BULK COLLECT QUERY REWRITE
I understand that Oracle 10g will rewrite your CURSOR FOR LOOP statements into a BULK COLLECT statement.
I am contemplating no longer explicitly writing the BULK COLLECT from now on as it reduces the number of lines of code and greatly simplifies code.
Can anyone see any serious flaws in this strategy?
Kind Regards
Chris
> I also think it is a good idea if people do take the
time to decide their strategy. You seem to be
suggesting that it is a bad idea to stop and think
about what you require from your loop.
Well, that depends on the type of programmer. When one deals with programmers that are not true PL/SQL developers and view PL/SQL.. well, as some kind of inferior database language (compared to something like Java for example).. you want to have templates and stuff to enforce best practises.
> I also don't agree with the 'package tuning knob'.
Each query may have different requirements and, as
with most things in programming, fixing one thing can
have a negative effect on another. It is about the
only place where would not advocate constants.
You have a point - but even so, defining these as constants (even if it has to be inside the actual proc doing the bulk fetch, one per bulk fetch) make it a lot easier to maintain than having to search out the actual bulk fetch statements in the code.
> But i would suggest that analysis is performed on a
close-to-live enviroment with a production level
server, large body of test data and multiple users. I
think we agree on that point.
Yeah.. but the problem there is that I have never really seen such an environment. Usually due to costs. How do you for example duplicate a large RAC, terabytes of SAN space, 1000's of users, for use as a close-to-live enviroment?
The usual approach (by management) is to spend as little as possible on development and Q&A platforms. Which at times means that the performance of dev vs. production can vary a lot.
So we have to play the hand we're dealt with unfortunately.
> Hmmm cmegar and I have never once said "don't worry
about, PL/SQL does it for you".
Yes - of course not. I'm just rambling on in general describing the usual attitudes I see when it comes to features like this.
It is managable in small dev team, but larger ones.. not really. There this attitude is often previlant in my experience. The "silver bullet" bullet syndrome.
> Would you use the same argument with regard to unit
testing. There are not many pl/sql developers who
unit test but should that prevent me using the
technique?
Well.. to be honest, I do not think that a developer that writes at least some basic unit tests for his/her code can be called a developer.
> >[i]Relying on implicit features to "fix" code for
you negates a deeper and better understanding of the
language and what you writeI don't see how using an implicit bulk collect is
fixing code.
Which is why I put it in brackets - "fixing" ito making it more performant, or "fixing" it as a FOR loop contains DMLs that can be changed to FORALLs.
> Don't tell me you've never
updated older code to take advangate of a new
feature.
I can never stop the urge to refactor old code I'm working with. :-)
> >[i]I think of features as an implicit bulk collect
behind the scenes, as crutches for mediocre
programmers.I take it that statement is suggesting that cmedgar
and I are mediocre programmers? Not a nice way to end
an otherwise constructive argument.
How does that saying go? You claim the cloth that I cut? :-)
My sincere apologies to both you guys - I did not intend that statement to personal at all.
Besides, I'm usually more blunt than that what it comes to throwing personal insults around. ;-)
This statement was just a general observation going back to my early days of writing Cobol and Natural. Programmers at time do not seem to care about grokking the features and apply them correctly. Actually I want to say "most programmers" and "a lot of times", but then I would be accused of generalisation. ;-)
I simply find it very frustrating dealing with programmers that does not simply love to write code. Programmers that see it as a mere job.
Someone once said that he never starts out to write beautiful code. But when he is done and the code is not beautiful and elegant (and simple), he knows he has screwed up.
In my experience.. many programmers will not understand this.
Similar Messages
-
Please help with an embedded query (INSERT RETURNING BULK COLLECT INTO)
I am trying to write a query inside the C# code where I would insert values into a table in bulk using bind variables. But I also I would like to receive a bulk collection of generated sequence number IDs for the REQUEST_ID. I am trying to use RETURNING REQUEST_ID BULK COLLECT INTO :REQUEST_IDs clause where :REQUEST_IDs is another bind variable
Here is a full query that use in the C# code
string sql = "INSERT INTO REQUESTS_TBL(REQUEST_ID, CID, PROVIDER_ID, PROVIDER_NAME, REQUEST_TYPE_ID, REQUEST_METHOD_ID, " +
"SERVICE_START_DT, SERVICE_END_DT, SERVICE_LOCATION_CITY, SERVICE_LOCATION_STATE, " +
"BENEFICIARY_FIRST_NAME, BENEFICIARY_LAST_NAME, BENEFICIARY_DOB, HICNUM, CCN, " +
"CLAIM_RECEIPT_DT, ADMISSION_DT, BILL_TYPE, LANGUAGE_ID, CONTRACTOR_ID, PRIORITY_ID, " +
"UNIVERSE_DT, REQUEST_DT, BENEFICIARY_M_INITIAL, ATTENDING_PROVIDER_NUMBER, " +
"BILLING_NPI, BENE_ZIP_CODE, DRG, FINAL_ALLOWED_AMT, STUDY_ID, REFERRING_NPI) " +
"VALUES " +
"(SQ_CDCDATA.NEXTVAL, :CIDs, :PROVIDER_IDs, :PROVIDER_NAMEs, :REQUEST_TYPE_IDs, :REQUEST_METHOD_IDs, " +
":SERVICE_START_DTs, :SERVICE_END_DTs, :SERVICE_LOCATION_CITYs, :SERVICE_LOCATION_STATEs, " +
":BENEFICIARY_FIRST_NAMEs, :BENEFICIARY_LAST_NAMEs, :BENEFICIARY_DOBs, :HICNUMs, :CCNs, " +
":CLAIM_RECEIPT_DTs, :ADMISSION_DTs, :BILL_TYPEs, :LANGUAGE_IDs, :CONTRACTOR_IDs, :PRIORITY_IDs, " +
":UNIVERSE_DTs, :REQUEST_DTs, :BENEFICIARY_M_INITIALs, :ATTENDING_PROVIDER_NUMBERs, " +
":BILLING_NPIs, :BENE_ZIP_CODEs, :DRGs, :FINAL_ALLOWED_AMTs, :STUDY_IDs, :REFERRING_NPIs) " +
" RETURNING REQUEST_ID BULK COLLECT INTO :REQUEST_IDs";
int[] REQUEST_IDs = new int[range];
cmd.Parameters.Add(":REQUEST_IDs", OracleDbType.Int32, REQUEST_IDs, System.Data.ParameterDirection.Output);
However, when I run this query, it gives me a strange error ORA-00925: missing INTO keyword. I am not sure what that error means since I am not missing any INTOs
Please help me resolve this error or I would appreciate a different solution
Thank youIt seems you are not doing a bulk insert but rather an array bind.
(Which you will also find that it is problematic to do an INSERT with a bulk collect returning clause (while this works just fine for update/deletes) :
http://www.oracle-developer.net/display.php?id=413)
But you are using array bind, so you simply just need to use a
... Returning REQUEST_ID INTO :REQUEST_IDand that'll return you a Rquest_ID[]
see below for a working example (I used a procedure but the result is the same)
//Create Table Zzztab(Deptno Number, Deptname Varchar2(50) , Loc Varchar2(50) , State Varchar2(2) , Idno Number(10)) ;
//create sequence zzzseq ;
//CREATE OR REPLACE PROCEDURE ZZZ( P_DEPTNO IN ZZZTAB.DEPTNO%TYPE,
// P_DEPTNAME IN ZZZTAB.DEPTNAME%TYPE,
// P_LOC IN ZZZTAB.LOC%TYPE,
// P_State In Zzztab.State%Type ,
// p_idno out zzztab.idno%type
// IS
//Begin
// Insert Into Zzztab (Deptno, Deptname, Loc, State , Idno)
// Values (P_Deptno, P_Deptname, P_Loc, P_State, Zzzseq.Nextval)
// returning idno into p_idno;
//END ZZZ;
//Drop Procedure Zzz ;
//Drop Sequence Zzzseq ;
//drop Table Zzztab;
class ArrayBind
static void Main(string[] args)
// Connect
string connectStr = GetConnectionString();
// Setup the Tables for sample
Setup(connectStr);
// Initialize array of data
int[] myArrayDeptNo = new int[3]{1, 2, 3};
String[] myArrayDeptName = {"Dev", "QA", "Facility"};
String[] myArrayDeptLoc = {"New York", "Chicago", "Texas"};
String[] state = {"NY","IL","TX"} ;
OracleConnection connection = new OracleConnection(connectStr);
OracleCommand command = new OracleCommand (
"zzz", connection);
command.CommandType = CommandType.StoredProcedure;
// Set the Array Size to 3. This applied to all the parameter in
// associated with this command
command.ArrayBindCount = 3;
command.BindByName = true;
// deptno parameter
OracleParameter deptNoParam = new OracleParameter("p_deptno",OracleDbType.Int32);
deptNoParam.Direction = ParameterDirection.Input;
deptNoParam.Value = myArrayDeptNo;
command.Parameters.Add(deptNoParam);
// deptname parameter
OracleParameter deptNameParam = new OracleParameter("p_deptname", OracleDbType.Varchar2);
deptNameParam.Direction = ParameterDirection.Input;
deptNameParam.Value = myArrayDeptName;
command.Parameters.Add(deptNameParam);
// loc parameter
OracleParameter deptLocParam = new OracleParameter("p_loc", OracleDbType.Varchar2);
deptLocParam.Direction = ParameterDirection.Input;
deptLocParam.Value = myArrayDeptLoc;
command.Parameters.Add(deptLocParam);
//P_STATE -- -ARRAY
OracleParameter stateParam = new OracleParameter("P_STATE", OracleDbType.Varchar2);
stateParam.Direction = ParameterDirection.Input;
stateParam.Value = state;
command.Parameters.Add(stateParam);
//idParam-- ARRAY
OracleParameter idParam = new OracleParameter("p_idno", OracleDbType.Int64 );
idParam.Direction = ParameterDirection.Output ;
idParam.OracleDbTypeEx = OracleDbType.Int64;
command.Parameters.Add(idParam);
try
connection.Open();
command.ExecuteNonQuery ();
Console.WriteLine("{0} Rows Inserted", command.ArrayBindCount);
//now cycle through the output param array
foreach (Int64 i in (Int64[])idParam.Value)
Console.WriteLine(i);
catch (Exception e)
Console.WriteLine("Execution Failed:" + e.Message);
finally
// connection, command used server side resource, dispose them
// asap to conserve resource
connection.Close();
command.Dispose();
connection.Dispose();
Console.WriteLine("Press Enter to finish");
Console.ReadKey();
} -
Please help with the query (INSERT RETURNING BULK COLLECT INTO)
I am trying to write a query inside the C# code where I would insert values into a table in bulk using bind variables. But I also I would like to receive a bulk collection of generated sequence number IDs for the REQUEST_ID. I am trying to use RETURNING REQUEST_ID BULK COLLECT INTO :REQUEST_IDs clause where :REQUEST_IDs is another bind variable
Here is a full query that use in the C# code
INSERT INTO REQUESTS_TBL(REQUEST_ID, CID, PROVIDER_ID, PROVIDER_NAME, REQUEST_TYPE_ID, REQUEST_METHOD_ID, SERVICE_START_DT, SERVICE_END_DT, SERVICE_LOCATION_CITY, SERVICE_LOCATION_STATE, BENEFICIARY_FIRST_NAME,
BENEFICIARY_LAST_NAME, BENEFICIARY_DOB, HICNUM, CCN, CLAIM_RECEIPT_DT, ADMISSION_DT, BILL_TYPE,
LANGUAGE_ID, CONTRACTOR_ID, PRIORITY_ID, UNIVERSE_DT, REQUEST_DT, BENEFICIARY_M_INITIAL,
ATTENDING_PROVIDER_NUMBER, BILLING_NPI, BENE_ZIP_CODE, DRG, FINAL_ALLOWED_AMT, STUDY_ID, REFERRING_NPI)
VALUES
(SQ_CDCDATA.NEXTVAL, :CIDs, :PROVIDER_IDs, :PROVIDER_NAMEs, :REQUEST_TYPE_IDs,
:REQUEST_METHOD_IDs, :SERVICE_START_DTs, :SERVICE_END_DTs, :SERVICE_LOCATION_CITYs,
:SERVICE_LOCATION_STATEs, :BENEFICIARY_FIRST_NAMEs, :BENEFICIARY_LAST_NAMEs, :BENEFICIARY_DOBs,
:HICNUMs, :CCNs, :CLAIM_RECEIPT_DTs, :ADMISSION_DTs, :BILL_TYPEs, :LANGUAGE_IDs,
:CONTRACTOR_IDs, :PRIORITY_IDs, :UNIVERSE_DTs, :REQUEST_DTs, :BENEFICIARY_M_INITIALs,
:ATTENDING_PROVIDER_NUMBERs, :BILLING_NPIs, :BENE_ZIP_CODEs, :DRGs, :FINAL_ALLOWED_AMTs,
:STUDY_IDs, :REFERRING_NPIs) RETURNING REQUEST_ID BULK COLLECT INTO :REQUEST_IDs
However, when I run this query, it gives me a strange error ORA-00925: missing INTO keyword. I am not sure what that error means since I am not missing any INTOs
Please help me resolve this error or I would appreciate a different solution
Thank youYou cannot use (and do not want to in this case) the BULK COLLECT.
create table for_testing
the_id number not null primary key,
some_data number
declare
l_return_value for_testing.the_id%type;
begin
4
insert into for_testing
the_id,
some_data
values
1,
5
returning the_id into l_return_value;
dbms_output.put_line('the return values is ' || l_return_value);
end;
20 /
the return values is 1
PL/SQL procedure successfully completed.
Elapsed: 00:00:00.02
TUBBY_TUBBZ?Is a simple example. In the future, please use the tags to preserve formatting on your code like i have so it remains readable . -
BULK COLLECT in select query inside a function
Hi All,
My query is :
SELECT col1,col2,col3 FROM table_a; --( consider this is a long running query with lot of joins)
Need to know how can i get the output of the above query from a function using BULK COLLECT.
and i tried this:
CREATE OR REPLACE TYPE tab_a_row
AS OBJECT (
col1 number(20),
col2 number(20),
col2 number(20)) ;
create or replace type tab_a_nt as table of tab_a_row;
create or replace function get_table_a
return sys_refcursor
is
tab_a_recs tab_a_nt;
rv sys_refcursor;
begin
SELECT tab_a_row(col1,col2,col3) BULK COLLECT INTO tab_a_recs FROM table_a;
open rv for select * from table(tab_a_recs);
return rv;
end;
Function created successfully. and i exec this from sql plus using
SQL> var rc refcursor;
SQL> exec :rc := get_table_a;
BEGIN :rc := get_table_a; END;
ERROR at line 1:
ORA-22905: cannot access rows from a non-nested table item
ORA-06512: at "GET_TABLE_A", line 12
ORA-06512: at line 1
Kindly share your ideas on how to use bulk collect and get set of outputs from a function.
Edited by: 887268 on Apr 18, 2013 3:10 AM>
If i use refcursor , then the JAVA code needs to be changed accordinglyto get the refcursor output.
>
Well, of course. Java has to know what the sql projection is. How else will it know how many columns there are and their datatypes.
But that is true no matter what method you use.
>
But if i use a PLSQL COLLECTION TYPE (nested tables ) , then i can get output as ,
select * from table(function_name(input1,input2));
>
No - using the 'table' function mean you are calling a PIPELINED function.
This is a sample of a PIPELINED procedure.
-- type to match emp record
create or replace type emp_scalar_type as object
(EMPNO NUMBER(4) ,
ENAME VARCHAR2(10),
JOB VARCHAR2(9),
MGR NUMBER(4),
HIREDATE DATE,
SAL NUMBER(7, 2),
COMM NUMBER(7, 2),
DEPTNO NUMBER(2)
-- table of emp records
create or replace type emp_table_type as table of emp_scalar_type
-- pipelined function
create or replace function get_emp( p_deptno in number )
return emp_table_type
PIPELINED
as
TYPE EmpCurTyp IS REF CURSOR RETURN emp%ROWTYPE;
emp_cv EmpCurTyp;
l_rec emp%rowtype;
begin
open emp_cv for select * from emp where deptno = p_deptno;
loop
fetch emp_cv into l_rec;
exit when (emp_cv%notfound);
pipe row( emp_scalar_type( l_rec.empno, LOWER(l_rec.ename),
l_rec.job, l_rec.mgr, l_rec.hiredate, l_rec.sal, l_rec.comm, l_rec.deptno ) );
end loop;
return;
end;
select * from table(get_emp(20))Or your function could return a collection like the example from this thread this morning.
Example of Collection as datatype of a function’s return value
CREATE OR REPLACE TYPE enamelist as VARRAY(20) of VARCHAR2(20)
/* Formatted on 4/18/2013 4:06:47 PM (QP5 v5.126.903.23003) */
CREATE OR REPLACE FUNCTION ename_fn
RETURN enamelist
AS
v_cursor_main enamelist := enamelist ();
BEGIN
SELECT ename
BULK COLLECT
INTO v_cursor_main
FROM emp;
RETURN v_cursor_main;
EXCEPTION
WHEN OTHERS
THEN
RETURN v_cursor_main;
END;
select * from table(ename_fn()) from dual;
COLUMN_VALUE
SMITH
ALLEN
WARD
JONES
MARTIN
BLAKE
CLARK
SCOTT
KING
TURNER
ADAMS
JAMES
FORD
MILLER -
Hi
I have the following PL/SQL code.
type temp_atlanta is table of atlanta_operator%rowtype
index by binary_integer;
atlanta_1 temp_atlanta;
begin
select * bulk collect into atlanta_1 from atlanta_operator;
forall k in 1..atlanta_1.count
insert into pt_atlanta_operator values atlanta_1(k);
end;
Query :- I want to commit after every 1000 records are inserted. Please tell me how to put rowcount in bulk inserts as %rowcount is not working here .You would probably want to use the LIMIT clause in your BULK COLLECT, i.e.
type temp_atlanta is table of atlanta_operator%rowtype
index by binary_integer;
atlanta_1 temp_atlanta;
CURSOR c1 IS
SELECT *
FROM atlanta_operator;
BEGIN
OPEN c1;
LOOP
FETCH c1
BULK COLLECT INTO atlanta_1
LIMIT 1000;
IF( temp_atlanta.count > 0 )
THEN
FORALL k IN 1..temp_atlanta.count
INSERT INTO pt_atlanta_operator VALUES atlanta_1(k);
END IF;
EXIT WHEN c1%NOTFOUND;
END LOOP;
END;Justin
Distributed Database Consulting, Inc.
http://www.ddbcinc.com/askDDBC -
How to tune a query which contains "Bulk Collect Into" clause
I want to tune the below query:
SELECT customer_master_num,
product_nam
BULK COLLECT INTO t_cont09_rec
FROM TB_CMA009_SUPRA_RE_AGNT_CONT
WHERE re_agent_customer_master_num = p_63cust_master_num
AND customer_master_num = p_63board_master_num
AND cancellation_dt IS NULL
AND NVL (is_training_key_flg, 'N') = 'N';
This contains "Bulk Collect Into" clause.
TYPE cont09cur IS RECORD (
customer_master_num TB_CMA009_SUPRA_RE_AGNT_CONT.customer_master_num%TYPE,
product_nam TB_CMA009_SUPRA_RE_AGNT_CONT.product_nam%TYPE);
t_cont09_rec cont09_rec;
"t_cont09_rec" This is of Record Type
Please help me out how to tune this one.[url http://forums.oracle.com/forums/thread.jspa?threadID=501834&tstart=0]When your query takes too long ...
Also, don't get too distracted by the PL/SQL bulk collect into construction. If it takes time, then you have more than 99% chance that the time is spent in the SQL.
Regards,
Rob. -
Oracle sql query - bulk collect into a 50 element array, to print them.
INSERT INTO temp_mobile_donors_au
select distinct d.donor_id
from [email protected] d,
blood_drives b,
ds_malaria_travel_history h
where b.site_code = 'NA19'
and d.coll_date between trunc(add_months(sysdate,-24))
and trunc(sysdate)
and d.drive_id = 'DRV'||b.drive_id
and h.donor_id = d.donor_id;
I have 105 donors. I am taking these donors and have to print a report. I was using loop to print the donor sheet for each donor. Usually we have about 20 to 30 donors. So reports prints without any problem. Now we have about 105 donors and may be next time we may have more. So report does not print for all donors.
I print reports using this command
Cursor C2 is
select distinct donor_id
from temp_mobile_donors_au;
Begin
Open C2;
Loop
Fetch C2 into :param_donor_id;
Exit when C2%notfound;
print_mobile_donor_sheets_au;
End loop;
Close C2;
End;
procedure for print_mobile_donor_sheet_au----- is as below
v_url_c := 'http://'||v_rep_server_c||'report=r_ds_travel_history1¶m_donor_id='||:control.param_donor_id;
Is there a way to iterate through a single cursor 50 rows at a time, bulk collect into a 50 element array, print them, bulk collect the next batch.
Any idea or help?Here is a bulk-collect-limit example that got by googling:
SET SERVEROUTPUT ON
DECLARE
TYPE t_bulk_collect_test_tab IS TABLE OF bulk_collect_test%ROWTYPE;
l_tab t_bulk_collect_test_tab;
CURSOR c_data IS
SELECT *
FROM bulk_collect_test;
BEGIN
OPEN c_data;
LOOP
FETCH c_data
BULK COLLECT INTO l_tab LIMIT 10000; /*I think this is what you are looking for*/
-- Process contents of collection here.
DBMS_OUTPUT.put_line(l_tab.count || ‘ rows’);
EXIT WHEN c_data%NOTFOUND;
END LOOP
CLOSE c_data;
END;
/So you need to three things:
1. TYPE t_bulk_collect_test_tab IS TABLE OF bulk_collect_test%ROWTYPE;
2. l_tab t_bulk_collect_test_tab;
3. FETCH c_data
BULK COLLECT INTO l_tab LIMIT 10000; /* use your reqd value */Change your code and try.
HTH -
Can I use Bulk Collect results as input parameter for another cursor
MUSIC ==> remote MUSIC_DB database, MUSIC table has 60 million rows
PRICE_DATA ==> remote PRICING_DB database, PRICE_DATE table has 1 billion rows
These two table once existed in same database, but size of database exceeded available hardware size and hardware budget, so the PRICE_DATA table was moved to another Oracle database. I need to create a single report that combines data from both of these tables, and a distributed join with DRIVING_SITE hint will not work because the size of both table is too large to push to one DRIVING_SITE location, so I wrote this PLSQL block to process in small blocks.
QUESTION: how can use bulk collect from one cursor and pass that bulk collected information as input to second cursor without specifically listing each cell of the PLSQL bulk collection? See sample pseudo-code below, I am trying to determine more efficient way to code than hard-coding 100 parameter names into 2nd cursor.
NOTE: below is truly pseudo-code, I had to change the names of everything to adhere to NDA, but below works and is fast enough for my purposes, but if I want to change from 100 input parameters to 200, I have to add more hard-coded values. There has got to be a better way.
DECLARE
-- define cursor that retrieves distinct SONG_IDs from MUSIC table in remote music database
CURSOR C_CURRENT_MUSIC
IS
select distinct SONG_ID
from MUSIC@MUSIC_DB
where PRODUCTION_RELEASE=1
/* define a parameterized cursor that accepts 100 SONG_IDs and retrieves
required pricing information
CURSOR C_get_music_price_data
P_SONG_ID_001 NUMBER, P_SONG_ID_002 NUMBER, P_SONG_ID_003 NUMBER, P_SONG_ID_004 NUMBER, P_SONG_ID_005 NUMBER, P_SONG_ID_006 NUMBER, P_SONG_ID_007 NUMBER, P_SONG_ID_008 NUMBER, P_SONG_ID_009 NUMBER, P_SONG_ID_010 NUMBER,
P_SONG_ID_011 NUMBER, P_SONG_ID_012 NUMBER, P_SONG_ID_013 NUMBER, P_SONG_ID_014 NUMBER, P_SONG_ID_015 NUMBER, P_SONG_ID_016 NUMBER, P_SONG_ID_017 NUMBER, P_SONG_ID_018 NUMBER, P_SONG_ID_019 NUMBER, P_SONG_ID_020 NUMBER,
P_SONG_ID_021 NUMBER, P_SONG_ID_022 NUMBER, P_SONG_ID_023 NUMBER, P_SONG_ID_024 NUMBER, P_SONG_ID_025 NUMBER, P_SONG_ID_026 NUMBER, P_SONG_ID_027 NUMBER, P_SONG_ID_028 NUMBER, P_SONG_ID_029 NUMBER, P_SONG_ID_030 NUMBER,
P_SONG_ID_031 NUMBER, P_SONG_ID_032 NUMBER, P_SONG_ID_033 NUMBER, P_SONG_ID_034 NUMBER, P_SONG_ID_035 NUMBER, P_SONG_ID_036 NUMBER, P_SONG_ID_037 NUMBER, P_SONG_ID_038 NUMBER, P_SONG_ID_039 NUMBER, P_SONG_ID_040 NUMBER,
P_SONG_ID_041 NUMBER, P_SONG_ID_042 NUMBER, P_SONG_ID_043 NUMBER, P_SONG_ID_044 NUMBER, P_SONG_ID_045 NUMBER, P_SONG_ID_046 NUMBER, P_SONG_ID_047 NUMBER, P_SONG_ID_048 NUMBER, P_SONG_ID_049 NUMBER, P_SONG_ID_050 NUMBER,
P_SONG_ID_051 NUMBER, P_SONG_ID_052 NUMBER, P_SONG_ID_053 NUMBER, P_SONG_ID_054 NUMBER, P_SONG_ID_055 NUMBER, P_SONG_ID_056 NUMBER, P_SONG_ID_057 NUMBER, P_SONG_ID_058 NUMBER, P_SONG_ID_059 NUMBER, P_SONG_ID_060 NUMBER,
P_SONG_ID_061 NUMBER, P_SONG_ID_062 NUMBER, P_SONG_ID_063 NUMBER, P_SONG_ID_064 NUMBER, P_SONG_ID_065 NUMBER, P_SONG_ID_066 NUMBER, P_SONG_ID_067 NUMBER, P_SONG_ID_068 NUMBER, P_SONG_ID_069 NUMBER, P_SONG_ID_070 NUMBER,
P_SONG_ID_071 NUMBER, P_SONG_ID_072 NUMBER, P_SONG_ID_073 NUMBER, P_SONG_ID_074 NUMBER, P_SONG_ID_075 NUMBER, P_SONG_ID_076 NUMBER, P_SONG_ID_077 NUMBER, P_SONG_ID_078 NUMBER, P_SONG_ID_079 NUMBER, P_SONG_ID_080 NUMBER,
P_SONG_ID_081 NUMBER, P_SONG_ID_082 NUMBER, P_SONG_ID_083 NUMBER, P_SONG_ID_084 NUMBER, P_SONG_ID_085 NUMBER, P_SONG_ID_086 NUMBER, P_SONG_ID_087 NUMBER, P_SONG_ID_088 NUMBER, P_SONG_ID_089 NUMBER, P_SONG_ID_090 NUMBER,
P_SONG_ID_091 NUMBER, P_SONG_ID_092 NUMBER, P_SONG_ID_093 NUMBER, P_SONG_ID_094 NUMBER, P_SONG_ID_095 NUMBER, P_SONG_ID_096 NUMBER, P_SONG_ID_097 NUMBER, P_SONG_ID_098 NUMBER, P_SONG_ID_099 NUMBER, P_SONG_ID_100 NUMBER
IS
select
from PRICE_DATA@PRICING_DB
where COUNTRY = 'USA'
and START_DATE <= sysdate
and END_DATE > sysdate
and vpc.SONG_ID IN
P_SONG_ID_001 ,P_SONG_ID_002 ,P_SONG_ID_003 ,P_SONG_ID_004 ,P_SONG_ID_005 ,P_SONG_ID_006 ,P_SONG_ID_007 ,P_SONG_ID_008 ,P_SONG_ID_009 ,P_SONG_ID_010,
P_SONG_ID_011 ,P_SONG_ID_012 ,P_SONG_ID_013 ,P_SONG_ID_014 ,P_SONG_ID_015 ,P_SONG_ID_016 ,P_SONG_ID_017 ,P_SONG_ID_018 ,P_SONG_ID_019 ,P_SONG_ID_020,
P_SONG_ID_021 ,P_SONG_ID_022 ,P_SONG_ID_023 ,P_SONG_ID_024 ,P_SONG_ID_025 ,P_SONG_ID_026 ,P_SONG_ID_027 ,P_SONG_ID_028 ,P_SONG_ID_029 ,P_SONG_ID_030,
P_SONG_ID_031 ,P_SONG_ID_032 ,P_SONG_ID_033 ,P_SONG_ID_034 ,P_SONG_ID_035 ,P_SONG_ID_036 ,P_SONG_ID_037 ,P_SONG_ID_038 ,P_SONG_ID_039 ,P_SONG_ID_040,
P_SONG_ID_041 ,P_SONG_ID_042 ,P_SONG_ID_043 ,P_SONG_ID_044 ,P_SONG_ID_045 ,P_SONG_ID_046 ,P_SONG_ID_047 ,P_SONG_ID_048 ,P_SONG_ID_049 ,P_SONG_ID_050,
P_SONG_ID_051 ,P_SONG_ID_052 ,P_SONG_ID_053 ,P_SONG_ID_054 ,P_SONG_ID_055 ,P_SONG_ID_056 ,P_SONG_ID_057 ,P_SONG_ID_058 ,P_SONG_ID_059 ,P_SONG_ID_060,
P_SONG_ID_061 ,P_SONG_ID_062 ,P_SONG_ID_063 ,P_SONG_ID_064 ,P_SONG_ID_065 ,P_SONG_ID_066 ,P_SONG_ID_067 ,P_SONG_ID_068 ,P_SONG_ID_069 ,P_SONG_ID_070,
P_SONG_ID_071 ,P_SONG_ID_072 ,P_SONG_ID_073 ,P_SONG_ID_074 ,P_SONG_ID_075 ,P_SONG_ID_076 ,P_SONG_ID_077 ,P_SONG_ID_078 ,P_SONG_ID_079 ,P_SONG_ID_080,
P_SONG_ID_081 ,P_SONG_ID_082 ,P_SONG_ID_083 ,P_SONG_ID_084 ,P_SONG_ID_085 ,P_SONG_ID_086 ,P_SONG_ID_087 ,P_SONG_ID_088 ,P_SONG_ID_089 ,P_SONG_ID_090,
P_SONG_ID_091 ,P_SONG_ID_092 ,P_SONG_ID_093 ,P_SONG_ID_094 ,P_SONG_ID_095 ,P_SONG_ID_096 ,P_SONG_ID_097 ,P_SONG_ID_098 ,P_SONG_ID_099 ,P_SONG_ID_100
group by
vpc.SONG_ID
,vpc.STOREFRONT_ID
TYPE SONG_ID_TYPE IS TABLE OF MUSIC@MUSIC_DB%TYPE INDEX BY BINARY_INTEGER;
V_SONG_ID_ARRAY SONG_ID_TYPE ;
v_commit_counter NUMBER := 0;
BEGIN
/* open cursor you intent to bulk collect from */
OPEN C_CURRENT_MUSIC;
LOOP
/* in batches of 100, bulk collect ADAM_ID mapped TMS_IDENTIFIER into PLSQL table or records */
FETCH C_CURRENT_MUSIC BULK COLLECT INTO V_SONG_ID_ARRAY LIMIT 100;
EXIT WHEN V_SONG_ID_ARRAY.COUNT = 0;
/* to avoid NO DATA FOUND error when pass 100 parameters to OPEN cursor, if the arrary
is not fully populated to 100, pad the array with nulls to fill up to 100 cells. */
IF (V_SONG_ID_ARRAY.COUNT >=1 and V_SONG_ID_ARRAY.COUNT <> 100) THEN
FOR j IN V_SONG_ID_ARRAY.COUNT+1..100 LOOP
V_SONG_ID_ARRAY(j) := null;
END LOOP;
END IF;
/* pass a batch of 100 to cursor that get price information per SONG_ID and STOREFRONT_ID */
FOR j IN C_get_music_price_data
V_SONG_ID_ARRAY(1) ,V_SONG_ID_ARRAY(2) ,V_SONG_ID_ARRAY(3) ,V_SONG_ID_ARRAY(4) ,V_SONG_ID_ARRAY(5) ,V_SONG_ID_ARRAY(6) ,V_SONG_ID_ARRAY(7) ,V_SONG_ID_ARRAY(8) ,V_SONG_ID_ARRAY(9) ,V_SONG_ID_ARRAY(10) ,
V_SONG_ID_ARRAY(11) ,V_SONG_ID_ARRAY(12) ,V_SONG_ID_ARRAY(13) ,V_SONG_ID_ARRAY(14) ,V_SONG_ID_ARRAY(15) ,V_SONG_ID_ARRAY(16) ,V_SONG_ID_ARRAY(17) ,V_SONG_ID_ARRAY(18) ,V_SONG_ID_ARRAY(19) ,V_SONG_ID_ARRAY(20) ,
V_SONG_ID_ARRAY(21) ,V_SONG_ID_ARRAY(22) ,V_SONG_ID_ARRAY(23) ,V_SONG_ID_ARRAY(24) ,V_SONG_ID_ARRAY(25) ,V_SONG_ID_ARRAY(26) ,V_SONG_ID_ARRAY(27) ,V_SONG_ID_ARRAY(28) ,V_SONG_ID_ARRAY(29) ,V_SONG_ID_ARRAY(30) ,
V_SONG_ID_ARRAY(31) ,V_SONG_ID_ARRAY(32) ,V_SONG_ID_ARRAY(33) ,V_SONG_ID_ARRAY(34) ,V_SONG_ID_ARRAY(35) ,V_SONG_ID_ARRAY(36) ,V_SONG_ID_ARRAY(37) ,V_SONG_ID_ARRAY(38) ,V_SONG_ID_ARRAY(39) ,V_SONG_ID_ARRAY(40) ,
V_SONG_ID_ARRAY(41) ,V_SONG_ID_ARRAY(42) ,V_SONG_ID_ARRAY(43) ,V_SONG_ID_ARRAY(44) ,V_SONG_ID_ARRAY(45) ,V_SONG_ID_ARRAY(46) ,V_SONG_ID_ARRAY(47) ,V_SONG_ID_ARRAY(48) ,V_SONG_ID_ARRAY(49) ,V_SONG_ID_ARRAY(50) ,
V_SONG_ID_ARRAY(51) ,V_SONG_ID_ARRAY(52) ,V_SONG_ID_ARRAY(53) ,V_SONG_ID_ARRAY(54) ,V_SONG_ID_ARRAY(55) ,V_SONG_ID_ARRAY(56) ,V_SONG_ID_ARRAY(57) ,V_SONG_ID_ARRAY(58) ,V_SONG_ID_ARRAY(59) ,V_SONG_ID_ARRAY(60) ,
V_SONG_ID_ARRAY(61) ,V_SONG_ID_ARRAY(62) ,V_SONG_ID_ARRAY(63) ,V_SONG_ID_ARRAY(64) ,V_SONG_ID_ARRAY(65) ,V_SONG_ID_ARRAY(66) ,V_SONG_ID_ARRAY(67) ,V_SONG_ID_ARRAY(68) ,V_SONG_ID_ARRAY(69) ,V_SONG_ID_ARRAY(70) ,
V_SONG_ID_ARRAY(71) ,V_SONG_ID_ARRAY(72) ,V_SONG_ID_ARRAY(73) ,V_SONG_ID_ARRAY(74) ,V_SONG_ID_ARRAY(75) ,V_SONG_ID_ARRAY(76) ,V_SONG_ID_ARRAY(77) ,V_SONG_ID_ARRAY(78) ,V_SONG_ID_ARRAY(79) ,V_SONG_ID_ARRAY(80) ,
V_SONG_ID_ARRAY(81) ,V_SONG_ID_ARRAY(82) ,V_SONG_ID_ARRAY(83) ,V_SONG_ID_ARRAY(84) ,V_SONG_ID_ARRAY(85) ,V_SONG_ID_ARRAY(86) ,V_SONG_ID_ARRAY(87) ,V_SONG_ID_ARRAY(88) ,V_SONG_ID_ARRAY(89) ,V_SONG_ID_ARRAY(90) ,
V_SONG_ID_ARRAY(91) ,V_SONG_ID_ARRAY(92) ,V_SONG_ID_ARRAY(93) ,V_SONG_ID_ARRAY(94) ,V_SONG_ID_ARRAY(95) ,V_SONG_ID_ARRAY(96) ,V_SONG_ID_ARRAY(97) ,V_SONG_ID_ARRAY(98) ,V_SONG_ID_ARRAY(99) ,V_SONG_ID_ARRAY(100)
LOOP
/* do stuff with data from Song and Pricing Database coming from the two
separate cursors, then continue processing more rows...
END LOOP;
/* commit after each batch of 100 SONG_IDs is processed */
COMMIT;
EXIT WHEN C_CURRENT_MUSIC%NOTFOUND; -- exit when there are no more rows to fetch from cursor
END LOOP; -- bulk fetching loop
CLOSE C_CURRENT_MUSIC; -- close cursor that was used in bulk collection
/* commit rows */
COMMIT; -- commit any remaining uncommitted data.
END;I've got a problem when using passing VARRAY of numbers as parameter to remote cursor: it takes a super long time to run, sometimes doesn't finish even after an hour as passed.
Continuing with my example in original entry, I replaced the bulk collect into PLSQL table collection with a VARRAY and i bulk collect into the VARRAY, this is fast and I know it works because I can DBMS_OUTPUT.PUT_LINE cells of VARRAY so I know it is getting populated correctly. However, when I pass the VARRAY containing 100 cells populated with SONG_IDs as parameter to cursor, execution time is over an hour and when I am expecting a few seconds.
Below code example strips the problem down to it's raw details, I skip the bulk collect and just manually populate a VARRAY with 100 SONG_ID values, then try to pass to as parameter to a cursor, but the execution time of cursor is unexpectedly long, over 30 minutes, sometime longer, when I am expecting seconds.
IMPORTANT: If I take the same 100 SONG_IDs and place them directly in the cursor query's where IN clause, the SQL runs in under 5 seconds and returns result. Also, if I pass the 100 SONG_IDs as individual cells of a PLSQL table collection, then it also runs fast.
I thought that since the VARRAY is used via select subquery that is it queried locally, but the cursor is remote, and that I had a distribute problem on my hands, so I put in the DRIVING_SITE hint to attempt to force the result of query against VARRAY to go to remote server and rest of query will run there before returning result, but that didn't work either, still got slow response.
Is something wrong with my code, or I am running into a Oracle problem that may require support to resolve?
DECLARE
/* define a parameterized cursor that accepts XXX number of in SONG_IDs and
retrieves required pricing information
CURSOR C_get_music_price_data
p_array_song_ids SYS.ODCInumberList
IS
select /*+DRIVING_SITE(pd) */
count(distinct s.EVE_ID)
from PRICE_DATA@PRICING_DB pd
where pd.COUNTRY = 'USA'
and pd.START_DATE <= sysdate
and pd.END_DATE > sysdate
and pd.SONG_ID IN
select column_value from table(p_array_song_ids)
group by
pd.SONG_ID
,pd.STOREFRONT_ID
V_ARRAY_SONG_IDS SYS.ODCInumberList := SYS.ODCInumberList();
BEGIN
V_ARRAY_SONG_IDS.EXTEND(100);
V_ARRAY_SONG_IDS( 1 ) := 31135 ;
V_ARRAY_SONG_IDS( 2 ) := 31140 ;
V_ARRAY_SONG_IDS( 3 ) := 31142 ;
V_ARRAY_SONG_IDS( 4 ) := 31144 ;
V_ARRAY_SONG_IDS( 5 ) := 31146 ;
V_ARRAY_SONG_IDS( 6 ) := 31148 ;
V_ARRAY_SONG_IDS( 7 ) := 31150 ;
V_ARRAY_SONG_IDS( 8 ) := 31152 ;
V_ARRAY_SONG_IDS( 9 ) := 31154 ;
V_ARRAY_SONG_IDS( 10 ) := 31156 ;
V_ARRAY_SONG_IDS( 11 ) := 31158 ;
V_ARRAY_SONG_IDS( 12 ) := 31160 ;
V_ARRAY_SONG_IDS( 13 ) := 33598 ;
V_ARRAY_SONG_IDS( 14 ) := 33603 ;
V_ARRAY_SONG_IDS( 15 ) := 33605 ;
V_ARRAY_SONG_IDS( 16 ) := 33607 ;
V_ARRAY_SONG_IDS( 17 ) := 33609 ;
V_ARRAY_SONG_IDS( 18 ) := 33611 ;
V_ARRAY_SONG_IDS( 19 ) := 33613 ;
V_ARRAY_SONG_IDS( 20 ) := 33615 ;
V_ARRAY_SONG_IDS( 21 ) := 33617 ;
V_ARRAY_SONG_IDS( 22 ) := 33630 ;
V_ARRAY_SONG_IDS( 23 ) := 33632 ;
V_ARRAY_SONG_IDS( 24 ) := 33636 ;
V_ARRAY_SONG_IDS( 25 ) := 33638 ;
V_ARRAY_SONG_IDS( 26 ) := 33640 ;
V_ARRAY_SONG_IDS( 27 ) := 33642 ;
V_ARRAY_SONG_IDS( 28 ) := 33644 ;
V_ARRAY_SONG_IDS( 29 ) := 33646 ;
V_ARRAY_SONG_IDS( 30 ) := 33648 ;
V_ARRAY_SONG_IDS( 31 ) := 33662 ;
V_ARRAY_SONG_IDS( 32 ) := 33667 ;
V_ARRAY_SONG_IDS( 33 ) := 33669 ;
V_ARRAY_SONG_IDS( 34 ) := 33671 ;
V_ARRAY_SONG_IDS( 35 ) := 33673 ;
V_ARRAY_SONG_IDS( 36 ) := 33675 ;
V_ARRAY_SONG_IDS( 37 ) := 33677 ;
V_ARRAY_SONG_IDS( 38 ) := 33679 ;
V_ARRAY_SONG_IDS( 39 ) := 33681 ;
V_ARRAY_SONG_IDS( 40 ) := 33683 ;
V_ARRAY_SONG_IDS( 41 ) := 33685 ;
V_ARRAY_SONG_IDS( 42 ) := 33700 ;
V_ARRAY_SONG_IDS( 43 ) := 33702 ;
V_ARRAY_SONG_IDS( 44 ) := 33704 ;
V_ARRAY_SONG_IDS( 45 ) := 33706 ;
V_ARRAY_SONG_IDS( 46 ) := 33708 ;
V_ARRAY_SONG_IDS( 47 ) := 33710 ;
V_ARRAY_SONG_IDS( 48 ) := 33712 ;
V_ARRAY_SONG_IDS( 49 ) := 33723 ;
V_ARRAY_SONG_IDS( 50 ) := 33725 ;
V_ARRAY_SONG_IDS( 51 ) := 33727 ;
V_ARRAY_SONG_IDS( 52 ) := 33729 ;
V_ARRAY_SONG_IDS( 53 ) := 33731 ;
V_ARRAY_SONG_IDS( 54 ) := 33733 ;
V_ARRAY_SONG_IDS( 55 ) := 33735 ;
V_ARRAY_SONG_IDS( 56 ) := 33737 ;
V_ARRAY_SONG_IDS( 57 ) := 33749 ;
V_ARRAY_SONG_IDS( 58 ) := 33751 ;
V_ARRAY_SONG_IDS( 59 ) := 33753 ;
V_ARRAY_SONG_IDS( 60 ) := 33755 ;
V_ARRAY_SONG_IDS( 61 ) := 33757 ;
V_ARRAY_SONG_IDS( 62 ) := 33759 ;
V_ARRAY_SONG_IDS( 63 ) := 33761 ;
V_ARRAY_SONG_IDS( 64 ) := 33763 ;
V_ARRAY_SONG_IDS( 65 ) := 33775 ;
V_ARRAY_SONG_IDS( 66 ) := 33777 ;
V_ARRAY_SONG_IDS( 67 ) := 33779 ;
V_ARRAY_SONG_IDS( 68 ) := 33781 ;
V_ARRAY_SONG_IDS( 69 ) := 33783 ;
V_ARRAY_SONG_IDS( 70 ) := 33785 ;
V_ARRAY_SONG_IDS( 71 ) := 33787 ;
V_ARRAY_SONG_IDS( 72 ) := 33789 ;
V_ARRAY_SONG_IDS( 73 ) := 33791 ;
V_ARRAY_SONG_IDS( 74 ) := 33793 ;
V_ARRAY_SONG_IDS( 75 ) := 33807 ;
V_ARRAY_SONG_IDS( 76 ) := 33809 ;
V_ARRAY_SONG_IDS( 77 ) := 33811 ;
V_ARRAY_SONG_IDS( 78 ) := 33813 ;
V_ARRAY_SONG_IDS( 79 ) := 33815 ;
V_ARRAY_SONG_IDS( 80 ) := 33817 ;
V_ARRAY_SONG_IDS( 81 ) := 33819 ;
V_ARRAY_SONG_IDS( 82 ) := 33821 ;
V_ARRAY_SONG_IDS( 83 ) := 33823 ;
V_ARRAY_SONG_IDS( 84 ) := 33825 ;
V_ARRAY_SONG_IDS( 85 ) := 33839 ;
V_ARRAY_SONG_IDS( 86 ) := 33844 ;
V_ARRAY_SONG_IDS( 87 ) := 33846 ;
V_ARRAY_SONG_IDS( 88 ) := 33848 ;
V_ARRAY_SONG_IDS( 89 ) := 33850 ;
V_ARRAY_SONG_IDS( 90 ) := 33852 ;
V_ARRAY_SONG_IDS( 91 ) := 33854 ;
V_ARRAY_SONG_IDS( 92 ) := 33856 ;
V_ARRAY_SONG_IDS( 93 ) := 33858 ;
V_ARRAY_SONG_IDS( 94 ) := 33860 ;
V_ARRAY_SONG_IDS( 95 ) := 33874 ;
V_ARRAY_SONG_IDS( 96 ) := 33879 ;
V_ARRAY_SONG_IDS( 97 ) := 33881 ;
V_ARRAY_SONG_IDS( 98 ) := 33883 ;
V_ARRAY_SONG_IDS( 99 ) := 33885 ;
V_ARRAY_SONG_IDS(100 ) := 33889 ;
/* do stuff with data from Song and Pricing Database coming from the two
separate cursors, then continue processing more rows...
FOR i IN C_get_music_price_data( v_array_song_ids ) LOOP
. (this is the loop where I pass in v_array_song_ids
. populated with only 100 cells and it runs forever)
END LOOP;
END; -
Use of FOR Cursor and BULK COLLECT INTO
Dear all,
in which case we prefer to use FOR cursor and cursor with BULK COLLECT INTO? The following contains two block that query identically where one is using FOR cursor, the other is using BULK COLLECT INTO . Which one that performs better given in the existing task? How do we measure performance between these two?
I'm using sample HR schema:
declare
l_start number;
BEGIN
l_start:= DBMS_UTILITY.get_time;
dbms_lock.sleep(1);
FOR employee IN (SELECT e.last_name, j.job_title FROM employees e,jobs j
where e.job_id=j.job_id and e.job_id LIKE '%CLERK%' AND e.manager_id > 120 ORDER BY e.last_name)
LOOP
DBMS_OUTPUT.PUT_LINE ('Name = ' || employee.last_name || ', Job = ' || employee.job_title);
END LOOP;
DBMS_OUTPUT.put_line('total time: ' || to_char(DBMS_UTILITY.get_time - l_start) || ' hsecs');
END;
declare
l_start number;
type rec_type is table of varchar2(20);
name_rec rec_type;
job_rec rec_type;
begin
l_start:= DBMS_UTILITY.get_time;
dbms_lock.sleep(1);
SELECT e.last_name, j.job_title bulk collect into name_rec,job_rec FROM employees e,jobs j
where e.job_id=j.job_id and e.job_id LIKE '%CLERK%' AND e.manager_id > 120 ORDER BY e.last_name;
for j in name_rec.first..name_rec.last loop
DBMS_OUTPUT.PUT_LINE ('Name = ' || name_rec(j) || ', Job = ' || job_rec(j));
END LOOP;
DBMS_OUTPUT.put_line('total time: ' || to_char(DBMS_UTILITY.get_time - l_start) || ' hsecs');
end;
/In this code, I put timestamp in each block, but they are useless since they both run virtually instantaneous...
Best regards,
ValIf you want to get 100% benifit of bulk collect then it must be implemented as below
declare
Cursor cur_emp
is
SELECT e.last_name, j.job_title
FROM employees e,jobs j
where e.job_id=j.job_id
and e.job_id LIKE '%CLERK%'
AND e.manager_id > 120
ORDER BY e.last_name;
l_start number;
type rec_type is table of varchar2(20);
name_rec rec_type;
job_rec rec_type;
begin
l_start:= DBMS_UTILITY.get_time;
dbms_lock.sleep(1);
/*SELECT e.last_name, j.job_title bulk collect into name_rec,job_rec FROM employees e,jobs j
where e.job_id=j.job_id and e.job_id LIKE '%CLERK%' AND e.manager_id > 120 ORDER BY e.last_name;
OPEN cur_emp;
LOOP
FETCH cur_emp BULK COLLECT INTO name_rec LIMIT 100;
EXIT WHEN name_rec.COUNT=0;
FOR j in 1..name_rec.COUNT
LOOP
DBMS_OUTPUT.PUT_LINE ('Name = ' || name_rec(j) || ', Job = ' || job_rec(j));
END LOOP;
EXIT WHEN cur_emp%NOTFOUND;
END LOOP;
CLOSE cur_emp;
DBMS_OUTPUT.put_line('total time: ' || to_char(DBMS_UTILITY.get_time - l_start) || ' hsecs');
end;
/ -
Issue in using Cursor+Dynamic SQL+ Bulk collect +FORALL
Hi,
I have a dynamic query which I need to use as a cursor to fetch records that inturn need to be inserted into a staging table.
The issue I am facing is I am not sure how to declare the variable to fetch the records into. Since I am using a dynamic cursor how do I declare it?
My code looks something like this -
TYPE c_details_tbl_type IS REF CURSOR;
c_details c_details_tbl_type;
TYPE c_det_tbl_type IS TABLE OF c_details%ROWTYPE INDEX BY PLS_INTEGER;
c_det_tbl c_det_tbl_type; -- ???
BEGIN
v_string1 := 'SELECT....'
v_string2 := ' UNION ALL SELECT....'
v_string3 := 'AND ....'
v_string := v_string1||v_string2||v_string3;
OPEN c_details FOR v_string;
LOOP
FETCH c_details BULK COLLECT
INTO c_det_tbl LIMIT 1000;
IF (c_det_tbl.COUNT > 0) THEN
DELETE FROM STG;
FORALL i IN 1..c_det_tbl.COUNT
INSERT INTO STG
VALUES (c_det_tbl(i));
END IF;
EXIT WHEN c_details%NOTFOUND;
END LOOP;
CLOSE c_details;
END
ThanksWhy the bulk collect? All that this does is slow down the read process (SELECT) and write process (INSERT).
Data selected needs (as a collection) to be pushed into the PGA memory of the PL/SQL engine. And then that very same data needs to be pushed again by the PL/SQL engine back to the database to be inserted. Why?
It is a lot faster, needs a lot less resources, with fewer moving parts, to simply instruct the SQL engine to do both these steps using a single INSERT..SELECT statement. And this can support parallel DML too for scalability when data volumes get large.
It is also pretty easy to make a single SQL statement like this dynamic and even support bind variables.
Simplicity is the ultimate form of elegance. Pushing data needlessly around is not simple and thus not a very elegant way to address the problem. -
Bulk Collect taking more time. Please suggest .
I am working on oracle 11g
I have one normal insert proc
CREATE OR REPLACE PROCEDURE test2
AS
BEGIN
INSERT INTO first_table
(citiversion, financialcollectionid,
dataitemid, dataitemvalue,
unittypeid, financialinstanceid,
VERSION, providerid, user_comment,
userid, insert_timestamp,
latestflag, finalflag, filename,
final_ytdltm_flag, nmflag , partition_key
SELECT citiversion, financialcollectionid,
dataitemid, dataitemvalue, unittypeid,
new_fi, VERSION, providerid,
user_comment, userid,
insert_timestamp, latestflag,
finalflag, filename, '', nmflag,1
FROM secon_table
WHERE financialinstanceid = 1
AND changeflag = 'A'
AND processed_flg = 'N';
END test2;
To impove performance i have normal insert into convert it to bulk collect :
CREATE OR REPLACE PROCEDURE test
AS
BEGIN
DECLARE
CURSOR get_cat_fin_collection_data(n_instanceid NUMBER) IS
SELECT citiversion,
financialcollectionid,
dataitemid,
dataitemvalue,
unittypeid,
new_fi,
VERSION,
providerid,
user_comment,
userid,
insert_timestamp,
latestflag,
finalflag,
filename,
nmflag,
1
FROM secon_table
WHERE financialinstanceid = n_instanceid
AND changeflag = 'A'
AND processed_flg = 'N';
TYPE data_array IS TABLE OF get_cat_fin_collection_data%ROWTYPE;
l_data data_array;
BEGIN
OPEN get_cat_fin_collection_data(1);
LOOP
FETCH get_cat_fin_collection_data BULK COLLECT
INTO l_data limit 100;
FORALL i IN 1 .. l_data.COUNT
INSERT INTO first_table VALUES l_data (i);
EXIT WHEN l_data.count =0;
END LOOP;
CLOSE get_cat_fin_collection_data;
END;
END test;
But bulk collect is taking more time.
below is the timings
SQL> set timing on
SQL> exec test
PL/SQL procedure successfully completed
Executed in 16.703 seconds
SQL> exec test2
PL/SQL procedure successfully completed
Executed in 9.406 seconds
SQL> rollback;
Rollback complete
Executed in 2.75 seconds
SQL> exec test
PL/SQL procedure successfully completed
Executed in 16.266 seconds
SQL> rollback;
Rollback complete
Executed in 2.812 seconds
Normal insert :- 9.4 second
Bulk insert:- 16.266 seconds
I am processing 1 lakh rows.
Can you please tell me the reason why bulk collect is taking more time. ? According to my knowledge it should take less time.
Please suggect do i need to check any parameter?
Please help.
Edited by: 976747 on Feb 4, 2013 1:12 AM>
Can you please tell me the reason why bulk collect is taking more time. ? According to my knowledge it should take less time.
Please suggect do i need to check any parameter?In that case, your knowledge is flawed.
Pure SQL is almost always faster than PL/SQL.
If your Insert into Select is executing slow, then it is probably because the Select statement is taking long to execute. How many rows are being Selected and Inserted from your query?
You might also consider tuning the Select statement. For more information on Posting a Tuning request, read {message:id=3292438} and post the relevant information. -
Procedure failed while using bulk collect into clause and works with cursor
hi all,
I am using "BULK collect into" clause in my procedure and it is failing after 21 minutes and gives the error "end of file communication channel".
after this error comes when i tried to connect database it is giving following error.
ORA -01034 - Oracle not available.
ORA- 27101 - shared memory realm does not exist.
svr4- error :2 : No such file or directory.
when i use cursor instead of BULK COLLECT INTO clause it is running successful.
Following code is working with cursor.
procedure work_kiosk_full(an_jobid in number ,ac_sqlcode out varchar2 ,ac_sqlerrm out varchar2) is
ld_curr_time Date;
cursor cur_work_kiosk is
select distinct jt.jt_id AS jt_id,
NVL ((ROUND ((jt_date_completed - jt_date_requested) * 24, 2)
0
) AS actual_hrs_to_complete,
NVL ((ROUND ((jt_date_responded - jt_date_requested) * 24, 2)
0
) AS actual_hrs_to_respond,
peo1.peo_name AS agent_name,
peo1.peo_user_name AS asagent_soe_id,
le.lglent_desc AS ap_system,
' ' AS assign_work_request_comment,
DECODE (jt.jt_bill_id,
138802, 'CLIENT BILLABLE',
138803, 'CONTRACTED',
138804, 'INTERNAL BILLABLE',
NULL, ' '
) AS billable,
bl.bldg_name_cc AS building, bl.bldg_id_ls AS building_id,
DECODE (bl.bldg_active_cc,
'Y', 'ACTIVE',
'INACTIVE'
) AS building_status,
DECODE (jt.jt_wrk_cause_id,
141521, 'STANDARD WEAR AND TEAR',
141522, 'NEGLIGENCE',
141523, 'ACCIDENTAL',
141524, 'MECHANICAL MALFUNCTION',
141525, 'OVERSIGHT',
141526, 'VANDAL',
141527, 'STANDARD',
141528, 'PROJECT WORK',
6058229, 'TEST',
NULL, ' '
) AS cause_type,
' ' AS comments, peo3.peo_name AS completed_by,
jt.jt_requestor_email AS contact_email,
jt.jt_requestor_name_first
|| ' '
|| jt.jt_requestor_name_last AS contact_name,
jt.jt_requestor_phone AS contact_phone,
cc.cstctrcd_apcode AS corp_code,
cc.cstctrcd_code AS cost_center,
jt.jt_date_closed AS date_closed,
jt.jt_date_completed AS date_completed,
jt.jt_date_requested AS date_requested,
jt.jt_date_responded AS date_responded,
jt.jt_date_response_ecd AS date_response_ecd,
jt.jt_date_scheduled AS date_scheduled,
DECODE (jt.jt_def_id,
139949, 'WTG VENDOR RESPONSE',
139950, 'WAITING ON PARTS',
139951, 'LABOR AVAILABILITY',
139952, 'DEFERRED- HI PRI WORK',
139953, 'WTG APPROVAL',
139954, 'FUNDING REQUIRED',
139955, 'ACCESS DENIED',
139956, 'WTG MATERIAL',
NULL, ' '
) AS deferral_reason,
jt.jt_description AS description,
jt.jt_date_resched_ecd AS ecd,
fmg.facility_manager AS facility_manager,
fl.floors_text AS FLOOR, gl.genled_desc AS general_ledger,
' ' AS kiosk_date_requested, ' ' AS kiosk_dispatch_confirmed,
' ' AS kiosk_dispatched,
eqp.equip_customer_code AS linked_equipment_alias,
eqp.equip_id AS linked_equipment_id,
eqp.equip_text AS linked_equipment_name,
DECODE (jt_originator_type_id,
1000, 'PROJECT MOVE REQUEST',
138834, 'CUSTOMER INITIATED CORRECTION',
138835, 'CUSTOMER INITIATED REQUEST',
138836, 'CORRECTIVE MAINTENANCE',
138837, 'CONFERENCE ROOM BOOKING',
138838, 'PROJECT INITIATED REQUEST',
138839, 'PLANNED PREVENTIVE MAINTENANCE',
138840, 'SELF INITATED REQUEST',
NULL, ' '
) AS originator_type,
' ' AS payment_terms, priority_text AS priority_code,
swoty.sworktype_text AS problem_type,
prop.property_name_cc AS property,
jt.jt_cost_quote_total AS quote_total,
par.levels_name AS region,
DECODE (jt.jt_repdef_id,
141534, 'ADJUSTED SETTING',
141535, 'TRAINING FOR END',
141536, 'NEW REQUEST',
141537, 'NO REPAIR REQUIR',
141538, 'REPLACED PARTS',
141539, 'REPLACE EQUIPMEN',
1000699, 'NEW REQUEST',
NULL, ' '
) AS repair_definitions,
jt.jt_repairdesc AS repair_description,
jt.jt_requestor AS requestor, ' ' AS requestor_cost_center,
jt.jt_requestor_email AS requestor_email,
jt.jt_requestor_name_first AS requestor_name,
jt.jt_requestor_phone AS requestor_phone,
' ' AS response_time, rm.room_name_cc AS room,
p1.peo_provider_code1 AS service_provider,
p1.peo_address_1 AS service_provider_address,
peocity.city_text service_provider_city,
p1.peo_provider_code1 AS service_provider_code,
peocity.city_country_name AS service_provider_country,
peocur.currency_text AS service_provider_currency,
p1.peo_name AS service_provider_description,
p1.peo_dispatch_method AS serv_prov_dispatc_hmethod,
p1.peo_rate_double AS serv_prov_double_time_rate,
p1.peo_email AS service_provider_email,
p1.peo_emergency_phone AS serv_prov_emergency_phone,
p1.peo_fax AS service_provider_fax_number,
p1.peo_home_phone AS service_provider_home_phone,
p1.peo_rate_hourly AS service_provider_hourly_rate,
p1.peo_title AS service_provider_job_title,
p1.peo_method_id AS service_provider_method,
p1.peo_cell_phone AS service_provider_mobile_phone,
p1.peo_pager AS service_provider_pager,
p1.peo_rate_differential AS service_provider_rates,
p1.peo_rate_differential AS ser_prov_shift_differential,
peocity.city_state_prov_text AS serv_prov_state_province,
DECODE (p1.peo_active,
'Y', 'ACTIVE',
'INACTIVE'
) AS service_provider_status,
p1.peo_url AS serv_prov_web_site_address,
p1.peo_phone AS service_provider_work_phone,
p1.peo_postal_code AS serv_prov_zip_postal_code, ' ' AS shift,
' ' AS skill,
DECODE (jt.jt_bigstatus_id,
138813, 'NEW',
138814, 'PENDING',
138815, 'OPEN',
138816, 'COMPLETED',
138817, 'CLOSED',
138818, 'CANCELLED',
NULL, ' '
) AS status,
lev.levels_name AS subregion, ' ' AS trade,
p1.peo_ls_interface_code1 AS vendor_id,
p1.peo_fax AS vendor_purchasing_fax,
p1.peo_vendor_site_code AS vendor_sitecode,
jt.jt_id AS vendor_ticket, p1.peo_name AS vendor_companyname,
jt.jt_requestor_vip AS vip, wo.wo_id AS work_order_no,
jt.jt_id AS work_request,
jt.jt_class_id AS work_request_class,
woty.worktype_text AS work_type, ' ' AS wr_cost,
jt.jt_description AS wr_description,
' ' AS wr_dispatch_method,
DECODE (jt.jt_bigstatus_id,
138813, 'NEW',
138814, 'PENDING',
138815, 'OPEN',
138816, 'COMPLETED',
138817, 'CLOSED',
138818, 'CANCELLED',
NULL, ' '
) AS wr_status,
ctry.country_name AS country
FROM citi.jobticket jt,
citi.property prop,
citi.bldg bl,
citi.bldg_levels bldglvl,
citi.LEVELS lev,
citi.LEVELS par,
(SELECT crstools.stragg (peo_name) facility_manager,
bldgcon_bldg_id
FROM citi.bldg_contacts, citi.people
WHERE bldgcon_peo_id = peo_id
AND bldgcon_contype_id IN (40181, 10142)
GROUP BY bldgcon_bldg_id) fmg,
citi.floors fl,
citi.room rm,
citi.general_ledger gl,
citi.legal_entity le,
citi.cost_center_codes cc,
citi.equipment eqp,
citi.worktype woty,
citi.subworktype swoty,
citi.work_order wo,
citi.jt_workers jtwo,
citi.priority,
citi.country ctry,
citi.people p1,
citi.people peo3,
citi.people peo1,
citi.city peocity,
citi.currency peocur
WHERE jt.jt_bldg_id = bl.bldg_id
AND bl.bldg_id = bldglvl.bldg_levels_bldg_id
AND bldglvl.bldg_levels_levels_id = lev.levels_id
AND lev.levels_parent = par.levels_id(+)
AND prop.property_id = bl.bldg_property_id
AND bl.bldg_active_ls <> 'N'
AND jt.jt_floors_id = fl.floors_id(+)
AND jt.jt_room_id = rm.room_id(+)
AND jt.jt_bldg_id = fmg.bldgcon_bldg_id(+)
AND jt.jt_genled_id = gl.genled_id(+)
AND gl.genled_lglent_id = le.lglent_id(+)
AND jt.jt_cstctrcd_id = cc.cstctrcd_id(+)
AND jt.jt_equip_id = eqp.equip_id(+)
AND jt.jt_id = jtwo.jtw_jt_id(+)
AND jt.jt_worktype_id = woty.worktype_id(+)
AND jt.jt_sworktype_id = swoty.sworktype_id(+)
AND jt.jt_wo_id = wo.wo_id
AND jt.jt_priority_id = priority_id(+)
--AND jt.jt_date_requested >= ADD_MONTHS (SYSDATE, -12)
AND jt.jt_last_update >= ADD_MONTHS (ld_curr_time, -12)
AND bl.bldg_country_id = ctry.country_id
AND jtwo.jtw_peo_id = p1.peo_id(+)
AND p1.peo_city_id = peocity.city_id(+)
AND jt.jt_completed_by_peo_id = peo3.peo_id(+)
AND p1.peo_rate_currency_id = peocur.currency_id(+)
AND jt.jt_agent_peo_id = peo1.peo_id(+);
BEGIN
execute immediate 'truncate table crstools.drt_bom_work_kiosk';
select sysdate into ld_curr_time from dual;
FOR cur_rec in cur_work_kiosk LOOP
IF MOD(cur_work_kiosk%rowcount,10000 ) = 0 then
COMMIT;
END IF;
INSERT INTO crstools.drt_bom_work_kiosk
( JT_ID
,ACTUAL_HRS_TO_COMPLETE
,ACTUAL_HRS_TO_RESPOND
,AGENT_NAME
,ASAGENT_SOE_ID
,AP_SYSTEM
,ASSIGN_WORK_REQUEST_COMMENT
,BILLABLE
,BUILDING
,BUILDING_ID
,BUILDING_STATUS
,CAUSE_TYPE
,COMMENTS
,COMPLETED_BY
,CONTACT_EMAIL
,CONTACT_NAME
,CONTACT_PHONE
,CORP_CODE
,COST_CENTER
,DATE_CLOSED
,DATE_COMPLETED
,DATE_REQUESTED
,DATE_RESPONDED
,DATE_RESPONSE_ECD
,DATE_SCHEDULED
,DEFERRAL_REASON
,DESCRIPTION
,ECD
,FACILITY_MANAGER
,FLOOR
,GENERAL_LEDGER
,KIOSK_DATE_REQUESTED
,KIOSK_DISPATCH_CONFIRMED
,KIOSK_DISPATCHED
,LINKED_EQUIPMENT_ALIAS
,LINKED_EQUIPMENT_ID
,LINKED_EQUIPMENT_NAME
,ORIGINATOR_TYPE
,PAYMENT_TERMS
,PRIORITY_CODE
,PROBLEM_TYPE
,PROPERTY
,QUOTE_TOTAL
,REGION
,REPAIR_DEFINITIONS
,REPAIR_DESCRIPTION
,REQUESTOR
,REQUESTOR_COST_CENTER
,REQUESTOR_EMAIL
,REQUESTOR_NAME
,REQUESTOR_PHONE
,RESPONSE_TIME
,ROOM
,SERVICE_PROVIDER
,SERVICE_PROVIDER_ADDRESS
,SERVICE_PROVIDER_CITY
,SERVICE_PROVIDER_CODE
,SERVICE_PROVIDER_COUNTRY
,SERVICE_PROVIDER_CURRENCY
,SERVICE_PROVIDER_DESCRIPTION
,SERV_PROV_DISPATC_HMETHOD
,SERV_PROV_DOUBLE_TIME_RATE
,SERVICE_PROVIDER_EMAIL
,SERV_PROV_EMERGENCY_PHONE
,SERVICE_PROVIDER_FAX_NUMBER
,SERVICE_PROVIDER_HOME_PHONE
,SERVICE_PROVIDER_HOURLY_RATE
,SERVICE_PROVIDER_JOB_TITLE
,SERVICE_PROVIDER_METHOD
,SERVICE_PROVIDER_MOBILE_PHONE
,SERVICE_PROVIDER_PAGER
,SERVICE_PROVIDER_RATES
,SER_PROV_SHIFT_DIFFERENTIAL
,SERV_PROV_STATE_PROVINCE
,SERVICE_PROVIDER_STATUS
,SERV_PROV_WEB_SITE_ADDRESS
,SERVICE_PROVIDER_WORK_PHONE
,SERV_PROV_ZIP_POSTAL_CODE
,SHIFT
,SKILL
,STATUS
,SUBREGION
,TRADE
,VENDOR_ID
,VENDOR_PURCHASING_FAX
,VENDOR_SITECODE
,VENDOR_TICKET
,VENDOR_COMPANYNAME
,VIP
,WORK_ORDER_NO
,WORK_REQUEST
,WORK_REQUEST_CLASS
,WORK_TYPE
,WR_COST
,WR_DESCRIPTION
,WR_DISPATCH_METHOD
,WR_STATUS
,COUNTRY
,CREATE_DATE
VALUES
(cur_rec.jt_id
,cur_rec.ACTUAL_HRS_TO_COMPLETE
,cur_rec.ACTUAL_HRS_TO_RESPOND
,cur_rec.AGENT_NAME
,cur_rec.ASAGENT_SOE_ID
,cur_rec.AP_SYSTEM
,cur_rec.ASSIGN_WORK_REQUEST_COMMENT
,cur_rec.BILLABLE
,cur_rec.BUILDING
,cur_rec.BUILDING_ID
,cur_rec.BUILDING_STATUS
,cur_rec.CAUSE_TYPE
,cur_rec.COMMENTS
,cur_rec.COMPLETED_BY
,cur_rec.CONTACT_EMAIL
,cur_rec.CONTACT_NAME
,cur_rec.CONTACT_PHONE
,cur_rec.CORP_CODE
,cur_rec.COST_CENTER
,cur_rec.DATE_CLOSED
,cur_rec.DATE_COMPLETED
,cur_rec.DATE_REQUESTED
,cur_rec.DATE_RESPONDED
,cur_rec.DATE_RESPONSE_ECD
,cur_rec.DATE_SCHEDULED
,cur_rec.DEFERRAL_REASON
,cur_rec.DESCRIPTION
,cur_rec.ECD
,cur_rec.FACILITY_MANAGER
,cur_rec.FLOOR
,cur_rec.GENERAL_LEDGER
,cur_rec.KIOSK_DATE_REQUESTED
,cur_rec.KIOSK_DISPATCH_CONFIRMED
,cur_rec.KIOSK_DISPATCHED
,cur_rec.LINKED_EQUIPMENT_ALIAS
,cur_rec.LINKED_EQUIPMENT_ID
,cur_rec.LINKED_EQUIPMENT_NAME
,cur_rec.ORIGINATOR_TYPE
,cur_rec.PAYMENT_TERMS
,cur_rec.PRIORITY_CODE
,cur_rec.PROBLEM_TYPE
,cur_rec.PROPERTY
,cur_rec.QUOTE_TOTAL
,cur_rec.REGION
,cur_rec.REPAIR_DEFINITIONS
,cur_rec.REPAIR_DESCRIPTION
,cur_rec.REQUESTOR
,cur_rec.REQUESTOR_COST_CENTER
,cur_rec.REQUESTOR_EMAIL
,cur_rec.REQUESTOR_NAME
,cur_rec.REQUESTOR_PHONE
,cur_rec.RESPONSE_TIME
,cur_rec.ROOM
,cur_rec.SERVICE_PROVIDER
,cur_rec.SERVICE_PROVIDER_ADDRESS
,cur_rec.SERVICE_PROVIDER_CITY
,cur_rec.SERVICE_PROVIDER_CODE
,cur_rec.SERVICE_PROVIDER_COUNTRY
,cur_rec.SERVICE_PROVIDER_CURRENCY
,cur_rec.SERVICE_PROVIDER_DESCRIPTION
,cur_rec.SERV_PROV_DISPATC_HMETHOD
,cur_rec.SERV_PROV_DOUBLE_TIME_RATE
,cur_rec.SERVICE_PROVIDER_EMAIL
,cur_rec.SERV_PROV_EMERGENCY_PHONE
,cur_rec.SERVICE_PROVIDER_FAX_NUMBER
,cur_rec.SERVICE_PROVIDER_HOME_PHONE
,cur_rec.SERVICE_PROVIDER_HOURLY_RATE
,cur_rec.SERVICE_PROVIDER_JOB_TITLE
,cur_rec.SERVICE_PROVIDER_METHOD
,cur_rec.SERVICE_PROVIDER_MOBILE_PHONE
,cur_rec.SERVICE_PROVIDER_PAGER
,cur_rec.SERVICE_PROVIDER_RATES
,cur_rec.SER_PROV_SHIFT_DIFFERENTIAL
,cur_rec.SERV_PROV_STATE_PROVINCE
,cur_rec.SERVICE_PROVIDER_STATUS
,cur_rec.SERV_PROV_WEB_SITE_ADDRESS
,cur_rec.SERVICE_PROVIDER_WORK_PHONE
,cur_rec.SERV_PROV_ZIP_POSTAL_CODE
,cur_rec.SHIFT
,cur_rec.SKILL
,cur_rec.STATUS
,cur_rec.SUBREGION
,cur_rec.TRADE
,cur_rec.VENDOR_ID
,cur_rec.VENDOR_PURCHASING_FAX
,cur_rec.VENDOR_SITECODE
,cur_rec.VENDOR_TICKET
,cur_rec.VENDOR_COMPANYNAME
,cur_rec.VIP
,cur_rec.WORK_ORDER_NO
,cur_rec.WORK_REQUEST
,cur_rec.WORK_REQUEST_CLASS
,cur_rec.WORK_TYPE
,cur_rec.WR_COST
,cur_rec.WR_DESCRIPTION
,cur_rec.WR_DISPATCH_METHOD
,cur_rec.WR_STATUS
,cur_rec.COUNTRY
,ld_curr_time
END LOOP;
COMMIT;
exception
when others then
rollback;
dbms_output.put_line('SQLCODE :'||sqlcode ||' Error :'||sqlerrm);
end work_kiosk_full;
Note : total record inserted 849000.
The same code does not work with bulk collect into caluse.
Please help me out why this is happening.
Thanks & regards
shyam~Shyam,
I agree with Billy.
Why are you not using an INSERT..SELECT ?
Also, what are you trying to achieve by
- incremental commits?
- copying data from one table to another (using expensive I/O)?
- using dynamic DML?
Most of these approaches are typically wrong - and not recommended for scalable and performant Oracle applications.I could see you using a CURSOR FOR LOOP if you were changing the data being inserted in such a way that you couldn't encapsulate the changes in a query, but you are doing a straight insert into the table from your Cursor. A much more efficient way would be to use the following modifications I made to your code example:
PROCEDURE WORK_KIOSK_FULL(AN_JOBID IN NUMBER,
AC_SQLCODE OUT VARCHAR2,
AC_SQLERRM OUT VARCHAR2) IS
BEGIN
EXECUTE IMMEDIATE 'truncate table crstools.drt_bom_work_kiosk';
/* Note: The APPEND hint forces a Direct Path INSERT (see Link below code sample) and is combined with the NOLOGGING Hint */
/* To dramtically increase performance. The Direct Path INSERT inserts records above the High-Water Mark on the table. */
INSERT /*+ APPEND NOLOGGING */ INTO CRSTOOLS.DRT_BOM_WORK_KIOSK
(JT_ID
,ACTUAL_HRS_TO_COMPLETE
,ACTUAL_HRS_TO_RESPOND
,AGENT_NAME
,ASAGENT_SOE_ID
,AP_SYSTEM
-- ,ASSIGN_WORK_REQUEST_COMMENT /* I commented out this COLUMN because it doesn't make sense to me to insert */
,BILLABLE /* a couple of space characters into a table. If the intent is to leave the column NULL */
,BUILDING /* don't include it in your INSERT statement and it will be NULL. If there is a valid reason */
,BUILDING_ID /* for inserting the spaces, then remove the "line comments" from the insert and select statments */
,BUILDING_STATUS
,CAUSE_TYPE
-- ,COMMENTS
,COMPLETED_BY
,CONTACT_EMAIL
,CONTACT_NAME
,CONTACT_PHONE
,CORP_CODE
,COST_CENTER
,DATE_CLOSED
,DATE_COMPLETED
,DATE_REQUESTED
,DATE_RESPONDED
,DATE_RESPONSE_ECD
,DATE_SCHEDULED
,DEFERRAL_REASON
,DESCRIPTION
,ECD
,FACILITY_MANAGER
,FLOOR
,GENERAL_LEDGER
-- ,KIOSK_DATE_REQUESTED
-- ,KIOSK_DISPATCH_CONFIRMED
-- ,KIOSK_DISPATCHED
,LINKED_EQUIPMENT_ALIAS
,LINKED_EQUIPMENT_ID
,LINKED_EQUIPMENT_NAME
,ORIGINATOR_TYPE
-- ,PAYMENT_TERMS
,PRIORITY_CODE
,PROBLEM_TYPE
,PROPERTY
,QUOTE_TOTAL
,REGION
,REPAIR_DEFINITIONS
,REPAIR_DESCRIPTION
,REQUESTOR
-- ,REQUESTOR_COST_CENTER
,REQUESTOR_EMAIL
,REQUESTOR_NAME
,REQUESTOR_PHONE
-- ,RESPONSE_TIME
,ROOM
,SERVICE_PROVIDER
,SERVICE_PROVIDER_ADDRESS
,SERVICE_PROVIDER_CITY
,SERVICE_PROVIDER_CODE
,SERVICE_PROVIDER_COUNTRY
,SERVICE_PROVIDER_CURRENCY
,SERVICE_PROVIDER_DESCRIPTION
,SERV_PROV_DISPATC_HMETHOD
,SERV_PROV_DOUBLE_TIME_RATE
,SERVICE_PROVIDER_EMAIL
,SERV_PROV_EMERGENCY_PHONE
,SERVICE_PROVIDER_FAX_NUMBER
,SERVICE_PROVIDER_HOME_PHONE
,SERVICE_PROVIDER_HOURLY_RATE
,SERVICE_PROVIDER_JOB_TITLE
,SERVICE_PROVIDER_METHOD
,SERVICE_PROVIDER_MOBILE_PHONE
,SERVICE_PROVIDER_PAGER
,SERVICE_PROVIDER_RATES
,SER_PROV_SHIFT_DIFFERENTIAL
,SERV_PROV_STATE_PROVINCE
,SERVICE_PROVIDER_STATUS
,SERV_PROV_WEB_SITE_ADDRESS
,SERVICE_PROVIDER_WORK_PHONE
,SERV_PROV_ZIP_POSTAL_CODE
-- ,SHIFT
-- ,SKILL
,STATUS
,SUBREGION
-- ,TRADE
,VENDOR_ID
,VENDOR_PURCHASING_FAX
,VENDOR_SITECODE
,VENDOR_TICKET
,VENDOR_COMPANYNAME
,VIP
,WORK_ORDER_NO
,WORK_REQUEST
,WORK_REQUEST_CLASS
,WORK_TYPE
-- ,WR_COST
,WR_DESCRIPTION
-- ,WR_DISPATCH_METHOD
,WR_STATUS
,COUNTRY
,CREATE_DATE
VALUES
(SELECT DISTINCT
JT.JT_ID AS JT_ID
,NVL((ROUND((JT_DATE_COMPLETED - JT_DATE_REQUESTED) * 24,2)),0) AS ACTUAL_HRS_TO_COMPLETE
,NVL((ROUND((JT_DATE_RESPONDED - JT_DATE_REQUESTED) * 24,2)),0) AS ACTUAL_HRS_TO_RESPOND
,PEO1.PEO_NAME AS AGENT_NAME
,PEO1.PEO_USER_NAME AS ASAGENT_SOE_ID
,LE.LGLENT_DESC AS AP_SYSTEM
-- ,' ' AS ASSIGN_WORK_REQUEST_COMMENT
,DECODE(JT.JT_BILL_ID,138802,'CLIENT BILLABLE'
,138803,'CONTRACTED'
,138804,'INTERNAL BILLABLE',NULL,' ') AS BILLABLE
,BL.BLDG_NAME_CC AS BUILDING
,BL.BLDG_ID_LS AS BUILDING_ID
,DECODE(BL.BLDG_ACTIVE_CC, 'Y', 'ACTIVE', 'INACTIVE') AS BUILDING_STATUS
,DECODE(JT.JT_WRK_CAUSE_ID,141521,'STANDARD WEAR AND TEAR'
,141522,'NEGLIGENCE'
,141523,'ACCIDENTAL'
,141524,'MECHANICAL MALFUNCTION'
,141525,'OVERSIGHT'
,141526,'VANDAL'
,141527,'STANDARD'
,141528,'PROJECT WORK'
,6058229,'TEST',NULL,' ') AS CAUSE_TYPE
-- ,' ' AS COMMENTS
,PEO3.PEO_NAME AS COMPLETED_BY
,JT.JT_REQUESTOR_EMAIL AS CONTACT_EMAIL
,JT.JT_REQUESTOR_NAME_FIRST || ' ' ||JT.JT_REQUESTOR_NAME_LAST AS CONTACT_NAME
,JT.JT_REQUESTOR_PHONE AS CONTACT_PHONE
,CC.CSTCTRCD_APCODE AS CORP_CODE
,CC.CSTCTRCD_CODE AS COST_CENTER
,JT.JT_DATE_CLOSED AS DATE_CLOSED
,JT.JT_DATE_COMPLETED AS DATE_COMPLETED
,JT.JT_DATE_REQUESTED AS DATE_REQUESTED
,JT.JT_DATE_RESPONDED AS DATE_RESPONDED
,JT.JT_DATE_RESPONSE_ECD AS DATE_RESPONSE_ECD
,JT.JT_DATE_SCHEDULED AS DATE_SCHEDULED
,DECODE(JT.JT_DEF_ID,139949,'WTG VENDOR RESPONSE'
,139950,'WAITING ON PARTS'
,139951,'LABOR AVAILABILITY'
,139952,'DEFERRED- HI PRI WORK'
,139953,'WTG APPROVAL'
,139954,'FUNDING REQUIRED'
,139955,'ACCESS DENIED'
,139956,'WTG MATERIAL',NULL,' ') AS DEFERRAL_REASON
,JT.JT_DESCRIPTION AS DESCRIPTION
,JT.JT_DATE_RESCHED_ECD AS ECD
,FMG.FACILITY_MANAGER AS FACILITY_MANAGER
,FL.FLOORS_TEXT AS FLOOR
,GL.GENLED_DESC AS GENERAL_LEDGER
-- ,' ' AS KIOSK_DATE_REQUESTED
-- ,' ' AS KIOSK_DISPATCH_CONFIRMED
-- ,' ' AS KIOSK_DISPATCHED
,EQP.EQUIP_CUSTOMER_CODE AS LINKED_EQUIPMENT_ALIAS
,EQP.EQUIP_ID AS LINKED_EQUIPMENT_ID
,EQP.EQUIP_TEXT AS LINKED_EQUIPMENT_NAME
,DECODE(JT_ORIGINATOR_TYPE_ID,1000,'PROJECT MOVE REQUEST'
,138834,'CUSTOMER INITIATED CORRECTION'
,138835,'CUSTOMER INITIATED REQUEST'
,138836,'CORRECTIVE MAINTENANCE'
,138837,'CONFERENCE ROOM BOOKING'
,138838,'PROJECT INITIATED REQUEST'
,138839,'PLANNED PREVENTIVE MAINTENANCE'
,138840,'SELF INITATED REQUEST',NULL,' ') AS ORIGINATOR_TYPE
-- ,' ' AS PAYMENT_TERMS
,PRIORITY_TEXT AS PRIORITY_CODE
,SWOTY.SWORKTYPE_TEXT AS PROBLEM_TYPE
,PROP.PROPERTY_NAME_CC AS PROPERTY
,JT.JT_COST_QUOTE_TOTAL AS QUOTE_TOTAL
,PAR.LEVELS_NAME AS REGION
,DECODE(JT.JT_REPDEF_ID,141534,'ADJUSTED SETTING'
,141535,'TRAINING FOR END'
,141536,'NEW REQUEST'
,141537,'NO REPAIR REQUIR'
,141538,'REPLACED PARTS'
,141539,'REPLACE EQUIPMEN'
,1000699,'NEW REQUEST',NULL,' ') AS REPAIR_DEFINITIONS
,JT.JT_REPAIRDESC AS REPAIR_DESCRIPTION
,JT.JT_REQUESTOR AS REQUESTOR
-- ,' ' AS REQUESTOR_COST_CENTER
,JT.JT_REQUESTOR_EMAIL AS REQUESTOR_EMAIL
,JT.JT_REQUESTOR_NAME_FIRST AS REQUESTOR_NAME
,JT.JT_REQUESTOR_PHONE AS REQUESTOR_PHONE
-- ,' ' AS RESPONSE_TIME
,RM.ROOM_NAME_CC AS ROOM
,P1.PEO_PROVIDER_CODE1 AS SERVICE_PROVIDER
,P1.PEO_ADDRESS_1 AS SERVICE_PROVIDER_ADDRESS
,PEOCITY.CITY_TEXT SERVICE_PROVIDER_CITY
,P1.PEO_PROVIDER_CODE1 AS SERVICE_PROVIDER_CODE
,PEOCITY.CITY_COUNTRY_NAME AS SERVICE_PROVIDER_COUNTRY
,PEOCUR.CURRENCY_TEXT AS SERVICE_PROVIDER_CURRENCY
,P1.PEO_NAME AS SERVICE_PROVIDER_DESCRIPTION
,P1.PEO_DISPATCH_METHOD AS SERV_PROV_DISPATC_HMETHOD
,P1.PEO_RATE_DOUBLE AS SERV_PROV_DOUBLE_TIME_RATE
,P1.PEO_EMAIL AS SERVICE_PROVIDER_EMAIL
,P1.PEO_EMERGENCY_PHONE AS SERV_PROV_EMERGENCY_PHONE
,P1.PEO_FAX AS SERVICE_PROVIDER_FAX_NUMBER
,P1.PEO_HOME_PHONE AS SERVICE_PROVIDER_HOME_PHONE
,P1.PEO_RATE_HOURLY AS SERVICE_PROVIDER_HOURLY_RATE
,P1.PEO_TITLE AS SERVICE_PROVIDER_JOB_TITLE
,P1.PEO_METHOD_ID AS SERVICE_PROVIDER_METHOD
,P1.PEO_CELL_PHONE AS SERVICE_PROVIDER_MOBILE_PHONE
,P1.PEO_PAGER AS SERVICE_PROVIDER_PAGER
,P1.PEO_RATE_DIFFERENTIAL AS SERVICE_PROVIDER_RATES
,P1.PEO_RATE_DIFFERENTIAL AS SER_PROV_SHIFT_DIFFERENTIAL
,PEOCITY.CITY_STATE_PROV_TEXT AS SERV_PROV_STATE_PROVINCE
,DECODE(P1.PEO_ACTIVE, 'Y', 'ACTIVE', 'INACTIVE') AS SERVICE_PROVIDER_STATUS
,P1.PEO_URL AS SERV_PROV_WEB_SITE_ADDRESS
,P1.PEO_PHONE AS SERVICE_PROVIDER_WORK_PHONE
,P1.PEO_POSTAL_CODE AS SERV_PROV_ZIP_POSTAL_CODE
-- ,' ' AS SHIFT
-- ,' ' AS SKILL
,DECODE(JT.JT_BIGSTATUS_ID,138813,'NEW'
,138814,'PENDING'
,138815,'OPEN'
,138816,'COMPLETED'
,138817,'CLOSED'
,138818,'CANCELLED',NULL,' ') AS STATUS
,LEV.LEVELS_NAME AS SUBREGION
-- ,' ' AS TRADE
,P1.PEO_LS_INTERFACE_CODE1 AS VENDOR_ID
,P1.PEO_FAX AS VENDOR_PURCHASING_FAX
,P1.PEO_VENDOR_SITE_CODE AS VENDOR_SITECODE
,JT.JT_ID AS VENDOR_TICKET
,P1.PEO_NAME AS VENDOR_COMPANYNAME
,JT.JT_REQUESTOR_VIP AS VIP
,WO.WO_ID AS WORK_ORDER_NO
,JT.JT_ID AS WORK_REQUEST
,JT.JT_CLASS_ID AS WORK_REQUEST_CLASS
,WOTY.WORKTYPE_TEXT AS WORK_TYPE
-- ,' ' AS WR_COST
,JT.JT_DESCRIPTION AS WR_DESCRIPTION
-- ,' ' AS WR_DISPATCH_METHOD
,DECODE(JT.JT_BIGSTATUS_ID,138813,'NEW'
,138814,'PENDING'
,138815,'OPEN'
,138816,'COMPLETED'
,138817,'CLOSED'
,138818,'CANCELLED',NULL,' ') AS WR_STATUS
,CTRY.COUNTRY_NAME AS COUNTRY
,SYSDATE --LD_CURR_TIME
FROM CITI.JOBTICKET JT,
CITI.PROPERTY PROP,
CITI.BLDG BL,
CITI.BLDG_LEVELS BLDGLVL,
CITI.LEVELS LEV,
CITI.LEVELS PAR,
(SELECT CRSTOOLS.STRAGG(PEO_NAME) FACILITY_MANAGER,
BLDGCON_BLDG_ID
FROM CITI.BLDG_CONTACTS, CITI.PEOPLE
WHERE BLDGCON_PEO_ID = PEO_ID
AND BLDGCON_CONTYPE_ID IN (40181, 10142)
GROUP BY BLDGCON_BLDG_ID) FMG,
CITI.FLOORS FL,
CITI.ROOM RM,
CITI.GENERAL_LEDGER GL,
CITI.LEGAL_ENTITY LE,
CITI.COST_CENTER_CODES CC,
CITI.EQUIPMENT EQP,
CITI.WORKTYPE WOTY,
CITI.SUBWORKTYPE SWOTY,
CITI.WORK_ORDER WO,
CITI.JT_WORKERS JTWO,
CITI.PRIORITY,
CITI.COUNTRY CTRY,
CITI.PEOPLE P1,
CITI.PEOPLE PEO3,
CITI.PEOPLE PEO1,
CITI.CITY PEOCITY,
CITI.CURRENCY PEOCUR
WHERE JT.JT_BLDG_ID = BL.BLDG_ID
AND BL.BLDG_ID = BLDGLVL.BLDG_LEVELS_BLDG_ID
AND BLDGLVL.BLDG_LEVELS_LEVELS_ID = LEV.LEVELS_ID
AND LEV.LEVELS_PARENT = PAR.LEVELS_ID(+)
AND PROP.PROPERTY_ID = BL.BLDG_PROPERTY_ID
AND BL.BLDG_ACTIVE_LS = 'N'
AND JT.JT_FLOORS_ID = FL.FLOORS_ID(+)
AND JT.JT_ROOM_ID = RM.ROOM_ID(+)
AND JT.JT_BLDG_ID = FMG.BLDGCON_BLDG_ID(+)
AND JT.JT_GENLED_ID = GL.GENLED_ID(+)
AND GL.GENLED_LGLENT_ID = LE.LGLENT_ID(+)
AND JT.JT_CSTCTRCD_ID = CC.CSTCTRCD_ID(+)
AND JT.JT_EQUIP_ID = EQP.EQUIP_ID(+)
AND JT.JT_ID = JTWO.JTW_JT_ID(+)
AND JT.JT_WORKTYPE_ID = WOTY.WORKTYPE_ID(+)
AND JT.JT_SWORKTYPE_ID = SWOTY.SWORKTYPE_ID(+)
AND JT.JT_WO_ID = WO.WO_ID
AND JT.JT_PRIORITY_ID = PRIORITY_ID(+)
--AND jt.jt_date_requested >= ADD_MONTHS (SYSDATE, -12)
AND JT.JT_LAST_UPDATE >= ADD_MONTHS(LD_CURR_TIME, -12)
AND BL.BLDG_COUNTRY_ID = CTRY.COUNTRY_ID
AND JTWO.JTW_PEO_ID = P1.PEO_ID(+)
AND P1.PEO_CITY_ID = PEOCITY.CITY_ID(+)
AND JT.JT_COMPLETED_BY_PEO_ID = PEO3.PEO_ID(+)
AND P1.PEO_RATE_CURRENCY_ID = PEOCUR.CURRENCY_ID(+)
AND JT.JT_AGENT_PEO_ID = PEO1.PEO_ID(+)
COMMIT;
EXCEPTION
WHEN OTHERS THEN
ROLLBACK;
DBMS_OUTPUT.PUT_LINE('SQLCODE :' || SQLCODE || ' Error :' || SQLERRM);
END WORK_KIOSK_FULL;Here's the link for infor on the [Oracle Direct-Path INSERT |http://download.oracle.com/docs/cd/B10501_01/server.920/a96524/c21dlins.htm#10778].
Also, if you are truly intent on using a CURSOR FOR LOOP with the BULK COLLECT, I suggest you read Steven Feuerstein's article [PL/SQL Practices: On BULK COLLECT |http://www.oracle.com/technology/oramag/oracle/08-mar/o28plsql.html].
Hope this helps.
Craig...
If my response or the response of another was helpful, please mark it accordingly -
How to call BULK COLLECT using JDBC in JSP?
Hi
I am using BULK COLLECT instead of CURSOR in my Stored Procedure.
How to use these stuff in JDBC.
If i am using CURSOR means
cs.registerOutParameter(1,oracle.jdbc.driver.OracleTypes.CURSOR);
is used.
Which types i have to use?
I am retriving morethan one record using FOR. LOOP in SP and before that
BULK COLLECT is used in SELECT query..
Can you give jsp code for this?
Thanksyou may find related sample jdbc code on otn - http://www.oracle.com/technology/sample_code/tech/java/sqlj_jdbc/index.html
Best regards. -
Dynamic sql with dynamic bulk collection variable
Hi,
I am facing the issue while bulk collecting dynamic sql query data into dynamic variable.
Eg:
query1:= << dynamic select query>>
Execute immediate query1 bulk collect into Dynamic_varibale;
here dynamic_varible is pl/sql table type with 1 column.
How do i declare "dynamic_variable" here????
please suggest...create type t_id is table of number
SQL> create type t_id is table of number
2 /
Type created.
SQL> declare
2
3 v_tid t_id;
4 v_results sys_refcursor;
5
6 v_employee_id number;
7 v_name varchar2(100);
8
9 v_sql varchar2(1000);
10
11
12 begin
13 v_tid := t_id(7902,7934);
14
15 --
16
17
18 v_sql := 'select empno, ename from scott.emp ' ||CHR(10)
19 || 'where empno in (select column_value from table(cast(:v_tid as
t_id)))';
20
21 dbms_output.put_line(v_sql);
22 dbms_output.put_line('----------');
23
24 open v_results for v_sql using v_tid;
25
26
27 IF v_results IS NOT NULL
28 THEN
29 LOOP
30 FETCH v_results
31 INTO v_employee_id, v_name;
32
33 EXIT WHEN (v_results%NOTFOUND);
34 dbms_output.put_line(v_name);
35 END LOOP;
36
37 IF v_results%ISOPEN
38 THEN
39 CLOSE v_results;
40 END IF;
41 END IF;
42
43 end;
44 /
select empno, ename from scott.emp
where empno in (select column_value from
table(cast(:v_tid as t_id)))
FORD
MILLER -
Which is better??? for loop or bulk collect
declare
cursor test is
select * from employees;
begin
open test;
loop
exit when test%notfound;
fetch test into myvar_a(i); CASE A
i:=i+1;
end loop;
close test;
open test;
fetch test bulk collect into myvar_b; CASE B
close test;
end;
Which case is better?? A or B?
Edited by: Kakashi on May 31, 2009 12:54 AMDepends on the meaning of better.
Generally case B should be faster although a bit more elaborate code is required.
But there may be exceptions. I think I read somewhere (I'm home now and I cannot find it at the moment) that in 10g (or 11g - not sure) 100 rows at a time are pre-fetched behind scenes even when you use case A. So using case B with a low limit could well be slower.
If I can express an additional opinion case F(irst) is nearly always the best i.e. plain SQL (no loops at all). I'm aware that sometimes it cannot be used, but should be the first approach to be tried.
Regards
Etbin
FOUND: http://asktom.oracle.com/pls/asktom/f?p=100:11:0::::P11_QUESTION_ID:213366500346264333
CONTAINS
Hey Tom, love the site. I noticed in your first fetch, which was in the first for loop that did an unconditional exit:
2 for x in ( select rownum r, t1.* from big_table.big_table t1 )
3 loop
4 exit;
5 end loop;
In looking at the TKPROF output for that query, it shows the number of rows being fetched as 100. Does that prove / demonstrate the bulk collecting optimization that Oracle added in 10g, where it implicitly and automatically does a bulk collect of limit 100 behind the scenes?
This came up at a discussion at my site very recently, and I think I can just point them to your example here as a demo rather than creating my own. I assume that if you ran the same thing in 9iR2, then that first fetch of rows in TKPROF would only show 1?
Followup April 18, 2007 - 1pm US/Eastern:
yes, that demonstrates the implicit array fetch of 100 rows...
in 9i, it would show 1 row fetched.
Edited by: Etbin on 31.5.2009 10:38
Maybe you are looking for
-
Ive installed a couple of apps such as dropbox, keycue and zooskmessanger, when i scan using iBoost they show in the startup items but when i go to system preferences they are not there and i want to disable them since they are making my startup very
-
Select query on two Database views
Hi all, Can i fetch the data by writing a select query on two DATABASE VIEWS Because i am able to fetch data by writing a selct query on ONE DATABASE VIEW and ON TRANSPARANT TABLE but i am not able to fetch data by writing a query on TWO DATABASE VIE
-
Can SQL injection output rows to hacker?
Can a hacker retrieve rows through SQL injection or simply just jumble up the data? I wouldn't see how they could get the rows without coldfusion code that will actually be instructed to output the query. If not, are there any hot cf/mssql hacking te
-
I have an autorun on my CD project and it takes too long to open. I have created a stub projector with Shockwave compression, Xtras removed and placed in an Xtras folder on the same directory level as the projector, and have included the application
-
Defualt Output Types on Billing Document Upon Creation
Need to know if it is possible through configuration to default specific output types on a Billing Document upon creation. We have created a few SAP Smart Forms for print for SD Billing Documents that typically we would have users set up via transact