HI Naresh
Instead of creating a substitution parameter and configuring in the system configuration level, you can directly create two different data store, which is used to extract the data from the same source and place it in the two different targets as HA1 and HA2 when the job is executed only once.
Here since the job has to be triggered only once, I hope this is the best practise to do so!!!
Best,
Sanjay