Need to run start routine only once

I understand that for every data package from ODS to Cube, the start routine is run. But I need to run it only once. Is there anyway to achieve this? As the programming logic doesn't really do anything to the data package, I don't need it to be in the start routine. But as I populate a database table using that routine, which in turn is used in update routines, I need to kick it off before the load from ODS to cube happens. I think an ABAP routine variant can be kicked off in process chains. Did anyone use the variant and know how to kick off the routine? Do I have to write this logic in a function module and then kick it off using ABAP variant in Process chains?
I would prefer a way to limit my start routine to run once for every data load rather than kicking it off through process chain. Can someone discuss which is a good approach and how I go about it? Thanks for your inputs.
Sam

Hi Sameer,
Let me recap you req. You want to populate a database table in your start routine and need to do it once for a load.
There is no possible way to limit the no. of times a Start routine is run, but you can find some worksaround.
Is there anyway you can identify that your DB table is updated with new records, i.e.
1> You delete and refill table every time
2> You have some kind of time stamps.
If one of the above case is true, then you can use that as a check in the Start routine to limit the no. of times the DB table is being populated.
Regards,
Sree

Similar Messages

(BW) Run the Start Routine Only Once

I understand that for every data package from ODS to Cube, the start routine is run. But I need to run it only once. Is there anyway to achieve this in ABAP?

I once had that same requirement, but in a datasource in R3. I'm not sure if the same solution would work in BW, though. I used a memory id to keep the variable value between packets:
DATA: ... n_globalvar TYPE n ...
then I added...
IMPORT n_globalvar FROM MEMORY ID 'ZMEMID01'.
...at the start of the routine to retrieve to the variable the value from the memory id.
At the end of the code, I exported the variable back to the same memory id...
EXPORT n_globalvar to MEMORY ID 'ZMEMID01'.

Execute ABAP code in Start Routine only once

I have ABAP code in a start routine that I only want to execute once. Is there a way to tell this is first execution of the start routine?
Also can I find out these value in the start routine ABAP:
1. How many data packets there are?
2. What data packet I am processing?
Regards,
Mike...

Hi,
I've just setup a little test:
TRules, Start Routine Global Decl:
DATA: GVI TYPE I, GVN(5) TYPE N.
Start Routine:
IF GVI IS INITIAL.
GVI = 1.
ELSE.
ADD 1 TO GVI.
ENDIF.
MOVE GVI TO GVN.
This is basically counting the times the start routine is executed and populating the var GVN; I am posting GVN in a char ZTESTGV in the target cube. I am posting as well the datapakID in my cube.
I have loaded two requests (each 4 packets)
the first one first to PSA and then from the PSA to the target: this is Serially using one single process. Here's the result in the cube:
Request________________________________DATAPAKID_____ZTESTGV
REQU_8B5ASGQNYYNIV1OJN68HFH1SR____________1____________1
REQU_8B5ASGQNYYNIV1OJN68HFH1SR____________2____________2
REQU_8B5ASGQNYYNIV1OJN68HFH1SR____________3____________3
REQU_8B5ASGQNYYNIV1OJN68HFH1SR____________4____________4
In this case the global variable is persistent accross packets.
the second request is loaded in paralell: this is, multiple processes are executed at the time. Here's the result in the cube:
Request________________________________DATAPAKID_____ZTESTGV
REQU_14QVH21BSVH44FAJW94BD7N2H____________1____________1
REQU_14QVH21BSVH44FAJW94BD7N2H____________2____________1
REQU_14QVH21BSVH44FAJW94BD7N2H____________3____________1
REQU_14QVH21BSVH44FAJW94BD7N2H____________4____________1
In this case the global variable is always1 !!
This is logic since several a process cannot access the internal memory used by another one...
Conclusion; the global variable will work only if the load is serial...
hope this helps...
Olivier.

Start routine: only once per teh load but not once per data packet

Hi,
I would like to execute some code in start routine (in update rules) only once per the load but not once per the data packet.
How can I implement this.
Regards,

I once had that same requirement, but in a datasource in R3. I'm not sure if the same solution would work in BW, though. I used a memory id to keep the variable value between packets:
DATA: ... n_globalvar TYPE n ...
then I added...
IMPORT n_globalvar FROM MEMORY ID 'ZMEMID01'.
...at the start of the routine to retrieve to the variable the value from the memory id.
At the end of the code, I exported the variable back to the same memory id...
EXPORT n_globalvar to MEMORY ID 'ZMEMID01'.

Why java allow start() method only once for a thread

Hi ,
Why java allows start method only once for thread . suppose
Thread t = new Thread();
t.start();
say at later stage if again we call t.start() IllegalStateException is thrown , even though isAlive method returns false.
Hence the question , why start() method is allowed only once.If you need start a thread , we need to create a new instance.

Really. Why do you think that? Do you have any evidence? It is one of the first things I would think of, personally.Considering that the Thread API doesn't allow you to specify a stack address (only stack size), I think it demonstrates they wanted to remove that capability from their Thread API all together. That missing "capability" makes me believe they want me to believe it's not something I need to worry about when using their API... I think the exact semantics of the Thread class and its methods were driven by how to make it most understandable and usable for their customers. I'm certain this issue was one of many that was given considerable thought during the design and implementation of the JVM and the underlying runtime classes.
Do I have any evidence? No. But if you can point me at some first-hand information on this, I'd love to read it. Most of what I've found is second or third hand accounts. (and I mean that sincerely, not as a smart-ass remark or rebuke of your comments).
On the one hand you seem to think the Java API designers are idiots, on the other hand you think that they should be. I can't make it out.I thought my position was that the Java developers were talented enough to implement platform in whatever way their API called for; hence, the designers made a choice about how they wanted their API to be used by their customers. They decided which capabilities they wanted to include or exclude, and created an API that was consistent with their vision for this technology. While I'm certain technical limitations had an effect on the API design, I certainly don't think the API was dictated by the them.
I think the current design of the Java Thread API was a reflection of their vision to make Threading easier and more accessible to Joe Programmer, not limitations in the implementation. I never said it was wrong or that I could do better... I just said I think they could have done something different if they decided it would have made for a better customer experience. But hey, maybe I'm wrong.

Running a query only once for a report in Reports 6i in a BEFORE REPORT trigger.

Hello all -
I am using Oracle Reports 6i on Windows NT 4.0 SP 6.0.
The report I am converting from Access to Oracle Reports is
rather complex, and features detail by decile (a rank) within
territory (geographical sales area). Also it features a summary
at the bottom of each territory/decile combination which
summarizes not only the territory information, but goes above
that to higher levels like district, region, and national.
With the help of a consultant, we have managed to get the report
almost finished. However, we are running into a snag with the
summary. While my two main queries need to run in the data
model as normal, I would like to run the summary query _only
once_ during the entire cycle of the report, and have the report
fields populated once for the entire report. This is possible
because for a district, the summary numbers are the same on
every page of the report, only the territory information
changes. So by only having the query run once and fill the
report values once, this would save considerably on the report
runtime.
However, I have tried a few different methods but cannot get the
report to recognize the fields in the query, most likely because
they are out of scope at report creation. Does anyone have any
ideas on how I can accomplish this in Reports? Any help would
be appreciated - I hope my question was clear, if not please let
me know.
Thanks,
-Jennifer Prichard

Hello!
You can place in the data model editor formula and placeholder
columns outside of any groups. I think you need for each of your
summary-attibutes a placeholder column and one formula column.
Use the formula column to populate the placeholders via PL/SQL.
Afterwords you can reference the placeholders anywhere you want.
Regards,
Hajo Winkler

Need print popup to display only once

Hi Experts ,
I have a customer account statement in smartform . For multiple customers the clients wants the print dialog to be displayed only once and the options mentioned in the first display to be carried for other customer accounts . Please any suggestions for the same .

Hi,
If using a custom print program you can achieve this using functions SSF_OPEN and SSF_CLOSE:
[Printing Several Forms in One Print Request|http://help.sap.com/saphelp_nw70/helpdata/en/64/bf2f12ed1711d4b655006094192fe3/content.htm]
This way the print dialog will be displayed only once no matter how many times the smartform is called.
Che.

Help needed for Graphical mapping.Tag only once produces

Hello experts,
I have a problem in my graphical mapping.This is the example:
from each segment A in IDOC has to be made Tag A or Tag B in output XML.
Segment A has a field USER_STATUS and if it is X than Tag B has to be made otherwise Tag A.
Segment A is 0..9999999. occurence
Tag A 0..unbounded
Tag B 0..unbounded.
I made a mapping
mapping for Tag A:
USER_STATUS -> mapwith default(empty)->length->EqualsA (with Constant =0)-> createif->Tag A
for Tag B as following:
USER_STATUS -> mapwith default(empty)->EqualsS (with Constant =X)-> createif->Tag B
The problem is that in output XML, I have only one Tag A or B (depending on field USER_STATUS in first occurence of the segment) while in inbound IDOC 5 segments are with different value of the field: USER_STATUS.
Does anyone idea to solve this problem!
Thank you very much for your effort.
King regards,
Danijela Zivanovic

HI,
CAn you try this way,
Use Maping Node UseOneAsMany for Tag B
and pass 3 inputs as
Constant[]----
>
USER_STATUS -> mapwith default(empty)->EqualsS (with Constant =X)----
>
USER_STATUS -> mapwith default(empty)->EqualsS (with Constant =X)----
>
Same way try for Tag A also.
Thanks
Swarup

Start routine in transformation

Hi,
    I need to read master data in a start routine only once.If I write a select statement in the start routine , the select is executed for every data package.I want the select to be executed only once and data can be used by all data packages.I tried keeping the code in the global section of the start routine class and that didn't help either.
   Any ideas.
Thanks in advance

Hi Siva,
Then..Use Static Internal table...
Structured Static Internal Table Definition
- STATICS itab TYPE tabtype [WITH HEADER LINE].
- STATICS itab TYPE tabkind OF linetype
[WITH [UNIQUE|NON-UNIQUE] keydef]
[INITIAL SIZE n] [WITH HEADER LINE].
- STATICS itab LIKE tabkind OF lineobj
[WITH [UNIQUE|NON-UNIQUE] keydef]
[INITIAL SIZE n] [WITH HEADER LINE].
- STATICS itab TYPE linetype OCCURS n [WITH HEADER LINE].
- STATICS itab LIKE lineobj OCCURS n [WITH HEADER LINE].
- STATICS: BEGIN OF itab OCCURS n,
END OF itab [VALID BETWEEN f1 AND f2].
STATICS itab TYPE RANGE OF type.
STATICS itab LIKE RANGE OF f
@@@@@@@@@@@@@22
Logic: make flag variable as static variable in global area.
First data package...
IF flag EQ ' '.
Read Master data table and put in STATIC internal TABLE.
So the next time(next data package)..not read from master data table
flag = 'X'.
ENDif.
LOOP DATA_PACKAGE.
ENDLOOP.
@@@@@@@@@@@@2
Regards,
San!
Message was edited by:
        Sandeep Khatri(San!)

Help: I want to auto schedule a load using file watcher but it runs only once for the first time and after that it is not running at all

Hi All,
I am trying to execute the below code as provided from one of the blogs. i am able to run the job only once based on a file watcher object(i.e. for very first time) and after that the job is not running at all and if i schedule the job to run automatically based on interval of 10 or more minutes it is executing properly). Please let me know or guide me if i have missed any step or configuration.that is needed.
Version of Oracle 11.2.0.1.0
OS : Windows 7 Prof
Given all the necessary privileges
BEGIN
DBMS_SCHEDULER.CREATE_CREDENTIAL(
     credential_name => 'cred',
     username        => 'XXXX',
     password        => 'XXXX');
END;
CREATE TABLE ZZZZ (WHEN timestamp, file_name varchar2(100),
   file_size number, processed char(1));
CREATE OR REPLACE PROCEDURE YYYY
(payload IN sys.scheduler_filewatcher_result) AS
BEGIN
INSERT INTO ZZZZ VALUES
     (payload.file_timestamp,
      payload.directory_path || '/' || payload.actual_file_name,
      payload.file_size,
      'N');
END;
BEGIN
DBMS_SCHEDULER.CREATE_PROGRAM(
    program_name        => 'prog1',
    program_type        => 'stored_procedure',
    program_action      => 'YYYY',
    number_of_arguments => 1,
    enabled             => FALSE);
DBMS_SCHEDULER.DEFINE_METADATA_ARGUMENT(
    program_name        => 'prog1',
    metadata_attribute => 'event_message',
    argument_position   => 1);
DBMS_SCHEDULER.ENABLE('prog1');
END;
BEGIN
DBMS_SCHEDULER.CREATE_FILE_WATCHER(
    file_watcher_name => 'file_watcher1',
    directory_path    => 'D:\AAAA',
    file_name         => '*.txt',
    credential_name   => 'cred',
    destination       => NULL,
    enabled           => FALSE);
END;
BEGIN
DBMS_SCHEDULER.CREATE_JOB(
    job_name        => 'job1',
    program_name    => 'prog1',
    queue_spec      => 'file_watcher1',
    auto_drop       => FALSE,
    enabled         => FALSE);
DBMS_SCHEDULER.SET_ATTRIBUTE('job1','PARALLEL_INSTANCES',TRUE);
END;
EXEC DBMS_SCHEDULER.ENABLE('file_watcher1,job1');
Regards,
kumar.

Please post a copy and paste of a complete run of a test case, similar to what I have shown below.
SCOTT@orcl12c> SELECT banner FROM v$version
2 /
BANNER
Oracle Database 12c Enterprise Edition Release 12.1.0.1.0 - 64bit Production
PL/SQL Release 12.1.0.1.0 - Production
CORE    12.1.0.1.0    Production
TNS for 64-bit Windows: Version 12.1.0.1.0 - Production
NLSRTL Version 12.1.0.1.0 - Production
5 rows selected.
SCOTT@orcl12c> CONN / AS SYSDBA
Connected.
SYS@orcl12c> -- set file watcher interval to one minute:
SYS@orcl12c> BEGIN
2    DBMS_SCHEDULER.SET_ATTRIBUTE
3       ('file_watcher_schedule',
4        'repeat_interval',
5        'freq=minutely; interval=1');
6 END;
7 /
PL/SQL procedure successfully completed.
SYS@orcl12c> CONNECT scott/tiger
Connected.
SCOTT@orcl12c> BEGIN
2    -- create credential using operating system user and password (fill in your own):
3    DBMS_SCHEDULER.CREATE_CREDENTIAL
4       (credential_name     => 'cred',
5        username          => '...',
6        password          => '...');
7 END;
8 /
PL/SQL procedure successfully completed.
SCOTT@orcl12c> -- create table to insert results into:
SCOTT@orcl12c> CREATE TABLE ZZZZ
2    (WHEN      timestamp,
3      file_name varchar2(100),
4      file_size number,
5      processed char(1))
6 /
Table created.
SCOTT@orcl12c> -- create procedure to insert results:
SCOTT@orcl12c> CREATE OR REPLACE PROCEDURE YYYY
2    (payload IN sys.scheduler_filewatcher_result)
3 AS
4 BEGIN
5    INSERT INTO ZZZZ VALUES
6        (payload.file_timestamp,
7         payload.directory_path || '/' || payload.actual_file_name,
8         payload.file_size,
9         'N');
10 END;
11 /
Procedure created.
SCOTT@orcl12c> -- create program, define metadata, and enable:
SCOTT@orcl12c> BEGIN
2    DBMS_SCHEDULER.CREATE_PROGRAM
3       (program_name          => 'prog1',
4        program_type          => 'stored_procedure',
5        program_action      => 'YYYY',
6        number_of_arguments => 1,
7        enabled          => FALSE);
8    DBMS_SCHEDULER.DEFINE_METADATA_ARGUMENT(
9       program_name         => 'prog1',
10       metadata_attribute => 'event_message',
11       argument_position   => 1);
12    DBMS_SCHEDULER.ENABLE ('prog1');
13 END;
14 /
PL/SQL procedure successfully completed.
SCOTT@orcl12c> BEGIN
2    -- create file watcher:
3    DBMS_SCHEDULER.CREATE_FILE_WATCHER
4       (file_watcher_name   => 'file_watcher1',
5        directory_path      => 'c:\my_oracle_files',
6        file_name          => 'f*.txt',
7        credential_name     => 'cred',
8        destination          => NULL,
9        enabled          => FALSE);
10 END;
11 /
PL/SQL procedure successfully completed.
SCOTT@orcl12c> BEGIN
2    -- create job:
3    DBMS_SCHEDULER.CREATE_JOB
4       (job_name          => 'job1',
5        program_name          => 'prog1',
6        queue_spec          => 'file_watcher1',
7        auto_drop          => FALSE,
8        enabled          => FALSE);
9    -- set attributes:
10    DBMS_SCHEDULER.SET_ATTRIBUTE ('job1', 'PARALLEL_INSTANCES', TRUE);
11 END;
12 /
PL/SQL procedure successfully completed.
SCOTT@orcl12c> -- enable:
SCOTT@orcl12c> EXEC DBMS_SCHEDULER.enable ('file_watcher1, job1');
PL/SQL procedure successfully completed.
SCOTT@orcl12c> -- write file (file must not exist previously):
SCOTT@orcl12c> CREATE OR REPLACE DIRECTORY upncommon_dir AS 'c:\my_oracle_files'
2 /
Directory created.
SCOTT@orcl12c> declare
2    filtyp utl_file.file_type;
3 begin
4    filtyp := utl_file.fopen ('UPNCOMMON_DIR', 'file1.txt', 'W', NULL);
5    utl_file.put_line (filtyp, 'File has arrived ' || SYSTIMESTAMP, TRUE);
6    utl_file.fclose (filtyp);
7 end;
8 /
PL/SQL procedure successfully completed.
SCOTT@orcl12c> -- wait long enough (may take more than one minute) for job to run:
SCOTT@orcl12c> EXEC DBMS_LOCK.SLEEP (100)
PL/SQL procedure successfully completed.
SCOTT@orcl12c> -- check for results:
SCOTT@orcl12c> SELECT * FROM zzzz
2 /
WHEN
FILE_NAME
FILE_SIZE P
22-OCT-13 10.12.28.309000 PM
c:\my_oracle_files/file1.txt
        57 N
1 row selected.
SCOTT@orcl12c> declare
2    filtyp utl_file.file_type;
3 begin
4    filtyp := utl_file.fopen ('UPNCOMMON_DIR', 'file2.txt', 'W', NULL);
5    utl_file.put_line (filtyp, 'File has arrived ' || SYSTIMESTAMP, TRUE);
6    utl_file.fclose (filtyp);
7 end;
8 /
PL/SQL procedure successfully completed.
SCOTT@orcl12c> -- wait long enough (may take more than one minute) for job to run:
SCOTT@orcl12c> EXEC DBMS_LOCK.SLEEP (100)
PL/SQL procedure successfully completed.
SCOTT@orcl12c> -- check for results:
SCOTT@orcl12c> SELECT * FROM zzzz
2 /
WHEN
FILE_NAME
FILE_SIZE P
22-OCT-13 10.12.28.309000 PM
c:\my_oracle_files/file1.txt
        57 N
22-OCT-13 10.14.08.580000 PM
c:\my_oracle_files/file2.txt
        57 N
2 rows selected.

Need to display records only once for that particular group

Hi,
I have a requirement like displaying repeated records only once for that particular group.
For eg, in the emp table, I need to display repeated deptno only once for the particular group of job.
And for the above requirement I have used the below query,
SELECT empno, DECODE (LAG (job, 1) OVER (ORDER BY job), job, NULL, job),
DECODE (LAG (deptno, 1) OVER (ORDER BY deptno), deptno, NULL, deptno)
FROM emp;
Output:
EMPNO     JOB     DEPTNO
7782          10
7934     CLERK
7839     PRESIDENT
7876          20
7788
7902     ANALYST
7566     MANAGER
7844          30
7900
7654     SALESMAN
7698
But in the above output you can find that, say deptno 10 is getting displayed only once, whereas I want that deptno 10 to be checked whether it is getting repeated within the group of JOB and than hide the deptno 10 only if it is repeated within that job group. If not, deptno 10 should be displayed again.
Please help me in this.
Regards,
Shiva

Hi,
Hope the below helps.
SELECT emp_id, job, deptno,
       CASE WHEN LAG(deptno, 1) OVER (partition by job order by deptno) = deptno THEN null else deptno end
FROM empRegards
Ameya

Need help in executing EEM applet only once

Do we have any sample script for EEM applet counter ? We want applet to execute once after reload after matching a string . We tried couple of solutions like “ server suspend “( customer does not like server suspend) and event counter ( this does not allow syslog string matching ). Please let me know if there are any options.
At the router reload, match the string , bring down the interface and run this applet only once . I have been trying multiple messages in Bootup log for this for bringing the interface down.
The issue we are facing is that most of the bootup log messages can be triggered when the router is up as well which will result in triggering of the applet which is what we want to avoid.
event manager applet Interface_down
event syslog pattern "SPA removed from subslot 0/0"
trigger delay 10
action 1.0 cli command "enable"
action 1.5 cli command "config t"
action 2.0 cli command "interface gi0/0/0"
action 2.5 cli command "shutdown"
action 3.0 cli command "end"
event manager applet Interface_up
event syslog pattern "Bulk Sync succeeded"
trigger delay 100
action 1.0 cli command "enable"
action 1.5 cli command "config t"
action 2.0 cli command "interface gi0/0/0"
action 2.5 cli command "no shutdown"
action 3.0 cli command "end"
end

This is a bit dirty but you can create an applet to run at startup that creates the event detection applet. Then at the end of the detection applet, have it remove itself from the running config. This means that the applet running at startup will always create the event detection applet but the event detection applet will only ever run once for that router boot period as it deletes itself after first run.
The problem is that when you want to have one applet create another applet, you have issues with the inverted commas being correctly configured for the second applet which will cause it to run once, but fail after that. The way you can get around this is to write the applet to a text file on the flash and then copy the contents of the text file to the running config at boot. This preserves the structure of the applet being created and ensures that it will function correctly.
The following example builds the event detection applet at system restart (BUILDAPPLET) by copying the file "eventapplet.txt" to running config. The event detection applet (EVENTDETECT) detects a syslog pattern, in this case exiting global config, and runs the applet sending a puts command and outputting HELLO. The event detection applet then removes itself from the running config.
!# configure the router
conf t
!# turn off file prompting to let the BUILDAPPLET run when copying
file prompt quiet
!# create the applet that runs at system restart and copies the flash applet to the running config
event manager applet BUILDAPPLET
event syslog pattern "%SYS-5-RESTART"
action 10 cli command "en"
action 11 cli command "copy flash:eventapplet.txt running-config"
!# exit
end
!######## Create the flash file containing your run once applet ########
! enter the tclsh so we can write our applet to a file - put your event detection applet in here between the {} that you want to run once
tclsh
puts [open "flash:eventapplet.txt" w+] {
event manager applet EVENTDETECT
event syslog pattern "%SYS-5-CONFIG_I"
action 1.0 puts "HELLO"
action 2.0 cli command "en"
action 3.0 cli command "conf t"
action 4.0 cli command "no event manager applet EVENTDETECT"

To run a scenario exactly once in a specific time every day

I would like to run a scenario which will read a file in a specific directory and will update the data in the database.
File -> JDBC
I would like to run the scenario only once a day in a specified time
i.e. say 12:00 midnight.
How can I specify that in the scenario?

Hi
Note:You will need to have the authorizations of the user group SAP_XI_ADMINISTRATOR with the role modify.
Go to Runtime Workbench -> Component Monitoring -> Communication Channel Monitoring
Locate the link Availability Time Planning on the top right corner of your Communication Channel Monitoring page.
In your case, the requirement is to schedule the Sender file adapter daily once at 12:00 at midnight.
In Availability Time Planning, choose the Availability time as daily and say create.
Provide the details like the time 12:00
Then select the communication channel , goto the Communication Channels tab and filter and add the respective channel (File Sender).
Once all the above has been done 'Save' the changes.
hope it helps
Thanks

Performance with select in Start routine

Hi,
I open a request because i have some troubles in a update rules.
In this update rule i have several select function on master data and i takes just 3-4 minutes by packets (which is OK for my daily extraction). But my users needs now a data taht i need to extract from SD invoice ODS. and in this ODS i have 3 000000 lines.
I made the test and it takes too many times.
I try several solutions as :
Select.... into internal table --> too many times
Select ... for all entries .... into internal table...--> Too many times
I tried to create a small ODS with just the key level and data i need for my start routine based on invoice ODS. And i now from 3000000 to 300000 records. but now it takes 15 minutes by packets.
My question is do you have any idea to improve performance ?
What is the max size or number of records for an internal tables ?
Thanks
Cyril

Reading that much, you might have some memory issue.
"Data objects that are defined either with the data type of an internal table, or directly as an internal table, are always fully defined in respect of their line type, key and access method. However, the number of lines is not fixed. Thus internal tables are dynamic data objects, since they can contain any number of lines of a particular type. The only restriction on the number of lines an internal table may contain are the limits of your system installation. The maximum memory that can be occupied by an internal table (including its internal administration) is 2 gigabytes. A more realistic figure is up to 500 megabytes. An additional restriction for hashed tables is that they may not contain more than 2 million entries. The line types of internal tables can be any ABAP data types - elementary, structured, or internal tables. The individual lines of an internal table are called table lines or table entries. Each component of a structured line is called a column in the internal table."
if you need the link to the above doc....
http://help.sap.com/saphelp_47x200/helpdata/en/fc/eb3660358411d1829f0000e829fbfe/frameset.htm
thanks.
Wond

Simple start routine

Hi friends,
I need a simple start routine code.
It should only load data for first quarter.
Fiscal period is the time characteristic.
3 months in a quarter Example : 200701 - 200703.
Thankyou.

Assumed that data_package can be filtered based on Year & Month.If it should be filtered based on some date field then there should be minor modifications.
DATA : V_YEAR TYPE C(4),
V_LOW TYPE C(6),
V_HIGH TYPE C(6).
RANGES : R_YRMON LIKE <Use the same field reference of data_package table for the date field>
V_YEAR = SY-DATUM+0(4).
CONCATENATE V_YEAR '01' INTO V_LOW.
CONCATENATE V_YEAR '03' INTO V_HIGH.
R_YRMON-SIGN = 'I'.
R_YRMON-OPTION = 'BT'.
R_YRMON-LOW = V_LOW.
R_YRMON-OPTION = V_HIGH.
APPEND R_YRMON.
CLEAR R_YRMON.
*This filters out all data other than the first quarter
DELETE DATA_PACKAGE WHERE <DATE FIELD> NOT IN R_YRMON.

Need to run start routine only once

Similar Messages

Maybe you are looking for