enter image description hereI'm a pentaho beginner i want to modify a column in my table (my table is an information transfer table, we empty it all the time) after each excel generation, knowing that i have several transformations and the update will be done only in the one transformation how i'm going to do this so that at each execution it will do the update only in the transformation that performs this update (since in the job all transformations will be executed in parallel)
I used the table input component and microsoft excel writer in my transformation, and in the job I used this transformation and the sql scripting component to make my request for the update, but don't forget I have other transformations in this job.
OK, if I'm understanding your problem correctly, you have N transformations T1.N followed by T2.N, that are executed in parallel, reading from one table but applying different filters to generate different excel files.
After ALL the parallel transformations are finished, you want to read/update the STATUS column in the source table (probably marking it as "row" processed or just take it to another table or whatever) and you want to do it once, that is why in your original screenshot you have the "SQL 2" script following the T2.1 transformation, but in reality you want it to be executed after all the T2.N transformations have finished, so the structure of the job in your original screenshot isn't valid. You can't put it where all the parallel branches converge, because it would be executed N times.
So instead of using one job, you use two jobs. In job J1 you put all the parallel executions:
You don't need a "Success step" to mark a job as completed, the job will be marked as finished when all the steps have finished correctly, I like to use it for clarity when running jobs sequentially, but for parallel jobs I prefer to omit it because it can be interpreted like the job will finished when any of the branches has finished.
Also, in your screenshot you haven't marked the entries to run in parallel (the "=" symbol and --- lines following the Start entry). To do that you right click in the Start entry and select the option to "Run next entries in parallel"
Then you create a second job, J2, that launches sequentially your first job and afterwards your SQL 2 script:
Now the SQL 2 script will be launched only after J1 has finished sucessfully.