Some troubles — especially those happening only sporadically — are not so easy to shoot and call for a deeper understanding of the matter. In the following real-world example this means: SQL*Net Tracing and some knowledge about the inner workings of the server’s operating system, particularly random number generation.
This case was suited well to demonstrate an approach to trouble-shoot connections to Oracle databases.
Up to Oracle 11.2 it was possible to display archived SQL plans from Statspack using DBMS_XPLAN. I make use of this in some of my scripts and SQL Developer Reports since I first saw this in Christian Antognini’s Book “Troubleshooting Oracle Performance“.
But in 12c (here: 188.8.131.52 on Linux), there’s a piece missing now:
select * from table(dbms_xplan.display(
table_name => 'perfstat.stats$sql_plan',
statement_id => null,
format => 'ALL -predicate -note',
filter_preds => 'plan_hash_value = '|| &&phv
ERROR: an uncaught error in function display has happened; please contact Oracle support
Please provide also a DMP file of the used plan table perfstat.stats$sql_plan
ORA-00904: "TIMESTAMP": invalid identifier
So it looks like STATS$SQL_PLAN wasn’t synchronized to the changes in 12c’s PLAN_TABLE. Maybe because the timestamp wouldn’t make much sense there, anyway, maybe simply because Oracle forgot.
==> Quick and most certainly unsupported workaround:
ALTER TABLE perfstat.stats$sql_plan ADD timestamp INVISIBLE AS (cast(NULL AS DATE));
Another workaround could be to create a separate view with an additional dummy timestamp column an reference the view. I chose to stick with the invisible column solution so I won’t have to create new objects in the DB and change scripts to use these objects.
Hopefully, this will be solved in 12.2 – at least, I had filed an SR / Enhancement Request with Oracle Support.
First things first:
Thou shalt not explicitly set AQ_TM_PROCESSES=0 !
Unless, of course, you want to disable the Queue Manager Process (QMNC in Oracle 11.x). Which you may want to do during database upgrades to prevent Streams or Advanced Queueing from interfering with the upgrade process.
However, if you don’t reset this parameter afterwards, you might run into the following scenario the next time you do a Data Pump export or import:
Just the other day I hit the unpublished Bug 16445132: DBMS_SCHEDULER PROBLEM WITH DATABASE MIGRATION VIA EXPDP/IMPDP.
It affects Scheduler Jobs that were created to end with a corresponding Scheduler Window.
Here’s an excerpt of a Data Pump Logfile:
ORA-39083: Object type PROCOBJ failed to create with error:
ORA-27469: STOP_ON_WINDOW_EXIT is not a valid job attribute
Actually, the correct attribute name for this procedure would be “stop_on_window_CLOSE” instead of “stop_on_window_EXIT”. The latter is an attribute of the JOB_DEFINITION object type which in my eyes is inconsistently named in respect to the SET_ATTRIBUTE procedure.
The bug is fixed, but as of Nov. 2014 there is no one-off patch for version 184.108.40.206. The existing workaround is to ignore the error and recreate the job after the import .
If you run into this bug and it bugs you a lot, the other option is to ask for a patch for this issue.
This is a translation of my German article which was requested through “Google Translate” many times. It seems that throughout the years this error and its mostly simple causes haven’t lost their significance.
Here’s a sample scenario:
- A table’s VARCHAR2 column shall be searched for numeric entries;
- We use a query with a WHERE clause that constrains the scanned rows to only those containing numeric values.
- Despite that, “ORA-01722” is thrown.
So, what happened?
- The Oracle documentation states: “When comparing a character value with a numeric value, Oracle converts the character data to a numeric value.” Think of an implicit TO_NUMBER being applied to the column.
- This implicit TO_NUMBER might be applied before the filter predicates of your WHERE clause.
- This in turn leads to an error, when an alphanumeric entry is hit — even when this entry wouldn’t be part of the scanned set if the filters were applied first.
Oracle’s “Security Alert Advisory for CVE-2012-3132” issued a warning about an attack vector that once again was discovered by security expert David Litchfield. The vulnerability allows to execute SQL code with SYS privileges by using object names containing quotation marks, if the attacker
- has authorized access to the database,
- has CREATE TABLE and CREATE PROCEDURE privileges and
- is allowed to execute DBMS_STATS.
A fix for this issue was published in July 2012 but there’s also Oracle’s advisory on how to cope with the threat without patching the RDBMS. This recommendation, however, has implications when such a database is recovered or cloned: