DSXchange

John, Whenever there is failure of the job, all the datastage logs you can write to the file.. Dsjob function can give you all the log depending upon your requirements. Here you can write all the info message/warnings/errors. So create one file which will be amended whenever a parallel job executes....

This can be catergozierd as one of the most broad quetions.. most of the times the job failure reasons are diffrent and you can not use the same method to debug the job using the log. Most of the times you can check where exactly the error is and what description that error is refering to 8) By look...

One of the options I can suggest is that you can put trigger on your table which is going to load.. And that trigger in turn can call the procedure. You can have trigger at trancation level or as row level. However the limitation with this apporach is that you can not have commit statements in the t...

You can use DSExecute function to execute the command. You can excecute the unix script using above command. In the unix script you can your insert statements. Routine - > DSExecute - > Unix Script - > Database inserts.. This could be one of the way doing what you want to achieve... ----------------...

Hi..

Please refer below link :-

http://publib.boulder.ibm.com/infocente ... 1325n.html

u will get better understanding what exactly caused the error..

Hope this helps..

----------
MK

What is your main purpose..? to run the unix script to get the sequential file processing to next job .. rite ?? :wink: or u r trying to test the funtionality of how unix script is executed when used in the DS jobs.. When unix script is called from unix job, unix excutes it and it does not run in pa...

If before join stage , in the same job u r using database stage, it will easier for you to modify the query..!

Just in case if your doing some transformations and then using the values from repsective tables then u can use the approach suggested by the post above..


hope this helps

MK

Hi.. this seems to be issue with DS version.. I was knowing one of the fatal error which is related to database stage and job was not aborting due to that error.. I guess it will need a patch from DS.. btw check out the both db and ds versions and corrosponding patches .. hope that will help.. -----...

if out of 11 datasets 6 are truncated successfully..
i will advise that u can try running again remaining 5 again !
as it seems to be unix enviornmental issue.

---hope this helps---

MK

-14 is means the it is waiting for the job, however job is not started for the certain time period.. i guess default value for this is 30 mins .. there could be some resource issues with your ds server that it is taking so much time to start that job.. you can try re-running the sequence again.. hop...

u can use trim also to remove the '-' ..
Use option as 'A'..

if you want to do it in unix ..use cut -c

hope this helps

--
Mk

The major performance imporvement can be acheived by avoiding the unnecessary sorting. As ray said, u can check the input properties of the most of the stages which will provide you the sorting option.. However if you still wants to check how sorting should be done check the suage of the APT_NO_SORT...

there is command dssearch..
it searches specific job used in the given number of seqeunces.

check out the parameters..
u need to give the job name ..
and the category..

once this is given, it will show the name of sequences in which the job is used..

----------

Hope this helps

---------

MK

When you are trying to import metadata..there is option of tables.. when u click on that.. it shows that what it should include 1.table 2.system tables 3.synonyms.. u can use the synonym to your view and i guess..it will appear in that list and you can use that.. ----------- hope this helps --------...

select data from physical table insert into temp table do some verification and calculation and insert into second temp table then select * from second temp table What you can do is like the following :- 1. Database stage to read form the physical table. 2.Here you dont have to use the temp table. ...