ROWID and IOTs

Greetings,
I've got some dynamic SQL that selects based on ROWID. Works OK until the table argument is an IOT.
ORA-01410: invalid ROWIDI have been made aware that IOTs do not carry valid ROWIDs, however, when a ROWID is selected from an IOT, values that can be handled as VARCHAR2 are returned.
SQL> select rowid from my_iot_table;
ROWID
*BAgA7EYCwgsCwQX+
*BAgA7EYCwgsCwQ7+
*BAgA7EYCwgsCwRX+If I handle these ROWIDs as character data, is there any guarantee of uniqueness?
-Kevin

In a single transaction, I've got a bit of (recursive) code that traverses a tree of related rows in various tables. At each node, it caches the current rowid and compares it to those of all visited nodes to determine if any cycles exist in the data.
Rowid seemed to be a convenient test for uniqueness.
-K

Similar Messages

Rowid in IOT table

It seems that rowid in IOT is
created by function involving primary key.
is there any built in function or package
that we use to simulate that ?
null

Think about what a ROWID actually is and it will become clear.
"The rowid of a row specifies the datafile and data block containing the row and the location of the row in that block."

What is the difference between logical ROWID and physical ROWID

hello
can u please explain the difference between logical ROWID and physical ROWID
regards,

from the docs (a 30 sec search)
Physical rowids store the addresses of rows in ordinary tables (excluding index-organized tables), clustered tables, table partitions and subpartitions, indexes, and index partitions and subpartitions.
Logical rowids store the addresses of rows in index-organized tables.

What are ROWID and ROWNUM? Are they stored in database and where?

Hi All,
can anybody please answer this question
What are ROWID and ROWNUM? Are they stored in database and where?
Thanks,
Srini

ROWID can be thought of as a pointer to the physical location (on disk) of the (table) row.
From a ROWID value, Oracle can extract the file, block-within-that-file and offset-of-the-row-within-that-block. Using these, Oracle can directly access a disk block to retrieve a row.
ROWNUM is a just sequence number of a row within a result set of a query.
As said by other repliers, both are not stored. They are 'constructed' when you reference them inside a query.

Rowid and urowid datatypes

I'm writing an exercise to learn more about different datatypes and am confused about the use of and difference between rowid and urowid. As I understand it using rowid should only be done when writing for backwards compatibility, and urowid should be used for new coding. I'm working in 9.2.0.1.0, so I assume I should use urowid.
The following code returns an error (shown below). Anyone know why this is happenning?
create or replace function lrnvrbls
return varchar2
is
v_rowid urowid;
v_data varchar2(100);
select rowid
into v_rowid
from mytable
where rownum = 10 -- 10 is just a number I chose
execute immediate
'select mycolumn ' ||
'from mytable ' ||
'where rowid = ' ||
v_rowid
into v_data;
return v_data;
end;
SQL>select lrnvrbls from dual;
select lrnvrbls from dual
ORA-00904: "AAAHCAAAMAAAADVAAI": invalid identifier
I've tried using both ROWID and UROWID datatypes with no luck. Also tried CHARTOROWID and ROWIDTOCHAR and failed the same way.
Anybody understand how to use this?

A single datatype called the universal rowid, or UROWID, supports both logical
and physical rowids, as well as rowids of foreign tables such as non-Oracle
tables accessed through a gateway.
A column of the UROWID datatype can store all kinds of rowids. The value of the
COMPATIBLE initialization parameter must be set to 8.1 or higher to use UROWID
columns.
DROP TABLE mytable
CREATE TABLE mytable
AS SELECT owner mycolumn FROM all_tables
WHERE rownum < 100
CREATE OR REPLACE
FUNCTION lrnvrbls
RETURN varchar2
IS
v_rowid urowid;
v_data varchar2(100);
strSQL varchar2(255);
BEGIN
select rowid
into v_rowid
from mytable
where rownum = 1;
strSQL := 'select mycolumn from mytable where rowid = '''||v_rowid ||'''';
execute immediate strSQL into v_data;
return v_data;
END lrnvrbls;
SELECT lrnvrbls FROM DUAL
DROP FUNCTION lrnvrbls
DROP TABLE mytable
COMMIT
16:39:37 SQL> DROP TABLE mytable
16:39:39 2 /
Table dropped.
Elapsed: 00:00:00.00
16:39:39 SQL> --
16:39:39 SQL> CREATE TABLE mytable
16:39:39 2 AS SELECT owner mycolumn FROM all_tables
16:39:39 3 WHERE rownum < 100
16:39:39 4 /
Table created.
Elapsed: 00:00:00.00
16:39:39 SQL> --
16:39:39 SQL> CREATE OR REPLACE
16:39:39 2 FUNCTION lrnvrbls
16:39:39 3 RETURN varchar2
16:39:39 4 IS
16:39:39 5 --
16:39:39 6 v_rowid urowid;
16:39:39 7 v_data varchar2(100);
16:39:39 8 strSQL varchar2(255);
16:39:39 9 --
16:39:39 10 BEGIN
16:39:39 11 --
16:39:39 12 select rowid
16:39:39 13 into v_rowid
16:39:39 14 from mytable
16:39:39 15 where rownum = 1;
16:39:39 16 --
16:39:39 17 strSQL := 'select mycolumn from mytable where rowid = '''||v_rowid ||'''';
16:39:39 18 --
16:39:39 19 execute immediate strSQL into v_data;
16:39:39 20 --
16:39:39 21 return v_data;
16:39:39 22 --
16:39:39 23 END lrnvrbls;
16:39:39 24 /
Function created.
Elapsed: 00:00:00.00
16:39:39 SQL> --
16:39:39 SQL> SELECT lrnvrbls FROM DUAL
16:39:39 2 /
LRNVRBLS
SYS
Elapsed: 00:00:00.00
16:39:39 SQL> --
16:39:39 SQL> DROP FUNCTION lrnvrbls
16:39:39 2 /
Function dropped.
Elapsed: 00:00:00.01
16:39:40 SQL> DROP TABLE mytable
16:39:40 2 /
Table dropped.
Elapsed: 00:00:00.00
16:39:40 SQL> COMMIT
16:39:40 2 /
Commit complete.
Elapsed: 00:00:00.00
16:39:40 SQL>
*/

VALIDATE_LAYER_WITH_CONTEXT - Rowid and NULL in result table

Hello,
Oracle 10g R2
I performed validity test using VALIDATE_LAYER_WITH_CONTEXT as follow:
CREATE TABLE RESULT_TABLE_CP(SDO_ROWID ROWID, RESULT VARCHAR2(2000));
CALL SDO_GEOM.VALIDATE_LAYER_WITH_CONTEXT('VESSEL_ZONE_GEOMETRY', 'POLYGON', 'RESULT_TABLE_CP');
What puzzled me was the result table contained the rowid and NULL as the result for each geometry. I was expecting to see what the error is, not just NULL.
However when I use VALIDATE_GEOMETRY_WITH_CONTEXT to test the same layer, I got TRUE for all of them, as:
select DISTINCT SDO_GEOM.VALIDATE_GEOMETRY_WITH_CONTEXT(a.POLYGON,b.DIMINFO)
from VESSEL_ZONE_GEOMETRY a,
USER_SDO_GEOM_METADATA b
where b.TABLE_NAME = 'VESSEL_ZONE_GEOMETRY'
and b.COLUMN_NAME = 'POLYGON'
and a.POLYGON is not null;
SDO_GEOM.VALIDATE_GEOMETRY_WITH_CONTEXT(A.POLYGON,B.DIMINFO) ------------------------------------------------------------------------------------
TRUE
Any suggestion would be much appreciated!
Thanks, CP

Can you get the geometry by running
select polygon from VESSEL_ZONE_GEOMETRY where rowid = '....';

How to get the rowid and the entire record selected in PLSQL?

The code given below does not work.
You cannot select the record and the rowid using a cursor in one-shot.
But you could do that in a direct SELECT in "SQL Plus" when you select records for display. Is this a bug in ORACLE PLSQL? Or is there another way to do this?
DECLARE
objid_ VARCHAR2(200);
rec_ xxx_tab%ROWTYPE;
CURSOR get_rec IS
SELECT t.*, t.rowid
FROM xxx_tab t;
BEGIN
OPEN get_rec;
FETCH get_rec INTO rec_, objid_;
CLOSE get_rec;
END;
-----------------------------------

You cannot fetch into both a record type and a variable. You have a few options, you can declare the record a s a rowtype of the cursor like this:
DECLARE
   CURSOR c IS
      SELECT t.*, rowid rid
      FROM t;
   l_rec c%ROWTYPE;
BEGIN
   OPEN c;
   FETCH c INTO l_rec;
   CLOSE c;
END;You could use an implicit cursor and let Oracle deal with the record type internally (not to mention the open fetch and close) like this:
BEGIN
   FOR rec in (SELECT t.*, rowid rid FROM t) LOOP
      do_stuff with rec.col_name
   END LOOP;
END;Note that in both of these you must alias the rowid column to some other name, you could also manually construct the record type to match the table and add a column od ROWID datatype to hold the rowid.
Finally, I think, depending on what you are actually going to do with the rowid, and how you feel about having records locked, you could look at declaring the cursor as FOR UPDATE and get rhe rowid for free.. This would be most appropriate if you are planning to update the table in the cursor (a bad practce by the way). Something like:
DECLARE
   l_rec t%ROWTYPE;
   CURSOR c IS
      SELECT t.*, rowid
      FROM t;
BEGIN
   OPEN c;
   FETCH c INTO l_rec;
   do_whatever with l_rec
   UPDATE t
   SET whatever
   WHERE current of c;
END;John

What's instance, ROWID and why we use cursor?

Dear memebers,
yesterday i have an interview there i faced the following questions.
1. what is instance?
2. what is ROWID?
3. why we use cursor?
i can give only the last one.
i am waiting for your answer.
thanks
Regards:
Muhammad Nadeem
[email protected]

HY
Row Identifiers
Each row in an Oracle database has a unique row identifier, or rowid, which is used internally by the Oracle database to access the row. A rowid is an 18-digit number that is represented as a base-64 number, and it contains the physical address of a row in an Oracle database. You can view the rowid value for rows in a table by specifying the ROWID column in the select list of a query. The query in the following example retrieves the ROWID and customer_id columns from the customers table:
SELECT ROWID, customer_id
FROM customers;
Oracle instance:
An Oracle instance is the combination of the background processes and
memory structures. The instance must be started to access the data in the database. Every
time an instance is started, a System Global Area (SGA) is allocated and Oracle
background processes are started. Background processes perform functions on behalf of
the invoking process. They consolidate functions that would otherwise be handled by
multiple Oracle programs running for each user. The background processes perform
input/output (I/O) and monitor other Oracle processes to provide increased parallelism
for better performance and reliability.
REGARDS

Partitioned tables and IOT

Hi All,
We had a database using IOT but now because of the performance reasons
we want to elimnate IOT table and using the Parititioned Tables.
Have you ever using the partitioned tables for ORACLE with your database?
If you have done that please share your experiences or refer me some
documents related to Partitioned tables for ORACLE to make my job more
easier.
Thanks in advance,
JP

Hi,
You will get good information from this Oracle site,
http://www.oracle.com/technology/documentation/index.html.
Thanks

MAX(ROWID) and inserting rows at the end of the table

Hi,
I know this will be another lame question but I simply didn't find the answer anywhere. I create ID (primary key) for rows manually so I have this code:
String query = "SELECT * FROM CIS_SEG_SKUPINY WHERE ROWID=(SELECT MAX(ROWID) FROM CIS_SEG_SKUPINY)";
rset = stmt.executeQuery(query);
if (!rset.next()) {
newID = new Number(1);
System.out.println(newID);
else {
rset.beforeFirst();
while (rset.next()) {
int lastID = rset.getInt("ID");
System.out.println(lastID);
newID = new Number(lastID + 1);
System.out.println(newID);
}
}
css.setId(newID);
Row globalsRow = getGlobals().first();
css.setKodSegSkupiny((String)globalsRow.getAttribute("GroupCode"));
css.setPopis((String)globalsRow.getAttribute("Description"));
css.setBarva((Number)globalsRow.getAttribute("Colour"));
getDBTransaction().commit();
When the table is empty newID is set to 1 and row inserted. Now I have one row in the table (with ID=1) so newID is set to 2 and row inserted BUT it is inserted BEFORE the first row. So when I try to insert third row lastID is 1 not 2. I tried to find some command which would insert new row after the last one but unfortunately I wasn't successfull.
Edited by: Herald on Mar 28, 2009 11:11 AM

Herald,
I am sorry to say that there are so many things wrong with your approach, that I don't know where to begin...
1). ROWID is a pseudo column in Oracle.. It has no meaning of first, last, middle, or otherwise. SELECT MAX(ROWID) will get you some row.
2). In a relational database, there is no concept of first or last row in a table. If you expect the rows in some order, use an order by.
3). Ignoring your implementation and addressing your intent - you should not be creating a primary key by getting the max already in the table and adding one. You will, using your technique, get duplicates in a multi-user scenario. If you fix your implementation to not give duplicates in a multi-user scenario (no, I won't show you how), you will use a table lock, and therefor inhibit scalability in a serious way.
The proper and commonly accepted way to generate a numeric primary key in Oracle would be to use a sequence and a trigger to populate the sequence. Lots of examples to be found in this forum, but just have a read of the ADF Developers Guide for Forms/4GL Developers (for 10g) or the Fusion Developers Guide (for 11g) to find one. The term to search for would be DBSequence.
John

ROWID and rownum

Hello World!
I have a large table I want to trancate to 10 rows.
I use the staement:
delete from aa where rowid in (select rowid from (select rowid from aa) where rownum > 11);
This result in 0 records deleted due to select rowid from (select rowid from aa) where rownum > 11; results in 0 records selected.
At the same time select rowid from (select rowid from aa) where rownum < 17;
produce correct result.
Any suugestions?
TIA

You should NOT use ";" symbol at the end of sql statement you call using
EXECUTE IMMEDIATE.
Change your function (and use binding variable whereever it's possible):
SQL> CREATE or replace function delete_rows (
2 table_name IN VARCHAR2,
3 nrows IN NUMBER DEFAULT NULL) return NUMBER IS
4 l_sql VARCHAR2(32767);
5 BEGIN
6
7 IF TABLE_NAME IS NOT NULL and nrows IS NOT NULL THEN
8 l_sql := 'delete from '||table_name ||
9 ' where rowid not in (select rowid from (select rowid from '|
10 table_name||') where rownum < :1)';
11 EXECUTE IMMEDIATE l_sql using nrows;
12 END IF;
13
14 commit;
15
16 return null;
17
18 EXCEPTION
19 WHEN OTHERS THEN
20 DBMS_OUTPUT.PUT_LINE(SQLERRM);
21 return null;
22 END delete_rows;
23 /
Function created.
SQL> select delete_rows('emp',11) from dual;
DELETE_ROWS('EMP',11)
ORA-14551: cannot perform a DML operation inside a query
SQL> begin
2 dbms_output.put_line(delete_rows('emp',11));
3 end;
4 /
PL/SQL procedure successfully completed.Rgds.

Rowid and insert

If I insert a row, on return, the cursor structure has the rowid. Is there any way I can, using the cursor structure, get the rowid of the inserted row without doing a select?
Please reply also to [email protected]
Thanks in advance for your help.
Elly Sokol
[email protected]

Something like this perhaps:
declare
   foo varchar2(4000);
begin
   insert into emp( empno )
     values( 2 )
     returning rowidtochar(rowid) into foo;
   dbms_output.put_line( foo );
end;Justin
Distributed Database Consulting, Inc.
http://www.ddbcinc.com/askDDBC

Oracle - ROWID (Physical and logical)

Hi,
Could you please clarify the below doubts?
- Difference between Physical and Logical ROWID's
- Difference between ROWID and UROWID datatypes
Regards,
Ragav

Instead of an answer, may I suggest you go through following link?
http://download.oracle.com/docs/cd/B19306_01/server.102/b14220/datatype.htm#sthref3894

Using Rownum and ROwid returns duplicate records

Hi All,
We have implemented pagination as below using rowid and rownum
SELECT
id
FROM
emp
WHERE
ROWID IN
SELECT RID FROM (SELECT
ROWID RID,
ROWNUM RNUM
FROM
SELECT ID FROM emp
WHERE
((T_ID IN (200005,200229,200230,200249,200250,200049))) AND
(dte >= sysdate-90) AND
(LOWER(DESC) = LOWER ('A') AND
LOWER(NVL(FLAG,'0')) != LOWER ('3') AND
LOWER(MODDE) like LOWER ('%210%')) ORDER BY dte ASC ))
WHERE ROWNUM < 11) WHERE RNUM>= 1)) ORDER BY dte emp.ASC
But, we face that - the query inserts duplicate records in consecutive pages. For Eg:
1.if a,b,c,d,e - is returned for first iteration, then for the next iteration - f,g,h,a,y is returned.
Is it because that the Order by clause doesnt have a Unique key column.
Please help. or suggest how to efficietly implement pagination without performance hit

try distinct since you are using only one column it will eliminate any duplicates.
SELECT distinct id
FROM emp
WHERE (ROWID IN ( SELECT RID
 FROM (SELECT ROWID RID,ROWNUM RNUM
 FROM (SELECT ID
 FROM emp
 WHERE ((T_ID IN (200005,200229,200230,200249,200250,200049)))
 AND (dte >= sysdate-90)
 AND (LOWER(DESC) = LOWER ('A')
 AND LOWER(NVL(FLAG,'0')) != LOWER ('3')
 AND LOWER(MODDE) like LOWER ('%210%'))
 ORDER BY dte ASC ))
 WHERE ROWNUM < 11)
 WHERE RNUM>= 1))
ORDER BY dte emp.ASC

Which one faster- With Rowid or PK

Hi,
Recently came across the concepts of IOT(Index organised table). It was mentioned that
You must specify a primary key for an index-organized table, because the primary key uniquely identifies a row. Use the primary key instead of the Rowid for directly
accessing index-organized rows. "
This surprised me as till date I was assuming that the fastest way to access the data from a table is by Rowid. Isn't this contadicting then?
Can the experts clarify more on these conceps related to IOT.
Rgds,
Aashish

what is the diff betn a Normal table with PK on some column and IOT with almost
the same structure?The difference is whatever is implied by that "almost".
Look, the purpose of index organized tables is to do away with a nugatory object when a table and and its primary key index - have basically the same structure. The classic example is the code lookup table, where the difference between the table's columns (code, code_descr) and its primary key (code) is a single column (code_descr). Nearly access of that table is going to be an indexed read on code to get the description. That usage makes it a good candidate for being an IOT.
Mere creation of PK on a column would give the same result.No. An index organized table is similar to creating an index on (code, code_descr) but without the need to maintain two sets of data. Furthermore the IOT allows us to include non-primary key columns without compromising our relational integrity.
Cheers, APC
Blog : http://radiofreetooting.blogspot.com/

ROWID and IOTs

Similar Messages

Maybe you are looking for