Technically you can for example use a PowerShell script or a batch file which regularly pings the Informatica domain (infacmd.bat ping -dn DomainName, documented in the Command Reference Guide, part of the Informatica Platform documentation and of the PowerCenter documentation).
If the output indicates that the ping succeeded, the INFA domain is still up and running.
If the output indicates that the domain cannot be reached, then you could restart the Windows service Informatica 10.* (depending on the exact version).
The big question is: does this make sense?
If the Informatica domain goes down, there is a reason why it collapsed. And this reason should first be investigated before trying to restart the domain.
So you should examine the logs (in particular the domain log) for any message indicating why the domain went down; and only if that reason was intermittent and can be resolved easily, restarting the domain makes sense.
In other environments we have set up similar scripts such that they try two times to restart the domain and, if both attempts fail, issue some administrator ticket and give up. Just as a side note how other customers deal with such requirements.
Thank you for your Reply, Nico!
I made Powershell Script for waking up Windows Service forcefully if the service is stopped.
Of Course, Your concern is very important.
In fact, this situation happened when PowerCenter did not connect to DataBase by 1 minnutet Network Interruption.
After that Trouble, DB worked normally, but PowerCenter Service is sttoped...
So, I want to retry PowerCenter Service Restart once time.
Sounds reasonable to me. Your approach (and its reasoning) is logical and clear. This is a typical situation for such "simple" surveillance scripts.