FAQ
hi:

My oracle version is 9205 and optimizer_features_enable is set to 9.2.0.

We have a query choose a bad plan.

The query is like this:

select /*+ index(e) */
max(event_id), sysdate from ANONYMOUS e where event_type = :1 and event_id
between :2 and :3 and creation_date < sysdate - :4
/

(I substitute the real table name with ANONYMOUS because of the policy)

The table has two indexes can be used for this query:

Index Name Column Name
------------------------------- -------------------------
---
ANONYMOUS_CE_IDX CREATION_DATE 1

EVENT_TYPE 2
ANONYMOUS_EEC_IDX EVENT_TYPE 1

EVENT_ID 2

CREATION_DATE 3

The query only returns one row, and obviously the second index
ANONYMOUS_EEC_IDX is more effecient.

But CBO chooses the first index and do a huge index range scan and table
block scan.

First I think maybe the statistics is not correct, so CBO choose the wrong
index.

But actually, CBO knows that using the second index will have the lower
cost.

SQL> explain plan for select /*+ index(e,ANONYMOUS_EEC_IDX,ANONYMOUS_CE_IDX)
*/
2 max(event_id), sysdate from ANONYMOUS e where
3 event_type = 'xxxxxxxx' and event_id between 3059913 and 3059914
4 and creation_date < sysdate - 1/100;

Explained.

SQL> @?/rdbms/admin/utlxpls

PLAN_TABLE_OUTPUT
Id | Operation | Name |
Rows | Bytes | Cost | Pstart| Pstop |
0 | SELECT STATEMENT | |
1 | 39 | 2771 | | |
1 | SORT AGGREGATE | |
1 | 39 | | | |
2 | PARTITION RANGE ALL |
1 | 13 |
* 3 | TABLE ACCESS BY LOCAL INDEX ROWID| ANONYMOUS | 1 |
39 | 2771 | 1 | 13 |
* 4 | INDEX RANGE SCAN | ANONYMOUS_CE_IDX |
119K| | 1412 | 1 | 13 |
---------------------------------------------------------------------------------------------------------------

SQL> explain plan for select /*+ index(e,ANONYMOUS_EEC_IDX) */
2 max(event_id), sysdate from ANONYMOUS e where
3 event_type = 'USER.BILLING_INFO.UPDATE' and event_id between 3059913
and 3059914
4 and creation_date < sysdate - 1/100;

Explained.

SQL> @?/rdbms/admin/utlxpls

PLAN_TABLE_OUTPUT
Id | Operation | Name | Rows | Bytes |
Cost | Pstart| Pstop |
0 | SELECT STATEMENT | | 1 | 39 |
40 | | |
1 | SORT AGGREGATE | | 1 | 39
2 | PARTITION RANGE ALL| | |
1 | 13 |
* 3 | INDEX RANGE SCAN | ANONYMOUS_EEC_IDX | 1 | 39 | 40
1 | 13 |
-------------------------------------------------------------------------------------------------

From plan, the cost of using second index is only 40 and the cost of using
first index is 2771. But CBO chooses the first one!! Why?

I did a 10053 trace on this sql and still can't understand the CBO behavior

BASE STATISTICAL INFORMATION

Table stats Table: ANONYMOUS Alias: E
(Using composite stats)
TOTAL :: CDN: 13276900 NBLKS: 612378 AVG_ROW_LEN: 156

-- Index stats
INDEX NAME: ANONYMOUS_CE_IDX COL#: 5 2

USING COMPOSITE STATS

TOTAL :: LVLS: 3 #LB: 152450 #DK: 207428 LB/K: 1 DB/K: 5 CLUF:

1056950
INDEX NAME: ANONYMOUS_EEC_IDX COL#: 2 1 5

USING COMPOSITE STATS

TOTAL :: LVLS: 3 #LB: 181250 #DK: 13035150 LB/K: 1 DB/K: 1 CLUF:

1647350
INDEX NAME: ANONYMOUS_PK COL#: 2 1 3

USING COMPOSITE STATS

TOTAL :: LVLS: 3 #LB: 167950 #DK: 13351700 LB/K: 1 DB/K: 1 CLUF:

1665200
INDEX NAME: ANONYMOUS_R_IDX COL#: 7

USING COMPOSITE STATS

TOTAL :: LVLS: 2 #LB: 84700 #DK: 201870 LB/K: 1 DB/K: 46 CLUF:

9445950
_OPTIMIZER_PERCENT_PARALLEL = 0

SINGLE TABLE ACCESS PATH

Column: EVENT_TYPE Col#: 2 Table: ANONYMOUS Alias: E
NDV: 7 NULLS: 0 DENS: 0.0000e+00
NO HISTOGRAM: #BKT: 1 #VAL: 2
Column: EVENT_ID Col#: 1 Table: ANONYMOUS Alias: E
NDV: 7429352 NULLS: 0 DENS: 0.0000e+00 LO: 91465 HI:

11861688

NO HISTOGRAM: #BKT: 1 #VAL: 2

Column: CREATION_D Col#: 5 Table: ANONYMOUS Alias: E
NDV: 604325 NULLS: 0 DENS: 0.0000e+00 LO: 2453787 HI:

2453878

NO HISTOGRAM: #BKT: 1 #VAL: 2

TABLE: ANONYMOUS ORIG CDN: 13276900 ROUNDED CDN: 1 CMPTD CDN: 0

Access path: index (scan)

Index: ANONYMOUS_CE_IDX
TABLE: ANONYMOUS

RSC_CPU: 0 RSC_IO: 2771

IX_SEL: 9.0000e-03 TB_SEL: 1.2857e-03
Access path: index (index-only)

Index: ANONYMOUS_EEC_IDX
TABLE: ANONYMOUS

RSC_CPU: 0 RSC_IO: 40

IX_SEL: 1.2137e-08 TB_SEL: 1.0923e-10

BEST_CST: 2771.00 PATH: 4 Degree: 1 -- why CBO chooses

the high cost plan here?

OPTIMIZER STATISTICS AND COMPUTATIONS

GENERAL PLANS

Join order[1]: ANONYMOUS[E]#0
Best so far: TABLE#: 0 CST: 2771 CDN: 1 BYTES: 39
prefetching is on for ANONYMOUS_CE_IDX
Final - All Rows Plan:
JOIN ORDER: 1

CST: 2771 CDN: 1 RSC: 2771 RSP: 2771 BYTES: 39

IO-RSC: 2771 IO-RSP: 2771 CPU-RSC: 0 CPU-RSP: 0

*** 2006-07-19 00:44:40.994

What do you think?

Search Discussions

  • Jared Still at Jul 19, 2006 at 3:55 pm

    On 7/19/06, eagle fan wrote:
    hi:

    My oracle version is 9205 and optimizer_features_enable is set to 9.2.0.

    We have a query choose a bad plan.

    ...
    What do you think?
    I think it's a bug.

    Check MetaLink.

    A possible candidate is *Bug 4112254 - CBO may not choose best index for
    single table access with unpeeked bind/function*

    I just searched MetaLink on CBO Bug 9.2.0.5

    You should probably check there.

    --
    Jared Still
    Certifiable Oracle DBA and Part Time Perl Evangelist

    --
    http://www.freelists.org/webpage/oracle-l
  • Eagle fan at Jul 20, 2006 at 5:20 am
    hi:

    I create a empty table with the same table/index structure in 9.2.0.5 &
    8.1.7.4 test environments and exp/imp the table/index statistics into test
    environments.

    The 9i test database still chose the bad index, but *8i test database chose
    the right plan.*

    Even when I set the *ANONYMOUS_CE_IDX index's *statistics to very big value,
    it still chose the bad one.

    I agree that this maybe a 9.2.0.5 bug.
    On 7/19/06, Jared Still wrote:
    On 7/19/06, eagle fan wrote:

    hi:

    My oracle version is 9205 and optimizer_features_enable is set to 9.2.0.

    We have a query choose a bad plan.

    ...
    What do you think?

    I think it's a bug.

    Check MetaLink.

    A possible candidate is *Bug 4112254 - CBO may not choose best index for
    single table access with unpeeked bind/function*

    I just searched MetaLink on CBO Bug 9.2.0.5

    You should probably check there.

    --
    Jared Still
    Certifiable Oracle DBA and Part Time Perl Evangelist
    --
    Eagle Fan

    Oracle DBA

    --
    http://www.freelists.org/webpage/oracle-l
  • Robert Freeman at Jul 19, 2006 at 5:08 pm
    All,

    I have a rather stange problem. I have a database on
    LINUX that we created. We then exported data from a
    VMS database and imported it to the Linux database. I
    can connect just fine in SQL*Plus... but our JDBC
    application is freezing up when trying to issue a
    query.

    What is *really* odd is:

    After about 12 hours of inactivity, it starts
    working ok again. So something is getting reset. We
    have been trying to figure out what it might be, but
    we have yet to be able to do so.

    We created another database, on the same box, with
    the same configuration, and imported the data again
    from the same dump file and we don't have the same
    problem.

    Any ideas???

    Robert G. Freeman
    Author:
    Portable DBA: Oracle (Oracle Press)
    Oracle Database 10g New Features (Oracle Press)
    Oracle9i RMAN Backup and Recovery (Oracle Press)
    Oracle9i New Features (Oracle Press)
    Oracle Replication (Rampant Tech Press)
    Mastering Oracle8i (Sybex)
    Oracle8 to 8i Upgrade Exam Cram (Coriolis )
    Oracle 7.3 to 8 Upgrade Exam Cram (Coriolis )
  • Thomas Day at Jul 19, 2006 at 6:37 pm
    There's not a lot of information here but I'd start by looking at the
    listener configuration.

    Are you sure that it's the database that stops working?
  • Robert Freeman at Jul 19, 2006 at 10:00 pm
    Thought about the listener, but both are configured
    the same except for the port. I can't think of
    anything about the listener that would reset itself
    after about 12 hours and then all of a sudden things
    work fine.....

    Thomas Day wrote:
    There's not a lot of information here but I'd start
    by looking at the
    listener configuration.

    Are you sure that it's the database that stops
    working?
    Robert G. Freeman
    Author:
    Portable DBA: Oracle (Oracle Press)
    Oracle Database 10g New Features (Oracle Press)
    Oracle9i RMAN Backup and Recovery (Oracle Press)
    Oracle9i New Features (Oracle Press)
    Oracle Replication (Rampant Tech Press)
    Mastering Oracle8i (Sybex)
    Oracle8 to 8i Upgrade Exam Cram (Coriolis )
    Oracle 7.3 to 8 Upgrade Exam Cram (Coriolis )
  • Mladen Gogala at Jul 20, 2006 at 12:31 am

    On 07/19/2006 01:08:22 PM, Robert Freeman wrote:
    All,

    I have a rather stange problem. I have a database on
    LINUX that we created. We then exported data from a
    VMS database and imported it to the Linux database. I
    can connect just fine in SQL*Plus... but our JDBC
    application is freezing up when trying to issue a
    query.

    What is *really* odd is:

    1. After about 12 hours of inactivity, it starts
    working ok again. So something is getting reset. We
    have been trying to figure out what it might be, but
    we have yet to be able to do so.

    2. We created another database, on the same box, with
    the same configuration, and imported the data again
    from the same dump file and we don't have the same
    problem.

    Any ideas???
    Rob, it looks like some stupid SELinux or IP_Tables trick.
    Is something blocking you and enforcing some kind of idiotic
    security policy? Turn that off.
  • Robert Freeman at Jul 20, 2006 at 2:21 am
    Don't think so... remember that both of the databases (one working and one
    not working) are on the same machine, though they are using a different
    port. What is also odd is that if the database is left up overnight, then it
    starts working.... if it's shutdown overnight it doesn't start working. We
    killed all the jobs (thinking it was the job scheduler) and that did not
    make a difference. We killed all CRON jobs, still nothing. No remote jobs
    are scheduled.

    It's the darndest thing I've ever seen. Still a few more things to test here
    and there though.

    RF

    -----Original Message-----
    From: Mladen Gogala
    Sent: Wednesday, July 19, 2006 6:32 PM
    To: robertgfreeman_at_yahoo.com
    Cc: oracle-l_at_freelists.org
    Subject: Re: Oracle... one DB connects, one does not...
    On 07/19/2006 01:08:22 PM, Robert Freeman wrote:
    All,

    I have a rather stange problem. I have a database on
    LINUX that we created. We then exported data from a
    VMS database and imported it to the Linux database. I
    can connect just fine in SQL*Plus... but our JDBC
    application is freezing up when trying to issue a
    query.

    What is *really* odd is:

    1. After about 12 hours of inactivity, it starts
    working ok again. So something is getting reset. We
    have been trying to figure out what it might be, but
    we have yet to be able to do so.

    2. We created another database, on the same box, with
    the same configuration, and imported the data again
    from the same dump file and we don't have the same
    problem.

    Any ideas???
    Rob, it looks like some stupid SELinux or IP_Tables trick.
    Is something blocking you and enforcing some kind of idiotic
    security policy? Turn that off.
  • Jeremiah Wilton at Jul 20, 2006 at 2:38 am
    Robert,

    You said:
    I can connect just fine in SQL*Plus... but our JDBC
    application is freezing up when trying to issue a
    query.
    So, on what event are all those JDBC sessions waiting when they are
    "freezing up?" Without that information, we are flying blind!

    Jeremiah Wilton
    ORA-600 Consulting
    http://www.ora-600.net
  • Robert Freeman at Jul 20, 2006 at 2:46 am
    Well.... thats kind of the problem, we are not sure. We know the SQL that is
    getting executed, and we can run it fine from SQL*Plus. The JDBC connection
    is made, and about 3 queries are successful and then the one that we have
    identified as problematic just seems to hang. We sent JDBC traces to Oracle,
    database traces, we have had the vendor here, etc...etc.... and nobody seems
    to have any idea what is causing this problem or exactly WHAT we are hanging
    on. The other problem I have is my general inability to release information

    in public... so my hands are somewhat tied as to what I can say. It's really
    frustrating... I was just thinking someone might have some idea that I'd not
    thought of....

    RF

    -----Original Message-----
    From: Jeremiah Wilton
    Sent: Wednesday, July 19, 2006 8:39 PM
    To: robertgfreeman_at_yahoo.com
    Cc: oracle-l_at_freelists.org
    Subject: RE: Oracle... one DB connects, one does not...

    Robert,

    You said:
    I can connect just fine in SQL*Plus... but our JDBC
    application is freezing up when trying to issue a
    query.
    So, on what event are all those JDBC sessions waiting when they are
    "freezing up?" Without that information, we are flying blind!

    Jeremiah Wilton
    ORA-600 Consulting
    http://www.ora-600.net
  • Jeremiah Wilton at Jul 20, 2006 at 2:52 am
    Robert,

    When you look at v$session_wait for one of the JDBC sessions that is
    hanging, what are the values of EVENT and SECONDS_IN_WAIT? Also is
    SEQUENCE# incrementing?

    That will help narrow it down to either the database or NOT the database.

    --
    Jeremiah Wilton
    ORA-600 Consulting
    http://www.ora-600.net

    Robert Freeman wrote:
    Well.... thats kind of the problem, we are not sure. We know the SQL that
    is getting executed, and we can run it fine from SQL*Plus. The JDBC
    connection is made, and about 3 queries are successful and then the one
    that we have identified as problematic just seems to hang. We sent JDBC
    traces to Oracle, database traces, we have had the vendor here,
    etc...etc.... and nobody seems to have any idea what is causing this
    problem or exactly WHAT we are hanging on....
    --
    http://www.freelists.org/webpage/oracle-l
  • Robert Freeman at Jul 20, 2006 at 4:34 am
    I know I checked waits early on and even ran a 10046 trace. It's been like 2
    weeks ago or more now. We will be doing some testing tomorrow and I'll look
    again. The 10046 trace file for the statement in question shows the EXEC,
    several waits on SQL*Net message to client and SQL*Net message from client
    and SQL*Net more data to client. There is also a db file scattered read
    wait.... but NONE of these waits amounts to much clock time at all (the
    longest is an elapsed of 6953 on the db file scattered read).

    We then get the fetch, with an e=169353, and then the trace simply stops. I
    think it's at this point that something is not right with the driver or the
    data or something. The question is, what and why...

    RF

    -----Original Message-----
    From: Jeremiah Wilton
    Sent: Wednesday, July 19, 2006 8:53 PM
    To: 'Robert Freeman'
    Cc: oracle-l_at_freelists.org
    Subject: RE: Oracle... one DB connects, one does not...

    Robert,

    When you look at v$session_wait for one of the JDBC sessions that is
    hanging, what are the values of EVENT and SECONDS_IN_WAIT? Also is
    SEQUENCE# incrementing?

    That will help narrow it down to either the database or NOT the database.

    --
    Jeremiah Wilton
    ORA-600 Consulting
    http://www.ora-600.net

    Robert Freeman wrote:
    Well.... thats kind of the problem, we are not sure. We know the SQL that
    is getting executed, and we can run it fine from SQL*Plus. The JDBC
    connection is made, and about 3 queries are successful and then the one
    that we have identified as problematic just seems to hang. We sent JDBC
    traces to Oracle, database traces, we have had the vendor here,
    etc...etc.... and nobody seems to have any idea what is causing this
    problem or exactly WHAT we are hanging on....
    --
    http://www.freelists.org/webpage/oracle-l
  • Steve Perry at Jul 20, 2006 at 2:54 am
    are you connecting as "/ as sysdba" or using a service name? Is the
    application on the same server or different server?
    if different server, can you use tcpdump to get any information or
    strace?
    Are you saying it connects fine, but can't issue sql or can't connect?
    I got called from a developer yesterday that couldn't get the
    database. the app was working the day before and nothing had been
    changed. at the end of the day, he said it was a corrupt jdbc driver.
    he said he switched it to the thin client driver and it "magically"
    started working. It doesn't sound like that's the case since you can
    connect to the 2nd db and things are fine.

    Have you tried changing the port of the first database and see the
    app works?
    I've also seen where the database doesn't register to the listener
    and had to issue "alter system register" or put the database in the
    listener.ora. probably only applicable if you can't connect though
    On Jul 19, 2006, at 09:21 PM, Robert Freeman wrote:

    Don't think so... remember that both of the databases (one working
    and one
    not working) are on the same machine, though they are using a
    different
    port. What is also odd is that if the database is left up
    overnight, then it
    starts working.... if it's shutdown overnight it doesn't start
    working. We
    killed all the jobs (thinking it was the job scheduler) and that
    did not
    make a difference. We killed all CRON jobs, still nothing. No
    remote jobs
    are scheduled.

    It's the darndest thing I've ever seen. Still a few more things to
    test here
    and there though.

    RF


    -----Original Message-----
    From: Mladen Gogala
    Sent: Wednesday, July 19, 2006 6:32 PM
    To: robertgfreeman_at_yahoo.com
    Cc: oracle-l_at_freelists.org
    Subject: Re: Oracle... one DB connects, one does not...

    On 07/19/2006 01:08:22 PM, Robert Freeman wrote:
    All,

    I have a rather stange problem. I have a database on
    LINUX that we created. We then exported data from a
    VMS database and imported it to the Linux database. I
    can connect just fine in SQL*Plus... but our JDBC
    application is freezing up when trying to issue a
    query.

    What is *really* odd is:

    1. After about 12 hours of inactivity, it starts
    working ok again. So something is getting reset. We
    have been trying to figure out what it might be, but
    we have yet to be able to do so.

    2. We created another database, on the same box, with
    the same configuration, and imported the data again
    from the same dump file and we don't have the same
    problem.

    Any ideas???
    Rob, it looks like some stupid SELinux or IP_Tables trick.
    Is something blocking you and enforcing some kind of idiotic
    security policy? Turn that off.


    --
    Mladen Gogala
    http://www.mgogala.com

    --
    http://www.freelists.org/webpage/oracle-l
    --
    http://www.freelists.org/webpage/oracle-l
  • Robert Freeman at Jul 20, 2006 at 2:56 am
    We have tried different JDBC clients, no joy there.
    The listener seems to be working fine, though I might just try switching
    ports.
    We are connecting to the two databases from the same environment, so all the
    application components are the same.

    We are using thin JDBC, so its like a direct connect... no service names.

    RF

    -----Original Message-----
    From: Steve Perry
    Sent: Wednesday, July 19, 2006 8:54 PM
    To: robertgfreeman_at_yahoo.com
    Cc: oracle-l_at_freelists.org Freelists
    Subject: Re: Oracle... one DB connects, one does not...

    are you connecting as "/ as sysdba" or using a service name? Is the
    application on the same server or different server?
    if different server, can you use tcpdump to get any information or
    strace?
    Are you saying it connects fine, but can't issue sql or can't connect?
    I got called from a developer yesterday that couldn't get the
    database. the app was working the day before and nothing had been
    changed. at the end of the day, he said it was a corrupt jdbc driver.
    he said he switched it to the thin client driver and it "magically"
    started working. It doesn't sound like that's the case since you can
    connect to the 2nd db and things are fine.

    Have you tried changing the port of the first database and see the
    app works?
    I've also seen where the database doesn't register to the listener
    and had to issue "alter system register" or put the database in the
    listener.ora. probably only applicable if you can't connect though
    On Jul 19, 2006, at 09:21 PM, Robert Freeman wrote:

    Don't think so... remember that both of the databases (one working
    and one
    not working) are on the same machine, though they are using a
    different
    port. What is also odd is that if the database is left up
    overnight, then it
    starts working.... if it's shutdown overnight it doesn't start
    working. We
    killed all the jobs (thinking it was the job scheduler) and that
    did not
    make a difference. We killed all CRON jobs, still nothing. No
    remote jobs
    are scheduled.

    It's the darndest thing I've ever seen. Still a few more things to
    test here
    and there though.

    RF


    -----Original Message-----
    From: Mladen Gogala
    Sent: Wednesday, July 19, 2006 6:32 PM
    To: robertgfreeman_at_yahoo.com
    Cc: oracle-l_at_freelists.org
    Subject: Re: Oracle... one DB connects, one does not...

    On 07/19/2006 01:08:22 PM, Robert Freeman wrote:
    All,

    I have a rather stange problem. I have a database on
    LINUX that we created. We then exported data from a
    VMS database and imported it to the Linux database. I
    can connect just fine in SQL*Plus... but our JDBC
    application is freezing up when trying to issue a
    query.

    What is *really* odd is:

    1. After about 12 hours of inactivity, it starts
    working ok again. So something is getting reset. We
    have been trying to figure out what it might be, but
    we have yet to be able to do so.

    2. We created another database, on the same box, with
    the same configuration, and imported the data again
    from the same dump file and we don't have the same
    problem.

    Any ideas???
    Rob, it looks like some stupid SELinux or IP_Tables trick.
    Is something blocking you and enforcing some kind of idiotic
    security policy? Turn that off.


    --
    Mladen Gogala
    http://www.mgogala.com

    --
    http://www.freelists.org/webpage/oracle-l
    --
    http://www.freelists.org/webpage/oracle-l
  • Andrey Kriushin at Jul 20, 2006 at 1:25 pm
    You can use service names with thin JDBC starting with JDK 1.4.1. The
    syntax is

    "jdbc:oracle:thin:@//host:port/service_name"

    Andrey

    Robert Freeman ÐÉÛÅÔ:
    We are using thin JDBC, so its like a direct connect... no service names.
  • Mladen Gogala at Jul 20, 2006 at 4:53 am

    On 07/19/2006 10:21:37 PM, Robert Freeman wrote:
    Don't think so... remember that both of the databases (one working and one
    not working) are on the same machine, though they are using a different
    port. What is also odd is that if the database is left up overnight, then it
    starts working.... if it's shutdown overnight it doesn't start working. We
    killed all the jobs (thinking it was the job scheduler) and that did not
    make a difference. We killed all CRON jobs, still nothing. No remote jobs
    are scheduled.

    It's the darndest thing I've ever seen. Still a few more things to test here
    and there though.
    Have you tried tracing the listener? Anything in the system log?
  • Wolfgang Breitling at Jul 20, 2006 at 5:42 am
    I don't think it has been volunteered and I'm surprised no one has asked:

    What Oracle version??

    Also, since it seems related to the JDBC connection, which jdbc
    driver are you using and have you tried tracing the jdbc driver and
    connection itself?
    At 11:08 AM 7/19/2006, Robert Freeman wrote:
    All,

    I have a rather stange problem. I have a database on
    LINUX that we created. We then exported data from a
    VMS database and imported it to the Linux database. I
    can connect just fine in SQL*Plus... but our JDBC
    application is freezing up when trying to issue a
    query.
    Regards

    Wolfgang Breitling
    Centrex Consulting Corporation
    www.centrexcc.com
  • Jaromir nemec at Jul 19, 2006 at 7:49 pm
    Hi E.F.,


    We have a query choose a bad plan.
    select /*+ index(e) */


    my experience is, if you use a hint INDEX you should also specify which index should be used. The result of saying only "use some index" could be surprising (see the thread "Does Oracle have a Sense of Humour" in this list for an example).

    In your particular case, Jared is probably right, it is a bug.



    Regards,



    Jaromir

    Original Message -----
    From: eagle fan
    To: oracle-l_at_freelists.org
    Sent: Wednesday, July 19, 2006 10:14 AM
    Subject: Why CBO doesn't choose the low cost plan

Related Discussions

Discussion Navigation
viewthread | post
Discussion Overview
grouporacle-l @
categoriesoracle
postedJul 19, '06 at 8:14a
activeJul 20, '06 at 1:25p
posts18
users10
websiteoracle.com

People

Translate

site design / logo © 2022 Grokbase