Skip to Main Content

Oracle Database Discussions

Announcement

For appeals, questions and feedback about Oracle Forums, please email oracle-forums-moderators_us@oracle.com. Technical questions should be asked in the appropriate category. Thank you!

"ORA-01722: invalid number" in alert log - caused memory dump.

GlenStromApr 5 2017 — edited Apr 6 2017

Database is 11.2.0.3 running on Solaris OS on a Sparc Enterprise M5000 server.

Had the error show up 4 times in succession, saying nothing more than a trace file had been generated. The generated trace file shows that a memory dump had occurred, with the following info showing that it was an OEM module running at the time.

=======================================================================

System name:    SunOS

Node name:      ****

Release:        5.10

Version:        Generic_150400-35

Machine:        ****

Instance name: ****

Redo thread mounted by this instance: 1

Oracle process number: 365

Unix process pid: 8691, image: oracle@****

*** 2017-04-04 09:11:50.618

*** SESSION ID:(2468.3759) 2017-04-04 09:11:50.618

*** CLIENT ID:() 2017-04-04 09:11:50.618

*** SERVICE NAME:(SYS$USERS) 2017-04-04 09:11:50.618

*** MODULE NAME:(OEM) 2017-04-04 09:11:50.618

*** ACTION NAME:([ACTIVE] ExecuteThread: '22' for) 2017-04-04 09:11:50.618

==========================================================================

I've never seen a ORA-01722 cause a memory dump before, and a google search for similar issues was fruitless.

One of the other DBA's said they may have been running a query analyzer on OEM at the time - so perhaps it's possible this may have caused it.

Other than that, I have no idea, there are no other occurrences of this in the alert log.  Has anyone else seen or run into this before?

Thanks in advance.

Comments
Locked Post
New comments cannot be posted to this locked post.
Post Details
Locked on May 4 2017
Added on Apr 5 2017
3 comments
497 views