT_CUST_ECPD".""
...
...
...
Few questions:
1) Can somebody provide me with a query to see how long this
job runs for and another query on how to extend it?
2) Does anybody have any opinions (good/bad) about changing
Oracle pre-defined jobs or should new jobs be put into place
and the std jobs left alone.
Thanks to all who answer
|
|
|
|
|
|
|
|
|
|
| Re: GATHER_STATS_JOB: Stopped by Scheduler [message #541994 is a reply to message #541988] |
Fri, 03 February 2012 11:27  |
John Watson
Messages: 8928
Registered: January 2010
Location: Global Village
|
Senior Member |
|
|
Agreed.
The GATHER_STATS_JOB issue should fix itself. Because of the way the the job is setup and the way the procedure runs, if it doesn't finish one night it will tend to start next evening from where it left off. Unless you regularly do a huge amount of DML in proportion to the size of the database, it should sort itself out after a few days.
|
|
|
|
| Re: GATHER_STATS_JOB: Stopped by Scheduler [message #541995 is a reply to message #541977] |
Fri, 03 February 2012 11:33 |
John Watson
Messages: 8928
Registered: January 2010
Location: Global Village
|
Senior Member |
|
|
Quote:2) Does anybody have any opinions (good/bad) about changing
Oracle pre-defined jobs or should new jobs be put into place
and the std jobs left alone. I would leave the autotask job(s) in place, unchanged. Except for changing the maintenance windows: they are awfully American, and may not be suitable if you operate in many time zones. Or in countries where Saturday and Sunday are working days.
If your database is any sort of upgrade, make sure that the old 10g scheduler job that gathers stats is not running. I've seen several 11g databases that were being analyzed twice a night, once by the 10g job and again by the autotask.
|
|
|
|
| GATHER_STATS_JOB: Stopped by Scheduler [message #542005 is a reply to message #541995] |
Fri, 03 February 2012 12:23 |
|
BeefStu
Messages: 208
Registered: October 2011
|
Senior Member |
|
|
Michel/John,
If I look in the .trc file I do see some failures..
I will work on fixing these and see if the GATHER STATS
message goes away.
I was a bit unsure if the job actually picked up where it
left off.
FYI, the DB was upgraded from 11107 to 11203 and there are NO
10G jobs running in the scheduler.
Thank you both for your expertise and advice.
DBMS_STATS: GATHER_STATS_JOB: GATHER_TABLE_STATS('"IFD_OWN"','"EXT_CUST_ECPD"','
""', ...)
DBMS_STATS: ORA-20011: Approximate NDV failed: ORA-29913: error in executing ODC
IEXTTABLEOPEN callout
ORA-29400: data cartridge error
KUP-04040: file cust_ecpd.dat in EDW_DATA_TRANSFER_DIR not found
|
|
|
|
] 
Blog
Search
Help
Members
Register
Login
Home
