Re: Taking export from SECONDARY_INSTANCE from 02 node RAC

From: Michael Austin <maustin_at_firstdbasource.com>
Date: Fri, 27 Feb 2009 16:40:21 -0600
Message-ID: <11_pl.11548$pr6.4807_at_flpi149.ffdc.sbc.com>

jgar the jorrible wrote:

> On Feb 27, 10:35 am, Michael Austin <maus..._at_firstdbasource.com>
> wrote:

>> Mladen Gogala wrote:

>>> On Thu, 26 Feb 2009 19:29:07 -0600, Michael Austin wrote:
>>>> I would caveat any discussion about licensing with "it all depends".
>>> Yes, of course. Oracle sales force is not unlike the car dealership.
>>> What, you think that $15,000 is too much for this only 4 years old
>>> Ford Taurus? No problem, I will lower the price, just for you, it's
>>> only $12,000  and I will throw in new tires, too. That is especially
>>> true if you are a big shop with a big name. I've seen some deals with
>>> cutting the prices in half - literally. Unfortunately, these days
>>> everybody is in dire need of money, especially the banks. Oracle cannot
>>> afford slashing the prices by much without layoffs.
>>> I do predict, however, that, during this crisis, many shops will turn to
>>> cheaper alternatives like PostgreSQL and MySQL, thus eventually forcing
>>> Oracle to cut the prices. My boss, an avid Oracle fan, started a pilot
>>> project for an office database with MySQL. I am an old oracle hack with
>>> many years of experience but I did accept the task to learn and manage
>>> MySQL. Should that go well, I am sure that we will see those projects
>>> multiply like rabbits. That is the thing that will eventually force Oracle
>>> to lower the prices and give some things like AWR report for free.
> 
> Many jobs available for many projects that will fail, I'm sure.
> 
> When most of Oracle's gross profit percentage comes from apps, Oracle
> will... make AWR an app.
> 
>>> I must say, though, that Oracle plans look so much nicer then this:
>>> mysql> explain select ar.artist_name,al.album_name
>>>     -> from artist ar,album al
>>>     -> where ar.artist_id=al.artist_id;
>>> +----+-------------+-------+------+---------------+---------+---------
>>> +---------
>>> -----------+------+-------+
>>> | id | select_type | table | type | possible_keys | key     | key_len |
>>> ref
>>>            | rows | Extra |
>>> +----+-------------+-------+------+---------------+---------+---------
>>> +---------
>>> -----------+------+-------+
>>> |  1 | SIMPLE      | ar    | ALL  | PRIMARY       | NULL    | NULL    |
>>> NULL
>>>            |    6 |       |
>>> |  1 | SIMPLE      | al    | ref  | PRIMARY       | PRIMARY | 2       |
>>> music.ar
>>> .artist_id |    1 |       |
>>> +----+-------------+-------+------+---------------+---------+---------
>>> +---------
>>> -----------+------+-------+
>>> 2 rows in set (0.20 sec)
>>> mysql>
>>> This is a simple NESTED LOOPS join.

>> And there are many other performance issues as well as standards issues..

>>

>> Try this in mysql and oracle: (and unless they fixed it- which they have
>> not in my server...) will be very troubling to REAL database folks...
>> Especially if you need to move data from MySQL to say a mainframe or
>> COBOL app :-( And BTW per their crack[ed?] developer staff - this is
>> NOT a BUG because it is "documented behavior".. How whacked is that!!!

>>

>> create table mytest (a char(10));
>> insert into mytest values ('ABCDE');
>> select '|'||a||'|' from mytest;
>> select '|'||cast(a as CHAR(20))||'|' from mytest;

>>

>> ORACLE: (correct result!!)
>> SQL> create table mytest (a char(10));
>> insert into mytest values ('ABCDE');
>> select '|'||a||'|' from mytest;
>> select '|'||cast(a as CHAR(20))||'|' from mytest;

>>

>> Table created.

>>

>> SQL>
>> 1 row created.

>>

>> SQL>
>> '|'||A||'|'
>> ------------
>> |ABCDE | <<--- note there are 5 chars+5spaces as per the data type.

>>

>> SQL>
>> '|'||CAST(AASCHAR(20))
>> ----------------------
>> |ABCDE |
>> ===========================================================================­===

>>

>> MySQL - the INCORRECT RESULT

>>

>> mysql> create table mytest (a char(10));
>> Query OK, 0 rows affected (1.62 sec)

>>

>> mysql> insert into mytest values ('ABCDE');
>> Query OK, 1 row affected (0.38 sec)

>>

>> mysql> select '|'||a||'|' from mytest;
>> +-------------+
>> | '|'||a||'|' |
>> +-------------+
>> | |ABCDE| | <<--- note there are 5 chars only - no spaces.
>> +-------------+
>> 1 row in set (0.24 sec)

>>

>> mysql> select '|'||cast(a as char(20))||'|' from mytest;
>> +-------------------------------+
>> | '|'||cast(a as char(20))||'|' |
>> +-------------------------------+
>> | |ABCDE| |
>> +-------------------------------+
>> 1 row in set (0.11 sec)

> 
> Preaching to the choir here.  Try complaining to this guy:
> http://monty-says.blogspot.com/
> 

[thought I would go visit the bug database to see the status before commenting further...]

It was logged as a bug... #24424 [19 Nov 2006 5:28] - apparently they added a parameter to change this behavior Oct 10th 2008 to versions  >=5.1 - the last time I responded to the bug was Jan 2008.

SET sql_mode = 'PAD_CHAR_TO_FULL_LENGTH';

So it took them almost 2 years to figure out how what a data type is...

> jg
> --
> _at_home.com is bogus.
> If at first you don't succeed... write a book, be profiled in
> newpapers and Reader's Digest, get on tv...
> http://www3.signonsandiego.com/stories/2009/feb/27/1m27silver235851-familys-sailing-adventure-now-sub/?uniontrib

Received on Fri Feb 27 2009 - 16:40:21 CST