FAQ
Good day, List,

I am playing with 11g and want to experiment with the Fault Diagnosability
Infrastructure. I would like to force an ora-600. I thought there was an
event one could set to crash smon with a ora-600, but looking through
$OH/rdbms/mesg/oraus.msg, I cannot find it. Anyone recall it off-hand?

Or if I could force a bug, that would be nice. So far the bugs I have found
for 11.1.0.6 are not reproducible. Especially when the testcase is not
given.

PS - gotta love those events that tell the customer "do not set this event".
Isn't that just asking to be set?

Search Discussions

  • Tim Gorman at Apr 4, 2008 at 9:32 pm
    Charles,

    Just type in "oracle" at the UNIX command-line (assuming you're on
    UNIX, haven't tried it on Windows), and you'll get an ORA-00600
    [12235] in the "alert log" file. Great way to wake up the DBAs...;-)

    Hope this helps...

    -Tim

    Quoting Charles Schultz :
    Good day, List,

    I am playing with 11g and want to experiment with the Fault Diagnosability
    Infrastructure. I would like to force an ora-600. I thought there was an
    event one could set to crash smon with a ora-600, but looking through
    $OH/rdbms/mesg/oraus.msg, I cannot find it. Anyone recall it off-hand?

    Or if I could force a bug, that would be nice. So far the bugs I have found
    for 11.1.0.6 are not reproducible. Especially when the testcase is not
    given.

    PS - gotta love those events that tell the customer "do not set this event".
    Isn't that just asking to be set?

    --
    Charles Schultz
    --
    http://www.freelists.org/webpage/oracle-l
  • Charles Schultz at Apr 5, 2008 at 2:49 am
    Perhaps this is no longer true for 11g (11.1.0.6)? Linux RHEL4.

    Jeremiah Wilton, your suggestion worked wonderfully, thank you. I also found
    that "blissful_ignorance" works quite well, also. =) No to go surprise my
    fellow DBAs....
    On Fri, Apr 4, 2008 at 4:32 PM, Tim Gorman wrote:

    Charles,

    Just type in "oracle" at the UNIX command-line (assuming you're on UNIX,
    haven't tried it on Windows), and you'll get an ORA-00600 [12235] in the
    "alert log" file. Great way to wake up the DBAs...;-)

    Hope this helps...

    -Tim



    Quoting Charles Schultz :

    Good day, List,
    I am playing with 11g and want to experiment with the Fault
    Diagnosability
    Infrastructure. I would like to force an ora-600. I thought there was an
    event one could set to crash smon with a ora-600, but looking through
    $OH/rdbms/mesg/oraus.msg, I cannot find it. Anyone recall it off-hand?

    Or if I could force a bug, that would be nice. So far the bugs I have
    found
    for 11.1.0.6 are not reproducible. Especially when the testcase is not
    given.

    PS - gotta love those events that tell the customer "do not set this
    event".
    Isn't that just asking to be set?

    --
    Charles Schultz
    --
    http://www.freelists.org/webpage/oracle-l

    --
    Charles Schultz

    --
    http://www.freelists.org/webpage/oracle-l
  • Jeremiah Wilton at Apr 5, 2008 at 4:09 am

    Charles Schultz wrote:

    Perhaps this is no longer true for 11g (11.1.0.6)? Linux RHEL4.
    I also was unable to raise ORA-00600[12235] on 11.1.0.6 Linux/32 using the
    tried and true method of invoking the oracle binary from the shell.
    Jeremiah Wilton, your suggestion worked wonderfully, thank you.
    NP. The story behind this is that Dan Morgan was somewhat miffed that ADR
    had no facility for generating a "generic" event that would result in a
    payload for Oracle Support similar to RDA. The premise of ADR at present is
    that some kind of error like an ORA-00600 has to happen to cause an event
    that can be packaged. If ADR payloads are meant to replace RDA, then there
    needs to be some way to create an event/package on demand.

    Dan bugged a few people at Oracle and was advised to raise ORA-700 with
    dbke_test when he wanted to create an event for purely diagnostic purposes
    (see blog entry). Dan generously shared this with me one day in the corridor
    at OOW.

    I liked the look of dbke_test and soon found that it has a usage prompt (as
    per my blog entry), and tried a few other functions. One lets you raise
    ORA-00600 (OERI) (dde_flow_kge_ora), with any arguments you like. The
    PL/SQL method has no way to pass arguments to the OERI. I also perused the
    Oracle binary for other unit_test components and found a few.

    Regards,

    Jeremiah Wilton
    ORA-600 Consulting
    http://www.ora-600.net
  • Tim Gorman at Apr 5, 2008 at 4:17 am
    <!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">

    If it no longer causes an ORA-00600 in 11.1, then it's about darned
    time they took care of that...
    ORA-00600 is supposed to be a "catch-all" for unforeseen or unhandled
    errors, like an "assert()" function in "C".  I recall being able to
    generate the OERI-12235 by simply entering "oracle" at the command-line
    since at least Oracle v6;  hardly an "unforeseen" or "unhandle-able"
    exception to have "argc == 0" -- every single blessed "C" program I
    ever wrote handled that, if when I was just goofing around...  ;-)

    Charles Schultz wrote:
    Perhaps this is no longer true for 11g (11.1.0.6)? Linux
    RHEL4.

    Jeremiah Wilton, your suggestion worked wonderfully, thank you. I also
    found that "blissful_ignorance" works quite well, also. =) No to go
    surprise my fellow DBAs....

    On Fri, Apr 4, 2008 at 4:32 PM, Tim Gorman
    wrote:
    Charles,

    Just type in "oracle" at the UNIX command-line (assuming you're on
    UNIX, haven't tried it on Windows), and you'll get an ORA-00600 [12235]
    in the "alert log" file.  Great way to wake up the DBAs...;-)

    Hope this helps...

    -Tim

    Quoting Charles Schultz <sacrophyte@gmail.com>:

    Good day, List,

    I am playing with 11g and want to experiment with the Fault
    Diagnosability
    Infrastructure. I would like to force an ora-600. I thought there was an
    event one could set to crash smon with a ora-600, but looking through
    $OH/rdbms/mesg/oraus.msg, I cannot find it. Anyone recall it off-hand?

    Or if I could force a bug, that would be nice. So far the bugs I have
    found
    for 11.1.0.6
    are not reproducible. Especially when the testcase is not
    given.

    PS - gotta love those events that tell the customer "do not set this
    event".
    Isn't that just asking to be set?
  • Jeremiah Wilton at Apr 4, 2008 at 9:36 pm

    Charles Schultz wrote:

    ...
    I would like to force an ora-600.
    Please see my blog entry on dbke_test:

    http://oradeblog.blogspot.com/2007/10/unittest-command-in-oradebug.html

    In short:

    SQL> oradebug unit_test dbke_test dde_flow_kge_ora stupid_idiot 0 0

    or

    declare
    v_bug exception;
    pragma exception_init(v_bug,-600);
    begin
    raise v_bug;
    end;
    /

    (Thanks to Jared for PL/SQL)

    Jeremiah Wilton
    ORA-600 Consulting
    http://www.ora-600.net
  • David Sharples at Apr 4, 2008 at 9:52 pm
    type oracle at the command line

    that will do it for you
    On 04/04/2008, Charles Schultz wrote:

    Good day, List,

    I am playing with 11g and want to experiment with the Fault Diagnosability
    Infrastructure. I would like to force an ora-600. I thought there was an
    event one could set to crash smon with a ora-600, but looking through
    $OH/rdbms/mesg/oraus.msg, I cannot find it. Anyone recall it off-hand?

    Or if I could force a bug, that would be nice. So far the bugs I have
    found for 11.1.0.6 are not reproducible. Especially when the testcase is
    not given.
    --
    http://www.freelists.org/webpage/oracle-l

Related Discussions

Discussion Navigation
viewthread | post
Discussion Overview
grouporacle-l @
categoriesoracle
postedApr 4, '08 at 8:23p
activeApr 5, '08 at 4:17a
posts7
users4
websiteoracle.com

People

Translate

site design / logo © 2019 Grokbase