Shell Programming and Scripting

Hi,
I want to create a recon shell script to trigger 3rd process.
Pl give some suggestions (as simple as possible) how to handle this scenario as I do not have a recon table in place.
e.g should I create a list of procsees files create by 1st etl ) after successful job run and do a recon but how do I handle case when a trigger which came has not been procssed by 1st etl

Process Description :
1.source system sends trigger files whenever they load data via etl into thier target tables for each branch and business date (not system date).
File format : xxx_TRIGGER_<branch_cod>_<file_control_id>.txt
file_control_id is unique per trigger file.
2.Once we receive trigger files,we have a sheell script to check for trigger and create a parameter file so that 1st ETL will read it and prepare sql using it and load into table A.
3.2nd process will at midnight or next day morning (since it is triggered by an existing etl running at midnight).
This process has to read data from table A join with other tables and load into table B,C .
the process has to run some limited no of times so that it procsses one by one all the data loaded in Table A(one file control id at a time) uisng a parameter file.
4.last 3rd process has to perform recon to verify there are no more source trigger files to be procssed and all file controls ids have been procssed in tbl A,B,C.
This recon script will in turn create one extract date for all the file control ids processed and trigger 3rd ETL.

Process Description: Doing it this method all your data will be processed and if exception/error occurs you can alert opertions/support with message
describing what/where/when it occurred.

hth

Spacebar your description upto pt 2 is fine.Pl assist in below

Pt 3 requires a correction.
eixsitng etl process runs a etl which filters the source system from existing table X and loads into a temp table T the filtered txns.
This process creates a touch file indicating it is complete.

new process(script) 2nd will have to check for the above touch file and
also based on trigger file of pt 2 (which loade data in table A).
This new process should run as many no of times as there are trigger files created by pt 2 and load data in another table B & C based on join with temp table and table A.

Also table A is a txn table which cannot be marked as completed .

3rd process should run once all trigger files created by 1st and 2nd process are completed.No more files left for processing.

We also need to handle exception case of reprocessing where a trigger file can be resent with new file control id based on same data existing in table A.