Executing Stored Procedure / Function From Database Link?
Feb 24, 2011
I have 2 databases:
Database A
Database B
In Database A I have a series of config tables and stored procedures/functions In Database B I have a lot of tables.
I would like to execute my stored procedures and all associated functions in database A on my data in database B.
Now I've figured out that creating a database link enables me to do SQL selects on data in both A and B...but how do I run SP/Funcs ? I've read something about packages but not sure if I'm heading in the right direction.
Do I need to create a simple synonym ? Can I use the existing DB link ? or is there a special way of calling them, or...
I like the A/B database set up since I can keep battle tested code in one location and have it work across multiple dbs...saves me having to create everything in every database.
There are 2 Oracle databases with pseudo names Remote and Local. I have a function in Remote called FUS.F_Return_10 which simply returns 10 for testing purposes, where FUS is a schema name. In Local I want to create a procedure that will call the above function. Here's the PL/SQL:
CREATE OR REPLACE PROCEDURE TEST ( V_COUNT OUT NUMBER ) AS V_FOO NUMBER(2,0); BEGIN
[Code]...
There's a Public Database Link called PER_ACC in Local. When I try to create this procedure I get: Encountered symbol "@" when expecting one of the following: .(*%&................
Procedure: CREATE OR REPLACE procedure test (a number, b varchar2) is
begin dbms_output.put_line(a ||'->'||b); end;
Anyonymous Block:
begin exec test(1,'m'); end; / When i run this i am getting this error ORA-06550: line 2, column 7: PLS-00103: Encountered the symbol "TEST" when expecting one of the following [code]...
I am trying to execute the SQLPLUS command (CONNECT) from the stored procedure.It is throwing below error message.
Stored Procedure:
SQL> select user from dual 2 ; USER ------------------------------ SYS SQL> create or replace 2 PROCEDURE PROCEDURE1 AS 3 BEGIN 4 sqlplus sys/sys@D as sysdba; 5 --execute immediate 'create user 'kkk' identified by 'kkk'; 6 END PROCEDURE1; 7 /
Warning: Procedure created with compilation errors.
SQL> show err Errors for PROCEDURE PROCEDURE1: LINE/COL ERROR -------- ----------------------------------------------------------------- 3/9 PLS-00103: Encountered the symbol "SYS" when expecting one of the following: := . ( @ % ; The symbol ":=" was substituted for "SYS" to continue. [code]....
Interviewer asked me "Tell me Diff. between Stored procedure vs. Function ".....I given technical answer which is mentioned in my Faq..But he asked me , dont gv me answer in technical manner..He was interested in which case u use Stored procedure and Function....
I have wrote a Stored Procedure Function that get all the rows from a Staging Table and assigns it to a CLOB and returns the CLOB. The issue is I'm getting the dreaded
ORA-06502: PL/SQL: numeric or value error ORA-06512: at "F_CLOB_TEST", line 21 ORA-06512: at line 7
The error does not occur when I remove the 'PRC_ID', 'S_FIL_NAME' & 'exportDt' from the query. It works fine. The PRC_ID has data such as "700$702$7 05$706$707$708$709$710$711$712$713$714$715$294404$294405$294407$294408$294409$294410"
and S_FIL_NAM is the same for all columns - "SBENE_FILID810-2010-04-07-10.59.17"
The query returns 829 rows. Also I have to include a few more columns in the query which have data larger than the 'PRC_ID', but I have not included them here in the sample code, as this code by itself returns the ORA-6502 error.
create or replace FUNCTION "F_CLOB_TEST" (job_id in Number) return clob is c_clob clob; [code]....
I need to call the VB function below from a Procedure's PL/SQL code and capture the returned variable into a varchar2 variable.I looked at the several means and nothing seems to work.
in a 9.2.0 db I create a db link but does not function :
CREATE DATABASE LINK "ONEDB" CONNECT TO "user1" IDENTIFIED BY "****" USING 'ONEDB';
SQL> select * from dual@"ONEDB"; select * from dual@"ONEDB" * ERROR : ORA-12154: TNS:could not resolve service nameBut on the same server I can connect to 11g in sqlplus :
Z:>sqlplus /nolog SQL*Plus: Release 11.2.0.3.0 Production on Mar. Juin 19 10:13:45 2012 Copyright (c) 1982, 2011, Oracle. All rights reserved.
SQL> connect user1@ONEDB password : Connected.Of cours the other DB links toward 9.2.0 databases work well.
I want Oracle stored function/procedure to calculate number of working days between two dates. We need to exclude Firdays and Saturdays as there are weekend holidays and also exclude official holidasy that lie between two dates.
I am trying to insert data from 9i to 11g using db link with the below query but while doing so the select statement is going for a full table scan even though rowid is used. But when we execute the same select statement without this insert command it is using proper index. Similar issues I am facing for updates.
Query:
INSERT INTO /*+ APPEND */ emp@db_link select * from emp where rowid ='AAC2SmAIIAAAHQgAAZ'
I have created a function for knowing the status of link which connect me with remote database.function is as follows
--------------------------------------------------------------------------------------------- CREATE OR REPLACE FUNCTION get_status_db_link (p_db_link_in VARCHAR2) RETURN NUMBER AS rows NUMBER; v_code NUMBER; v_errm VARCHAR2(64); [code]....
Now functin should returns 1 when link is up and it should return 0 when link down but sometime when link is down it continues showing executing it hang my pc.
The Oracle DB in question is 11.2.0.1, x64, Server 2008.I also have a SQL Server 2005 database that runs a third party product, "PaperVision", which we use to manage documents of various kinds. This SQL Server server is also Win 2008, x64.Now, on the server that runs SQL Server, I have a simple view which is defined as such :
This view works great from the SQL Server side. I also created a database link from Oracle to the SQL Server machine, and it also works great.It is defined as such :
CREATE PUBLIC DATABASE LINK PVE_SQLSERVER CONNECT TO EVP_PVE_USER IDENTIFIED BY <PWD> USING 'PVE_SQLSERVER';
Where EVP_PVE_USER is a user created on the SQL Server machine with rights to select from this view.I know it works because I get results with a sql command like :
select * from VW_PVE_DOCS_1_1@PVE_SQLSERVER;
I also created a view on the Oracle server that refines this information. It is defined as such :
CREATE OR REPLACE FORCE VIEW EVPDBA.VW_PVE_CONTRACTS_INALERT ( DOCID, EFFECTIVE_DATE, EXPIRATION_TYPE, ALERT_PERIOD_START, ALERT_PERIOD_END, ACRONYM, [code]....
This view also works fine, i.e., I can select * from it from the sql command line.Now, the problem comes in when I need to run a procedure that processes this view every night and/or week.I have stripped everything out of this procedure that is not relevant, and it is defined as such for this forum :
CREATE OR REPLACE PROCEDURE EVPDBA.TESTME is tnum number := 0; begin select count(*) into tnum from VW_PVE_CONTRACTS_INALERT; end; /
If I execute this procedure from the sql command line, all is well.When I run it from a scheduler job, I get
ORA-01010: invalid OCI operation ORA-02063: preceding line from PVE_SQLSERVER ORA-06512: at "EVPDBA.TESTME", line 5 ORA-06512: at line 1 ORA-06512: at "SYS.DBMS_ISCHED", line 185 ORA-06512: at "SYS.DBMS_SCHEDULER", line 486 ORA-06512: at line 1
I am aware that DBMS_SCHEDULER performs a commit when scheduling a job, however, this is not scheduled from a trigger.I scoured the forums and have found a few things that seemed relevant, but not much. One had to do with the version of the JDBC driver between two Oracle databases, but I wonder if the age difference between Oracle 11 and SQL Server 2005 (Express) might be an issue. The fact that all command line select statements and running the procedure work fine implies to me that there is an additional issue raised due to the scheduler.
The other posts I found talked about performing a commit just before any select that ultimately pulls across a db link. I did this, and still no luck.One other useful fact - the job appeared to run succesfully at 5am, yet trying again at 8am threw the error, so it may be sporadic. (Although during regular daytime hours it is a very repeatable error).
I am looking into reformatting things to use the older DBMS_JOB, however, I really like the log history of job details and other functionality available with SCHEDULER.
There is one DBMSJOB i created schedule to run every day at 22:00 hrs. The jobname is IT_TO_DUM_LOADING and the job type is Stored procedure.
I queried all_source and found stored procedure does not exist in database.
SQL> select text from all_source 2 where type = 'PROCEDURE' 3 and name ='IT_TO_DUM_LOADING';
no rows selected
But i could see the job details in dba_scheduler_jobs view. I cannot use WHERE clause in dba_scheduler_jobs to restrict and find only the job row as i receive time zone error. How it is possible to view the job in dba_scheduler job view whereas i could not find the stored procedure of its in all_source?
All these stored procedures deals with insert/updated transactions . i need to create a new stored procedure to execute all this in a single stored procedure which will be something like
create procedure sp4(param1...param8) as begin Execute sp1 param1...param6 rollback if any error Execute sp2 param1...param8 rollback if any error Execute sp3 param1...param4 rollback if any error end;
I'm trying to find a simple way of getting around this. I have a PL/SQL procedure which loops through a list of values in a table. These values are the actual names of the functions I want it to run in sequence. The problem is I can get the string value each time but I need to store the return value of each function into a number variable e.g.
BEGIN open all_tests; fetch all_tests BULK COLLECT INTO test_tabl; close all_tests;
for myindex IN 1..test_tabl.count LOOP
time_taken := test_source1 ==> this will work but how do I avoid hardcoding the name of the function to be executed and use the test_name value instead?? time_taken is declared as a number (ie the return value of each function is a number)
test_detail := test_tabl(myindex).test_name; ==> test_detail now contains the string name of the function dbms_output.put_line('Test detail is ' || test_detail); end loop;
This error is returned when executing a Function with the RETURN X%ROWTYPE is used.Here is the code.
CREATE OR REPLACE function FE_GET_addr ( p_PIDM IN addr.addr_PIDM%TYPE, p_atyp1 IN VARCHAR2, p_atyp2 IN VARCHAR2, p_atyp3 IN VARCHAR2, p_atyp4 IN VARCHAR2 [code]....
I have created a Package Body and Package Spec for a function to select a username and a password from a table and return the username.
The code i have created is this:-
CREATE OR REPLACE PACKAGE BODY USER_LOGIN AS FUNCTION user_select (USERNAME_IN VARCHAR2, PASSWORD_IN VARCHAR2) RETURN VARCHAR2 IS USERNAME_OUT VARCHAR2(12); BEGIN [code]........
The package body and spec compiles successfully but i am having trouble when i execute this function. I am trying to do this :-
ERROR at line 1: ORA-06550: line 1, column 7: PLS-00306: wrong number or types of arguments in call to 'USER_SELECT' ORA-06550: line 1, column 7: PL/SQL: Statement ignored
I am getting errors while executing the following block.
create TYPE c_Rec as object(a VARCHAR2(1), b NUMBER); DECLARE -- TYPE c_Rec as object(a VARCHAR2(1), b NUMBER); TYPE c_collection IS TABLE OF c_Rec; l_coll c_collection := c_collection(); BEGIN [code]........
error
06530. 00000 - "Reference to uninitialized composite"
I'm trying to execute a procedure within a Declare/Begin/End statement and using variables as input parameters in my procedure but I keep on getting an Invalid SQL Statement Error. My code is below:
declare
START_dt VARCHAR2(30); END_DT VARCHAR2(30);
begin
SELECT '01-APR-2011' INTO END_DT FROM DUAL; [code]....
The table the procedure is pulling data from doesn't have proper date/time stamps but my procedure takes the varchar dates above and turns them into dates in the procedure so the input date parameters are left as just string characters.
How to execute Sql Server procedure from Oracle.I have created a dblink and when calling directly to that procedure it is throwing error.
--creationg sqlserver procedure create proc getRecords as print 'testing'; --creating db link create database link remote connect by ss identified by ss using 'remote'; [code]....
I have found DBMS_HS_PASSTHROUGH package is capable of doing it but not sure how to do. Nathan
