Shell Programming and Scripting

Hello Gurus ,

Could you help me with my below requirement

In my script I will start a process ,it takes some time to complete the process.After completion it will create a file
after that I have to restart the process again
The number of time it will be done will vary but that count we should know before

so here is the algorithm

for in 1...count

delete indicator file

step 1 start process
step 2 check indicator file exist [ denotes process completed ]
no then sleep for 30 second
go to step 2
else
go to step 1 [run the process again ]

do this until the for loop completed

Could anyone please give me the code or suggestion how to procees
Your help is very much appreciated

Is this a homework assignment? If not, please explain what is going on here.

No Don its not

We have to run informatica (an ETL tool) workflow from the unix script
Now the unix command will start the workflow but will never know whether its finished that process or not .

It may took 5 mins to 1 hours depending on the load to complete the workflow .

Now we have to run the workflow depending on the number of trigger files in a directory so we have to start the informatica workflow in a loop.

Thats why we have the above approach .
take the count of the number of trigger files
start the workflow (workflow will create a indicator file after its successful completion)
check about the indicator file if its there means the workflow finished then again start it after deleting the workflow

Else wait to finish the workflow

Does it make sense?

Why count the trigger files? Can't you use them directly?

Where are they stored? What are they named?

After well over 100 posts and having received our help in solving more than 50 issues you have submitted, please show us what you have tried so far.

Have you considered changing the UNIX command that starts the workflow so it does know when the workflow completes and whether or not it completes successfully?