Archive

Archive for the ‘sql’ Category

Do you know the ORA- nonerrors?

August 17th, 2010

In one of my script, I am checking at the end for any ORA- error. And if I have any ORA- error, I quit with an error.

So far so good.

Also when I run a report from the shell, I do set the sqlplus settings I like and I expect the script to receive the output with no ORA- error at the beginning.

But watch this !

$ echo "set feed off hea off newp none
conn scott/tiger
select * from dual;"|sqlplus /nolog
ERROR:
ORA-28002: the password will expire within 10 days

X

My password will expire in 10 days. And this screw up my day :(

sql, sqlplus

Hardcoding SYSDATE

August 16th, 2010

I see TRUNC(SYSDATE) in the code very often. Is this good or bad?

Over my past two years as a developer, I can tell you it is pretty bad. What does TRUNC(SYSDATE) mean? It is today’s date.

But when does today starts and ends? And at the time of the writing, it is already tomorrow in Thailand.

We could argue about using CURRENT_DATE instead of sysdate, but it does not solve everything. Maybe you want your day to end at 10:30pm and initialize for the next day after some post processing. Maybe you want some components running in different timezones. Maybe you want to rollback and rollforward the date in your test systems…

Also the bad news, if your report hardcodes SYSDATE, kind of SELECT SUM(AMT) FROM T WHERE EXPIRY_DATE>TRUNC(SYSDATE), you will not be able to run it tomorrow in case it fails today.

No, I think that you would better store the date in a table and update it daily.

So you will have SELECT SUM(T.AMT) FROM T, TRADINGDAY WHERE T.EXPIRY_DATE>TRADINGDAY.TRADINGDAY and a daily job running UPDATE TRADEDAY SET TRADINGDAY=TRADINGDAY+1;.

You may want to consider the performance overhead of reading from a table instead of using the very performant SYSDATE function…

plsql, sql

connect by and recursive with (part 2)

May 20th, 2010

According to the doc
The subquery_factoring_clause now supports recursive subquery
factoring (recursive WITH), which lets you query hierarchical data.
This feature is more powerful than CONNECT BY in that it
provides depth-first search and breadth-first search, and supports
multiple recursive branches. A new search_clause and cycle_clause
let you specify an ordering for the rows and mark cycles in the
recursion

As written there and there, recursive with has more capabilities. It is also ANSI and implemented in DB2, MS SQL Server. CONNECT BY is an Oracle oddity. But does recursive with perfoms as well as connect by?

  1. Create a big emp
  2. 
    create table big_emp as
    with t(empno,mgr) as
    (select 1, null from dual
    union all 
    select empno+1,trunc(dbms_random.value(1+empno/10,empno))
    from t 
    where empno<100000)
    select * from t
    

  3. execution plan
  4. CONNECT BY

    
    select empno,mgr
    from big_emp
    connect by mgr = prior empno
    start with mgr is null;
    
    E M 
    - -
    1 -  
    2 1 
    4 2 
    6 2 
    8 6 
    
    Operation                 Object    Rows Time Cost   Bytes
    ------------------------- ------- ------ ---- ---- -------
    SELECT STATEMENT                       3    3  185      78
    CONNECT BY WITH FILTERING
    TABLE ACCESS FULL         BIG_EMP      1    1   61      10
    HASH JOIN                              2    2  122      46
    CONNECT BY PUMP
    TABLE ACCESS FULL         BIG_EMP 100000    1   61 1000000
    

    recursive WITH

    
    with e(empno,mgr) as (
    select empno, mgr 
    from big_emp 
    where mgr is null
    union all
    select f.empno,f.mgr 
    from big_emp f, e 
    where e.empno=f.mgr)
    select empno,mgr
    from e;
    
    E M
    - -
    1 -  
    2 1 
    3 1 
    4 2 
    5 3 
    ...
    
    Operation                 Object    Rows Time Cost   Bytes
    ------------------------- ------- ------ ---- ---- -------
    SELECT STATEMENT                       3    3  183      78
    VIEW                                   3    3  183      78
    UNION ALL (RECURSIVE WITH) BREADTH FIRST
    TABLE ACCESS FULL BIG_EMP              1    1   61      10
    HASH JOIN                              2    2  122      46
    RECURSIVE WITH PUMP
    TABLE ACCESS FULL BIG_EMP         100000    1   61 1000000
    

    In this particular simple case, it seems CONNECT BY have a 1% higher cost.

  5. execution time
  6. CONNECT BY

    
    select sum(mgr)
    from 
    (
    select empno,mgr
    from big_emp
    connect by mgr = prior empno
    start with mgr is null
    )
    
    SUM(MGR) 
    ------------
    2745293877 
    
    1 rows returned in 0.73 seconds 
    

    recursive with

    
    with e(empno,mgr) as (
    select empno, mgr 
    from big_emp 
    where mgr is null
    union all
    select f.empno,f.mgr 
    from big_emp f, e 
    where e.empno=f.mgr)
    select sum(mgr)
    from e;
    
    SUM(MGR) 
    ------------
    2745293877 
    
    1 rows returned in 1.24 seconds 
    

Honestly I am not surprised that CONNECT BY is faster, CONNECT BY has been in the Oracle database forever and has been massively tuned in 8.1.7.4.

And then one day you find, ten years have got behind you.

11gR2, sql

CONNECT BY and Recursive CTE

May 19th, 2010

11gR2 introduced a new mechanism to build up hierarchies.

I remembered a thread in developpez.net that reveals the dubious implementation of nocycle in 10g.

For the CONNECT BY ISLEAF, I have read the technique on amis.nl.

Ok, here is my graph

The 10g query


with o as
(
SELECT 'A' obj, 'B' link from dual union all
SELECT 'A', 'C' from dual union all
SELECT      'C', 'D' from dual union all
SELECT           'D', 'C' from dual union all
SELECT           'D', 'E' from dual union all
SELECT                'E', 'E' from dual)
select connect_by_root obj root,level,obj,link,
  sys_connect_by_path(obj||
'->'
||link,','),
  connect_by_iscycle,
  connect_by_isleaf
from o 
connect by nocycle obj=prior link
start with obj='A';

ROOT LEVEL O L PATH                 CYCLE  LEAF
---- ----- - - -------------------- ----- -----
A        1 A B ,A->B                    0     1
A        1 A C ,A->C                    0     0
A        2 C D ,A->C,C->D               1     0
A        3 D E ,A->C,C->D,D->E          1     1

Obviously in 10g the connect by nocycle does works well with that kind of graphs, D-C and E-E are missing and C-D and D-E are marked as cycling…

Let’s try the 11gR2 equivalency.


with o(obj,link) as
(
SELECT 'A', 'B' from dual union all
SELECT 'A', 'C' from dual union all
SELECT      'C', 'D' from dual union all
SELECT           'D', 'C' from dual union all
SELECT           'D', 'E' from dual union all
SELECT                'E', 'E' from dual),
t(root,lev,obj,link,path) as (
select obj,1,obj,link,cast(obj||'->'||link 
as varchar2(4000))
from o 
where obj='A'  -- START WITH
union all
select 
  t.root,t.lev+1,o.obj,o.link,
  t.path||', '||o.obj||
    '->'
    ||o.link
from t, o 
where t.link=o.obj
)
search depth first by obj set ord
cycle obj set cycle to 1 default 0
select root,lev,obj,link,path,cycle,
    case
    when (lev - lead(lev) over (order by ord)) < 0
    then 0
    else 1
    end is_leaf
 from t;

ROOT LEV  OBJ  LINK PATH                        CYCLE IS_LEAF
---- ---- ---- ---- --------------------------- ----- -------
A    1    A    B    A->B                            0       1
A    1    A    C    A->C                            0       0
A    2    C    D    A->C, C->D                      0       0
A    3    D    C    A->C, C->D, D->C                0       0
A    4    C    D    A->C, C->D, D->C, C->D          1       1
A    3    D    F    A->C, C->D, D->E                0       0
A    4    F    F    A->C, C->D, D->E, E->E          0       0
A    5    F    F    A->C, C->D, D->E, E->E, E->E    1       1

It looks good :)

If you exclude the rows with cycle=1, you get the six rows for the graph.

11gR2, sql

number series

May 17th, 2010

Patrick Wolf wrote about the newest Apex release, which contains a 11.2 db engine, so I had to play with recursive queries ;)


with t(x) as (select 1 from dual 
union all
select x+1 from t where x<5 )
select x from t;

X
1
2
3
4
5

with t(x,y) as (select 1 x, 1 y from dual 
union all
select x+1,y*(x+1) from t where x<5 )
select x,y "X!" from t;

X  X!
1  1
2  2
3  6
4  24
5  120

with t(r,x,y) as (select 1,1,1 from dual 
union all
select r+1,y,x+y from t where r<5)
select x fib from t

FIB 
1 
1 
2 
3 
5 

with t1(x) as (select 1 from dual 
union all
select x+1 from t1 where x<4),
t2(x,y,z) as (select x, 1,x from t1
union all 
select x,y+1,x*(y+1) from t2 where y<2)
select listagg(z,';') within group (order by x) s from t2 group by y;

S 
1;2;3;4 
2;4;6;8 

11gR2, sql

where is the TRIGGER ANY TABLE privilege?

May 13th, 2010

You have your table data in one schema and your procedures in another one. But can you have triggers and tables in different schemas?


SYS@lsc01> create user u1 identified by u1;

User created.

SYS@lsc01> create user u2 identified by u2;

User created.

SYS@lsc01> grant create table, unlimited tablespace to u1;

Grant succeeded.

SYS@lsc01> grant create session, create trigger to u2;

Grant succeeded.

SYS@lsc01> create table u1.t(x number);

Table created.

SYS@lsc01> grant select on u1.t to u2;

Grant succeeded.

SYS@lsc01> connect u2/u2
Connected.
U2@lsc01> create trigger u2.tr after insert on u1.t for each row
  2  begin
  3  null;
  4  end;
  5  /
create trigger u2.tr after insert on u1.t for each row
                                        *
ERROR at line 1:
ORA-01031: insufficient privileges

What’s the missing privilege? To create a trigger on another schema, you need the CREATE ANY TRIGGER privilege.

 
U2@lsc01> connect / as sysdba
Connected.
SYS@lsc01> grant CREATE ANY TRIGGER to u2;

Grant succeeded.

SYS@lsc01> 
SYS@lsc01> connect u2/u2
Connected.
U2@lsc01> create trigger u2.tr after insert on u1.t for each row
  2  begin
  3  null;
  4  end;
  5  /

Trigger created.

dba, sql

where c not in (:b1, :b2, … , :b9999)

March 18th, 2010

I do not like this kind of dynamic NOT IN clauses. It is better to have a temporary table, a bit like in


create global temporary table t(x number not null);
insert into t(x) values (:b1);
insert into t(x) values (:b2);
...
insert into t(x) values (:b9999);
select foo from bar where c not in (select x from t);

If you want to however do this in one query you can still use AND


SQL> select foo from bar where c not in (
:b1,
:b2,
...
:b9999);
*
ERROR at line 1002:
ORA-01795: maximum number of expressions in a list is 1000

SQL> select foo from bar where 
c!=:b1 and
c!=:b2 and
...
c!=:b9999;

FOO
---
foo

Translate c NOT IN (exprlist) into c!=expr1 and c!=expr2…
Translate c IN (exprlist) into c=expr1 or c=expr2…

If you use dynamic expression list, this will bypass the ORA-01795 error

sql

on analytics and superaggregation

February 28th, 2010

When I wrote my book, I did not expect having the advanced Oracle sql features available on other dabatase engine.

This week-end I downloaded for fun a try of db2, v9.7.
1) download db2 linux64bit trial on ibm.com
2) install enterprise edition (next-next-install principle)
3) create the instance
4) create the sample database (where database and instance means something different than in Oracle)

$ db2 create db test pagesize 32 k
$ db2 connect to test
$ db2 create user temporary tablespace temp
$ db2 connect reset

5) connect with clpplus, which looks similar to sqlplus, even cooler at first look
$ clpplus db2inst1/***@vaio:50001/TEST

I tried a few examples from my book. Left, full and cross join works, but not partitioned outer join.

Scalar, Inline and nested query works too.

Surprisely many function that I thought Oracle specific work too, like sysdate and to_char.

In aggegation, the KEEP FIRST/LAST does not work,
In Super aggregations, all GROUPING SETS, CUBE, ROLLUP do work.
No Pivot or Unpivot.

No dual, but a powerful TABLE function

SQL> select * from table(values(1),(2));       
          1
-----------
          1
          2

Analytics works, With range, rows, order by partition, all the fancy stuff.

I have not checked at XML, but it sounds to be differently implemented.
Some things will work however

SQL> select xmlquery('1 to 10') from dual;

1                   
--------------------
1 2 3 4 5 6 7 8 9 10

No CONNECT BY, but as in Oracle 11gR2, recursive query factoring does build the hierarchy

No Model, obviously.

sql

what is the type of NULL

February 9th, 2010

I was a bit surprised to see a VARCHAR2(0) column in my schema


select column_name, data_type, data_length 
from user_tab_columns 
where table_name='V';
COLUMN_NAME     DATA_TYPE            DATA_LENGTH
--------------- -------------------- -----------
X               VARCHAR2                       0

What’s this datatype? It is the datatype of NULL !!!

SQL> create or replace view v as select null x from dual
View created.
SQL> desc v
VIEW v
 Name              Null?    Type        
 ----------------- -------- ------------
 X                          VARCHAR2(0) 

sql

How many decimals do you need?

January 21st, 2010

Do you user NUMBER or NUMBER(p,s) for your datatypes?

I posted last year about 1!=1

Today I realized this could be solved with the scale :)

SQL> drop table lsc_t
Table dropped.
SQL> create table lsc_t(x number, y number(*,6))
Table created.
SQL> insert into lsc_t values (1/3*3,1/3*3)
1 row created.
SQL> commit
Commit complete.
SQL> select * from lsc_t where x=1
no rows selected.
SQL> select * from lsc_t where y=1

         X          Y
---------- ----------
1.00000000          1

So if you have amounts in US$ or GB£ or CHF, do not use NUMBER, use NUMBER(*,6) or NUMBER(*,2) or whatever is relevant to your business!

sql

.plz dump file

October 6th, 2009

What are those .plz dump files in my user dump directory ?


-rw-r--r--   1 oracle   dba        15168 Oct  6 14:34 _anon__3ca8c5e38__AB.plz
-rw-r-----   1 oracle   dba        15883 Oct  6 14:45 db01_ora_10061.trc
-rw-r--r--   1 oracle   dba        15168 Oct  6 14:45 _anon__3c929b088__AB.plz
-rw-r-----   1 oracle   dba        15895 Oct  6 14:47 db01_ora_10666.trc
-rw-r--r--   1 oracle   dba        15168 Oct  6 14:47 _anon__3c8651198__AB.plz

let’s check one, briefly :


*** ASSERT at file pdw4.c, line 2080; Type 0xffffffff7d79fb40 has no MAP method.
Source Location = _anon__3d2474b28__AB[1, 7]

== Dump of OPT Context Object ffffffff7c519ec8. ==
  Tue Oct  6 16:31:11 2009
  Event 10944 = 0
  plsp          = ffffffff7fff67a8
  lu            = 3c9c18210
  Diana root    = 0x20014 = 131092
  Diana proc    = 0x20012 = 131090
  Graph         = ffffffff7d774d70
  Dump file     = /app/oracle/admin/DB01/udump/_anon__3d2474b28__AB.plz
  CG            = 0
...

It is a kind of dump file, apparently. I could not find details on metalink. I generate the one above in 10.2.0.4 sparc with the following code

create or replace type t1 as object (x number)
/
create or replace type t2 as table of t1
/
exec if t1(1)member of t2()then null;end if

BEGIN if t1(1)member of t2()then null;end if; END;
Error at line 10
ORA-06550: line 1, column 7:
PLS-00801: internal error [*** ASSERT at file pdw4.c, line 2080; Type 0xffffffff7d7ba280 has no MAP method.; _anon__3c929b088__AB[1, 7]]

dba, sql

stragg in 11gR2

September 1st, 2009

This will be a killer in the Oracle forums ;)
listagg

LISTAGG (measure_expr [, 'delimiter_expr'])
  WITHIN GROUP (order_by_clause) [OVER query_partition_clause]

for instance

SELECT LISTAGG(ename,',') WITHIN GROUP (ORDER BY ename) FROM emp;
ADAMS,ALLEN,BLAKE,CLARK,FORD,JAMES,JONES,KING,MARTIN,MILLER,SCOTT,SMITH,TURNER,WARD

11gR2, sql

le compte est bon

August 19th, 2009

I am back from my vacations, I was at nice places in Switzerland like Rhone Glacier, underground lake of Saint-Leonard, Salt Mines of Bex, Rhine Waterfalls and more …

To keep up with the fun, here is a little quiz :

You have the numbers 1-3-4-6 and you need to achieve the number 24. The allowed operations are +, -, * and /

If I try to achieve 49 it is easy :

SQL> /
Enter value for n1: 1
old  14:      (SELECT &n1 n
new  14:      (SELECT 1 n
Enter value for n2: 3
old  17:       SELECT &n2 n
new  17:       SELECT 3 n
Enter value for n3: 4
old  20:       SELECT &n3 n
new  20:       SELECT 4 n
Enter value for n4: 6
old  23:       SELECT &n4 n
new  23:       SELECT 6 n
Enter value for result: 49
old 143:              ) = &result
new 143:              ) = 49

result
------------------------------------
(4+3)*(6+1)
(3+4)*(6+1)
(6+1)*(4+3)
(1+6)*(4+3)
(6+1)*(3+4)
(1+6)*(3+4)
(4+3)*(1+6)
(3+4)*(1+6)

8 rows selected.

Elapsed: 00:00:11.28

But for 24 it is not that simple :-) at least for human !

Ok, in SQL I am using a plsql function to evaluate expression

CREATE OR REPLACE FUNCTION lsc_eval (expr VARCHAR2)
   RETURN NUMBER
IS
   x   NUMBER;
BEGIN
   EXECUTE IMMEDIATE 'begin :x := ' || expr || ';end;'
               USING OUT x;
   RETURN x;
EXCEPTION
   WHEN OTHERS
   THEN
      RETURN NULL;
END;
/

I will post the rest of the code as a comment later ;)

fun, game, sql

CTAS and NOT NULL

July 13th, 2009

When you create a table as select (CTAS), you lose a lot of information like grants, partitioning, organization, referential integrity, check constraints. But the NOT NULL constraints remain. Sometimes

Let’s see when the not null constraints are not copied to the new table.

Here is the test case :

create table lsc_t1(
  c0 number constraint C_PRIMARY_KEY primary key,
  c1 number constraint C_DEFERRABLE not null deferrable, 
  c2 number constraint C_NOVALIDATE not null novalidate,
  c3 number constraint C_DISABLE not null disable,
  c4 number constraint C_DISABLE_VALIDATE not null disable validate, 
  c5 number constraint C_NOT_NULL not null  
  );
create table lsc_t2 as select * from lsc_t1;

let’s describe the tables :

SQL> desc lsc_t1
 Name                          Null?    Type
 ----------------------------- -------- --------------------
 C0                            NOT NULL NUMBER
 C1                                     NUMBER
 C2                                     NUMBER
 C3                                     NUMBER
 C4                            NOT NULL NUMBER
 C5                            NOT NULL NUMBER

SQL> desc lsc_t2
 Name                          Null?    Type
 ----------------------------- -------- --------------------
 C0                                     NUMBER
 C1                                     NUMBER
 C2                                     NUMBER
 C3                                     NUMBER
 C4                            NOT NULL NUMBER
 C5                            NOT NULL NUMBER

The NOT NULL of c0 has been lost. C0 is the primary key, and the primary key is not transferred to the target table. We can see the non-deferrable validated not-null constraints c4 and c5.

Let’s compare all the constraints :

select * from user_constraints where table_name in ('LSC_T1','LSC_T2');

  LSC_T1 LSC_T2
C0 PRIMARY KEY

 
C1 DEFERRABLE

 
C2 ENABLED NOT VALIDATED

 
C3 DISABLED NOT VALIDATED

 
C4 DISABLED VALIDATED

ENABLED VALIDATED
C5 ENABLED VALIDATED

ENABLED VALIDATED

The deferrable and not-validated check constraints are lost. The regular ENABLE VALIDATE constraint is ok, but the DISABLE VALIDATE constraint is now enabled.

sql ,

How to tune WHERE NAME LIKE ‘%ABC%’

July 9th, 2009

More than once customers wanted me to tune queries where the LIKE clause made the query very slow…

The easy answer is : you cannot. If you want to search for some characters in the middle of the string, Oracle will not use an index.

Is it a correct answer? Maybe not.

Let’s imagine my application allow you to search for a string in the middle of the name, but to avoid scanning too large amount of data, the application enforces the string to be at least 3 characters long. For example ABC.

In this case, instead of doing a full table scan of the table to retrieve only a few rows, we can use an hint to tell Oracle to use an index on the name:

CREATE TABLE lsc_t AS
SELECT
  ROWNUM ID,
  SUBSTR(DBMS_RANDOM.STRING ('n', DBMS_RANDOM.VALUE (2, 8))
    || ' '
    || DBMS_RANDOM.STRING ('n', DBMS_RANDOM.VALUE (2, 8))
    ,1,17) NAME,
  TRUNC(SYSDATE-ABS(10000*DBMS_RANDOM.NORMAL)) birthdate,
  LPAD('X',4000,'X') address
FROM DUAL
CONNECT BY LEVEL <= 1e5;

ALTER TABLE lsc_t ADD PRIMARY KEY(ID);

CREATE INDEX lsc_i ON lsc_t(NAME);

EXEC dbms_stats.gather_table_stats(user,'LSC_T',cascade=>true)

let’s measure the time for a full table scan

SQL> set timi on
SQL> SELECT ID, NAME, birthdate
  FROM lsc_t WHERE NAME LIKE '%ABC%';

       ID NAME              BIRTHDATE
--------- ----------------- ---------
    60249 ABCBIFAB KRKBCRN  11-MAR-90
    16714 AF YABCG          09-OCT-95
    55571 BABCIQ GESGLW     27-MAR-50
    77561 BP GABC           24-APR-90
    80027 DALSABC TZLOAWDV  05-NOV-01
    49817 EABCTFIY XWB      10-FEB-88
    23283 EMMOGGBF DABCB    20-DEC-87
    39530 FMABCKB AB        18-SEP-87
    68605 FTPGOHE ABCC      28-SEP-28
    74615 KIFDWABC CSSUQ    08-AUG-82
    31772 KNOABCT BO        08-SEP-77
    68730 KRYIEN LMABC      10-APR-07
    43317 LUFJKZJT AUABCZR  19-DEC-88
    76851 MZABC TEIFG       14-SEP-92
    54589 NXE YABCDX        03-MAY-88
     6940 OIWABCZ DLFFXY    29-MAR-88
    59070 ONIB ADGABCI      29-JUL-07
    27264 PGHOABC ZY        05-OCT-90
    38157 QABC OPZHE        13-JUN-87
    17511 QPDKD CIABCJ      08-AUG-69
    25507 RX OWULOABC       24-FEB-92
    62159 SEABC DAILK       25-JUN-02
     3845 SK CCABCG         22-JAN-80
    50059 SPABC BVHRHW      18-MAR-86
    54700 UABCPC WUHAJS     28-OCT-71
    70207 UKY OIDUABC       23-APR-88
    39484 WABC TJLYHVJZ     14-MAR-78
    14561 WDRWABC XZKDH     29-MAR-86
    61501 YBYU RYABCGI      28-JUN-78
    30578 YEWENGX ABCHARA   12-SEP-67
    35397 YHBEABC HFKO      25-AUG-85
    26450 YOABCVG HJT       23-DEC-98
    87224 ZKNLNY YAABC      13-NOV-61

33 rows selected.

Elapsed: 00:00:02.56

about 3 seconds for retrieving 33 rows out of 100000

let’s try with an index

SQL> SELECT /*+INDEX(LSC_T,LSC_I)*/  ID, NAME, birthdate
  FROM lsc_t WHERE NAME LIKE '%ABC%';

       ID NAME              BIRTHDATE
--------- ----------------- ---------
    60249 ABCBIFAB KRKBCRN  11-MAR-90
    16714 AF YABCG          09-OCT-95
    55571 BABCIQ GESGLW     27-MAR-50
    77561 BP GABC           24-APR-90
    80027 DALSABC TZLOAWDV  05-NOV-01
    49817 EABCTFIY XWB      10-FEB-88
    23283 EMMOGGBF DABCB    20-DEC-87
    39530 FMABCKB AB        18-SEP-87
    68605 FTPGOHE ABCC      28-SEP-28
    74615 KIFDWABC CSSUQ    08-AUG-82
    31772 KNOABCT BO        08-SEP-77
    68730 KRYIEN LMABC      10-APR-07
    43317 LUFJKZJT AUABCZR  19-DEC-88
    76851 MZABC TEIFG       14-SEP-92
    54589 NXE YABCDX        03-MAY-88
     6940 OIWABCZ DLFFXY    29-MAR-88
    59070 ONIB ADGABCI      29-JUL-07
    27264 PGHOABC ZY        05-OCT-90
    38157 QABC OPZHE        13-JUN-87
    17511 QPDKD CIABCJ      08-AUG-69
    25507 RX OWULOABC       24-FEB-92
    62159 SEABC DAILK       25-JUN-02
     3845 SK CCABCG         22-JAN-80
    50059 SPABC BVHRHW      18-MAR-86
    54700 UABCPC WUHAJS     28-OCT-71
    70207 UKY OIDUABC       23-APR-88
    39484 WABC TJLYHVJZ     14-MAR-78
    14561 WDRWABC XZKDH     29-MAR-86
    61501 YBYU RYABCGI      28-JUN-78
    30578 YEWENGX ABCHARA   12-SEP-67
    35397 YHBEABC HFKO      25-AUG-85
    26450 YOABCVG HJT       23-DEC-98
    87224 ZKNLNY YAABC      13-NOV-61

33 rows selected.

Elapsed: 00:00:00.06

Much better :)

reposted due to % in url

dba, sql ,

on recycle bin

July 7th, 2009

more than one user may wondered who created those BIN$ when they first connected to a 10g database.


create table lsc_t(x number) 
partition by range(x) 
(partition LESS_THAN_ONE values less than (1));

drop table lsc_t;

select object_name, subobject_name, created 
from user_objects 
where object_name like 'BIN$%';
OBJECT_NAME                    SUBOBJECT_NAME  CREATED  
------------------------------ --------------- ---------
BIN$bh2VJ6FqFJ3gRAAUT+rFpg==$0 LESS_THAN_ONE   07-JUL-09

Ok, it is quite easy to get rid of it. Either at DROP time with a DROP TABLE LSC_T PURGE or later with PURGE RECYCLEBIN. Most of the objects disappear from USER_OBJECTS when dropped actually. The recyclebin view is called : RECYCLEBIN.


purge recyclebin;

select object_name, subobject_name, created 
from user_objects 
where object_name like 'BIN$%';

no rows selected.

select * from recyclebin;

no rows selected.

So far so good…

Let’s see what’s happening with my primary keys


purge recyclebin;
create table lsc_t(x number constraint lsc_t_pk primary key);
drop table lsc_t;
select object_name, original_name, type from recyclebin;

OBJECT_NAME                    ORIGINAL_NAME TYPE 
------------------------------ ------------- -----
BIN$bh23ggtBHALgRAAUT+rFpg==$0 LSC_T         TABLE
BIN$bh23ggtAHALgRAAUT+rFpg==$0 LSC_T_PK      INDEX

The primary key index is now in the recycle bin as well.

let’s recover the recycle bin version :


flashback table lsc_t to before drop;
select index_name from user_indexes where table_name='LSC_T';

INDEX_NAME                    
------------------------------
BIN$bh3GgNi1HR3gRAAUT+rFpg==$0

select object_name, original_name, type from recyclebin;

no rows selected.

:evil:

So if you cannot exclude BIN$ objects from your dba maintenance scripts, you will need to deal with thoses as they may be recovered indexes!

dba, sql

select from comma-separated list

July 3rd, 2009

This is asked over and over in the forums, but why not proposing an 11g solution here ;)

create table t(description varchar2(12) primary key, 
  numbers varchar2(4000));
insert into t(description, numbers) values ('PRIME','2,3,5,7');
insert into t(description, numbers) values ('ODD','1,3,5,7,9');
commit;

DESCRIPTION NUMBERS
PRIME 2,3,5,7
ODD 1,3,5,7,9

Now I want to unpivot numbers in rows


select description,(column_value).getnumberval()  
from t,xmltable(numbers)

DESCRIPTION (COLUMN_VALUE).GETNUMBERVAL()
PRIME 2
PRIME 3
PRIME 5
PRIME 7
ODD 1
ODD 3
ODD 5
ODD 7
ODD 9

It is that simple :)

Works also with strings :


select (column_value).getstringval() 
from xmltable('"a","b","c"');

(COLUMN_VALUE).GETSTRINGVAL()
a
b
c

11g, sql, xml

to_number(to_char(:n))

June 9th, 2009

Is it safe to do a TO_NUMBER(TO_CHAR(:n)) ?

var n number
exec :n :=  9.9999999999999999999999999999999999e125
select to_number(to_char(:n)) from dual;
ORA-01722: invalid number

Why?
SQL Reference
If a positive NUMBER value is extremely large and cannot be represented in the specified format, then the infinity sign (~) replaces the value

select to_char(:n) from dual;
TO_CHAR(:N)
--------------
~

I wrote more about infinity here

sql

committing transaction

June 4th, 2009

When and how do you commit a transaction?

SQL> insert into lsc_t(x) values(1);         ... (1) not committed
SQL> commit;                                 ... (1) committed

SQL> insert into lsc_t(x) values(2);         ... (2) not committed
SQL> alter table lsc_t disable primary key;  ... (2) committed

SQL> set autocommit ON
SQL> insert into lsc_t(x) values(3);         ... (3) committed
SQL> set autocommit OFF

SQL> insert into lsc_t(x) values(4);         ... (4) not committed
SQL> disc                                    ... (4) committed

Row 1 is committed after the commit keyword.

Row 2 is implicitely committed after before the alter table ddl statement.

Row 3 is autocommitted. Autocommit exists in sqlplus, but is more often seen in stateless applications (for instance a web application).

Row 4 is committed after a successfull disconnection.

But what appends if you kill your current session?


SQL> insert into lsc_t(x) values (5);

1 row created.

SQL> !ps
   PID TTY      TIME CMD
 13903 pts/33   0:00 ksh
 22384 pts/33   0:00 sqlplus

SQL> !kill 22384
Terminated

The row 5 is not commited

Ref: Overview of Transaction Management

A transaction ends when any of the following occurs:
- A user issues a COMMIT or ROLLBACK statement without a SAVEPOINT clause.
- A user runs a DDL statement such as CREATE, DROP, RENAME, or ALTER (*).
If the current transaction contains any DML statements, Oracle Database
first commits the transaction, and then runs and commits the DDL statement
as a new, single statement transaction.
- A user disconnects from Oracle Database. The current transaction is committed.
- A user process terminates abnormally. The current transaction is rolled back.

(*) but not ALTER SESSION

sql

CONNECT BY NOCYCLE

May 25th, 2009

the idea of this post was largely inspired by nuke_y on developpez.net (in French)

I will try to rephrase it in a hierarchical way.

Let’s take all managers of Adams in EMP :

SELECT     empno, ename, mgr
      FROM emp
CONNECT BY NOCYCLE PRIOR mgr = empno
START WITH ename = 'ADAMS';
     EMPNO ENAME             MGR
---------- ---------- ----------
      7876 ADAMS            7788
      7788 SCOTT            7566
      7566 JONES            7839
      7839 KING                 

So far so good. Let’s imagine SCOTT has two managers, JONES and PAUL.

create table lsc_t AS
     (SELECT ename, empno, mgr
        FROM emp
      UNION ALL
      SELECT 'SCOTT', 7788, 9999
        FROM DUAL
      UNION ALL
      SELECT 'PAUL', 9999, NULL
        FROM DUAL);
SELECT     empno, ename, mgr
      FROM lsc_t
CONNECT BY NOCYCLE PRIOR mgr = empno
START WITH ename = 'ADAMS';
     EMPNO ENAME             MGR
---------- ---------- ----------
      7876 ADAMS            7788
      7788 SCOTT            7566
      7566 JONES            7839
      7839 KING                 
      7788 SCOTT            9999
      9999 PAUL                 

EMP is not supposed to allow this (primary key on empno), so I created a table LSC_T.

So far still fine, Paul is in my list of managers.

Let’s imagine Scott is the manager of Paul. In 9i and below, this would result in an ORA-0146 CONNECT BY loop in user data, but in 10g, and according to the doc The NOCYCLE parameter instructs Oracle Database to return rows from a query even if a CONNECT BY loop exists in the data.

Should I try ???

UPDATE lsc_t
   SET mgr = 7788
 WHERE ename = 'PAUL';
COMMIT ;
SELECT     empno, ename, mgr
      FROM lsc_t
CONNECT BY NOCYCLE PRIOR mgr = empno
START WITH ename = 'ADAMS';
     EMPNO ENAME             MGR
---------- ---------- ----------
      7876 ADAMS            7788
      7788 SCOTT            7566
      7566 JONES            7839
      7839 KING                 
      7788 SCOTT            9999

Due to a loop in user data, PAUL is no longer returned. I have no explanation so far.

sql

Eastern challenge

April 9th, 2009

I just wanted to give to my readers a bit fun for Eastern…

Using the well known EMP table, retrieve the name of the job that is the less common…

The solution must be a one-liner SELECT statement (max 80 chars) :)

Have fun!

sql

not a group by expression?

April 1st, 2009

How do I count all objects in the current schema?


SQL*Plus: Release 10.2.0.1.0 - Production on Wed Apr 1 09:58:46 2009

Copyright (c) 1982, 2005, Oracle.  All rights reserved.

Connected to:
Oracle 12g Enterprise Edition Release 12.1.0.1.0 - 64bit Production
With the Expensive option

SQL> select
  2    user,
  3    count(*)
  4  from
  5    user_objects
  6  where
  7    status='INVALID';
  user,
  *
ERROR at line 2:
ORA-00937: not a single-group group function

Disconnected from Oracle 12g Enterprise Edition Release 12.1.0.1.0 - 64bit Production
With the Expensive option

It does not seem to work … What’s wrong?

sql

locking and database link

March 31st, 2009

Last week-end a new release went to production. This week we got plenty of ORA-2049. What in the hell went wrong? And why did not we find out this locking before?

We just did not have that concurrency effect in test.

How does locking work ?

With regular tables:
session 1
select * from lsc_t for update;
session 2
update lsc_t set id=3;

Session 2 waits until transaction in session 1 finishes.

But in case the update is using a subquery that access a remote table, the update will fail after 1 minute

session 1
select * from lsc_t for update;
session 2
update lsc_t set x=(select * from lsc_t@db02);
Session 2 will end with ORA-02049: timeout: distributed transaction waiting for lock

The timeout of one minute is defined by distributed_lock_timeout, and it is not modifiable in your session (chances are, you will not be able to modify this parameter).

sql

1!=1

March 26th, 2009

In the process of validating some calculation, I noticed some strange differences from one run to another run, with exactly the same data and exactly the same function.

I suppose the only difference was the order in which the transactions were processed (and since I had no ORDER BY, the order was not deterministic).

To limit it to the minimum, I noticed that, in Oracle, 1/3*3 is not 1 !


SQL> select * from dual where 1/3*3=1
no rows selected.

Let me tell you, this was unexpected. Ok, on my first pocket calculator 30 years ago, I noticed than 1/3=.3333333 and .3333333*3=.9999999. But since then, I used to expect 1.

perl :


$ perl -e 'print 1 if 1/3*3==1'
1

OK, I tried in java


public class X {
  public static void main(String[] args){
    if (1/3*3<1) {
      System.out.println("<n");
    }
  }
}
$ javac X.java
$ java X
<

In java it seems different too.

So what is 1/3*3 in Oracle?


SQL> set numw 50
SQL> select 1/3*3 from dual;

                                             1/3*3
--------------------------------------------------
         .9999999999999999999999999999999999999999

Surprised?

sql

package version control

March 24th, 2009

Oracle does not offer any kind of version control like CVS or subversion in the database. CVS and subversion could be used in the filesystem, then the code could be deployed with sqlplus.

To quickly compare packages in Test and Production I used :


select 
  env,name,type,line,text 
from (
  select 
    env,name,type,line,text,
    count(distinct text) over(partition by name,type,line) c
  from (
    select 
      'TEST' env,name,type,line,text 
    from 
      user_source 
    union all
    select 
      'PROD' env,name,type,line,text 
    from 
      user_source@prod
  )
)
where c>1 ;

where “PROD” is a database link to production database. My security friends will scream to hear I have a database link to production in my test schema, but you are free to do it the other way round :mrgreen:

I also explained TOAD method in how to compare schema a few months ago.

dba, sql

commit suicide;

March 16th, 2009

Tanel Poder directed me to Miladin Modrakovic blog, I will definitely add this to my prefered feed.

You have to try this :


SQL> alter session set events 'immediate crash';
alter session set events 'immediate crash'
*
ERROR at line 1:
ORA-03113: end-of-file on communication channel

A very decent method for killing yourself :twisted:

dba, sql

ORA-2070

March 12th, 2009

Today I received an error which was totally new to me. And totally unexpected…

For the purpose of this post, I reduced it to a minimum

CREATE TABLE LSC_T1
( x number);
CREATE TABLE LSC_T2
( x number);
with c as (select nls_charset_decl_len(1, x) 
  from lsc_t1 join lsc_t2 
  using (x))
select 1 from c
union 
select 1 from c;

Error at line 1
ORA-02070: database  does not support operator NLS_CHARSET_DECL_LEN in this context

My basic query was actually joining two dictionary views, and NLS_CHARSET_DECL_LEN is used in USER_TAB_COLUMNS.CHAR_COL_DECL_LENGTH.
Kind of

with c as (
  select table_name, column_name, constraint_name 
  from user_tab_columns 
  join user_constraints  
  using (table_name)
  where constraint_type='P'
)
select 
  constraint_name,column_name 
from c
union all
select 
  column_name,constraint_name 
from c;
Error at line 1
ORA-02070: database  does not support operator NLS_CHARSET_DECL_LEN in this context

I spent some time to understand the reason for getting an ORA-2070 in a simple query that did seem to be totally unrelated to charset. The database used for this test is 9.2.0.8/Solaris.

It is quite easy to avoid the bug, for instance by not using the ANSI join ;)

When analyzing such a defect, my approach is to simplify the query as much as possible. Very often on forums you see users posting queries of more than 100 lines. Those users just expect the community to debug their code. Posting a tiny reproducible test case is key to receive a proper solution!

sql

dynamic database link

March 10th, 2009

How do I select thru a database link, where the database link is not fixed?


SQL> var db_link varchar2(255);
SQL> exec :db_link := 'DB02';
SQL> select x from LSC_T@:db_link where x>0;  
select x from LSC_T@:db_link where x>0;  
                    *
ERROR at line 1:
ORA-01729: database link name expected

My solution using dynamic SQL
Remote databases DB02


create table lsc_t(x number);
insert into lsc_t(x) values (2);
commit;

Remote databases DB03


create table lsc_t(x number);
insert into lsc_t(x) values (3);
commit;

Local database DB01

create or replace type lsc_t_o as object(x number);
/
create or replace type lsc_t_c as table of lsc_t_o;
/
create or replace function lsc_f(str varchar2) 
return lsc_t_c is  
  rc lsc_t_c;  
begin  
  execute immediate 
    'select lsc_t_o(x) from '
    || dbms_assert.qualified_sql_name(str) bulk collect into rc;  
  return rc;  
end;  
/

Note the DBMS_ASSERT function. The DBMS_ASSERT just enhances the security by prevent SQL injection

Ok, let’s try

SQL> var db_link varchar2(255);
SQL> exec :db_link := 'DB02';
SQL> select x from table(lsc_f('LSC_T@'
  ||:db_link)) where x>0;  
         X
----------
         2
SQL> exec :db_link := 'DB03';
SQL> select x from table(lsc_f('LSC_T@'||:db_link)) where x>0;  
         X
----------
         3

Seems to work :)

sql

date or timestamp literals

February 24th, 2009

Yesterday I commented on a post on the French developer forums http://www.developpez.net about using DATE and TIMESTAMP literals.

A date literal is DATE '2000-01-01'. It is not possible to have a date literal at 6am, like DATE '2000-01-01 06:00:00', the syntax simply does not exist. You can either use TO_DATE function, or if you are really hooked on litterals, maybe DATE '2000-01-01'+INTERVAL '+0 06:00:00' DAY TO SECOND.

One suggestion was to use TIMESTAMP, because TIMESTAMP literals do have time in it.

However, comparing a date with a timestamp will result in an implicit conversion to timestamp, and if the date colum is indexed, the index will not be used.

SQL> create table lsc_t(x number, d date)
Table created.
SQL> create index lsc_i on lsc_t(d)
Index created.
SQL> insert into lsc_t(d) values (to_date('2000-01-01 06:00:00',
  'YYYY-MM-DD HH24:MI:SS'))
1 row created.
SQL> select * from lsc_t where d>=timestamp '2000-01-01 06:00:00'
         X D
---------- -------------------
           2000-01-01 06:00:00
Execution Plan
----------------------------------------------------------
   0       SELECT STATEMENT Optimizer Mode=CHOOSE
   1    0    TABLE ACCESS FULL TGPOWNER.LSC_T
SQL> select * from lsc_t where d>=to_date('2000-01-01 06:00:00',
  'YYYY-MM-DD HH24:MI:SS')
         X D
---------- -------------------
           2000-01-01 06:00:00

Execution Plan
----------------------------------------------------------
   0       SELECT STATEMENT Optimizer Mode=CHOOSE
   1    0    TABLE ACCESS BY INDEX ROWID TGPOWNER.LSC_T
   2    1      INDEX RANGE SCAN TGPOWNER.LSC_I

As shown above, the index is not used when comparing to timestamp.

Why does Oracle convert the date to timestamp ? Because timestamp may contain nanoseconds.

SQL> select * from lsc_t where 
  d>=timestamp '2000-01-01 06:00:00.000000001';

no rows selected

SQL> select * from lsc_t where 
  cast(d as timestamp)>=timestamp '2000-01-01 06:00:00.000000001';

no rows selected

SQL> select * from lsc_t where 
  d>=cast(timestamp '2000-01-01 06:00:00.000000001' as date);

         X D
---------- -------------------
           2000-01-01 06:00:00

Converting to date would deliver wrong result.

sql

Tuning query over database link

February 11th, 2009

I just learnt from a colleague a very useful hint for remote databases. Over a db link, Oracle does not have the necessary statistics to take the right decision.

Here is the test case :
DB01


create database link lsc_link_2 using 'DB02';
create table lsc_small_1 as 
  select rownum x from dual connect by level<100;
alter table lsc_small_1 add primary key(x);
create table lsc_big_1 as 
  select rownum x from dual connect by level<1000000;
alter table lsc_big_1 add primary key(x);

DB02


create table lsc_small_2 as 
  select rownum x from dual connect by level<100;
alter table lsc_small_2 add primary key(x);
create table lsc_big_2 as 
  select rownum x from dual connect by level<1000000;
alter table lsc_big_2 add primary key(x);

and here are the two test queries (run on DB01)


select count(*) from lsc_big_1 b1 
  join lsc_small_2@lsc_link_2 s2 using (x);
  COUNT(*)
----------
        99
Elapsed: 00:00:00.10

select count(*) from lsc_big_2@lsc_link_2 b2 
  join lsc_small_1 s1 using (x);
  COUNT(*)
----------
        99
Elapsed: 00:00:10.31

As shown above, first one is 100x faster. Anything to tune? For sure!

The trick is to execute the join remotely when the remote table is much larger than the local one!


select /*+ DRIVING_SITE(b2) */ count(*) 
  from lsc_big_2@lsc_link_2 b2 
  join lsc_small_1 s1 using (x);
  COUNT(*)
----------
        99
Elapsed: 00:00:00.06

Ref: the DRIVING_SITE hint

Uncategorized, sql

track DDL changes

January 16th, 2009

Why is my package being invalidated? This is the question I asked myself a few times those days. In order to find out what is happening in my schema, I created this simple DDL trigger which tracks all DDL changes in my schema.


CREATE TABLE AUDIT_DDL (
  d date,
  OSUSER varchar2(255),
  CURRENT_USER varchar2(255),
  HOST varchar2(255),
  TERMINAL varchar2(255),
  owner varchar2(30),
  type varchar2(30),
  name varchar2(30),
  sysevent varchar2(30),
  sql_txt varchar2(4000));
   
create or replace trigger audit_ddl_trg after ddl on schema
declare
  sql_text ora_name_list_t;
  stmt VARCHAR2(4000) := '';
  n number;
begin
  if (ora_sysevent='TRUNCATE')
  then
    null;
  else
    n:=ora_sql_txt(sql_text);
    for i in 1..n
    loop
      stmt:=substr(stmt||sql_text(i),1,4000);
    end loop; 
    insert into audit_ddl(d, osuser,current_user,host,terminal,owner,type,name,sysevent,sql_txt)
    values(
      sysdate,
      sys_context('USERENV','OS_USER') ,
      sys_context('USERENV','CURRENT_USER') ,
      sys_context('USERENV','HOST') , 
      sys_context('USERENV','TERMINAL') ,
      ora_dict_obj_owner,
      ora_dict_obj_type,
      ora_dict_obj_name,
      ora_sysevent,
      stmt
    );
  end if;
end;
/

Thanks Yas for your comment, I added the SQL text

dba, security, sql