Best. Error. Message. Ever.
Got this little gem while running sqlldr:
SQL*Loader-926: OCI error while uldlfca: unexpected cvtret with ldret of no_data for table xxyyzz
From the documentation:
SQL*Loader-00926: OCI error while string for table string <!-- class="msg" -->
Cause: An OCI error has occurred.<!-- class="msgexplan" -->
Action: Check the message below this one in the log file for more information. The table or column referenced may not be accessible.
You did read the documentation, right?
l.<!-- class="msgaction" --><!-- class="msgentry" -->
I did, and I fixed the problem. I was just expecting a little more descriptive, so I wouldn't have to dig out the docs to fix a simple problem.
I dunno, I might have it beat with this one. Seems to happen whenever there is a database timeout in Oracle Forms.
Oracle Forms are a horrible abomination anyway.
While patching a patch I found the following patch description interesting:
Feb 27, 2007 4743226 Patch RDBMS Server Generic Platform 10.2.0.3 BUG 3680460 INCORRECTLY BROKE BUG 3588054 available
I don't know if I would be happier if the bugs break each other correctly.
That's what happens when you set NLS_LANG=Martian_Martian.MAR64. You're an Earthling, I take it?
I got everyone here beat.
During a routine switchover to a standby of our biggest (5 TB), most important database:
ORA-16139 DURING ALTER DATABASE COMMIT TO SWITCHOVER TO STANDBY;
We were testing a disaster recovery and switching our databases to our standby sites. We had done this many times and had saved our most important database for last. Did I mention it was 5TB, 24x7 and had 10k users? After receiving this error, we couldn't open either the primary nor the standby. What was supposed to be about a 1-2 hour outage (that management had difficulty approving) turned into a 14 hour nightmare. We got no meaningful support from Oracle and there was time in which we thought we would have to restore this database from tape and do incomplete media recovery. And did I mention it was 5TB?
not to be an a$$, but it took 14 hours to do a forced recovery? i'm guessing because it was 5 tb?
the oracle forms error was classic though...