What is the best way to implement resending instances?

General
1

Hello. We want to implement resending instances logic. We just send them on NEW_INSTANCE callback asynchronously. And if it is failed to send, just add instance id to file and resend each time on new study uploads:

def OnChange(changeType, level, resource):
    if changeType == orthanc.ChangeType.NEW_INSTANCE: 
        orthanc.RestApiPost('/peers/server/store', json.dumps({
           "Resources" : [instance_id],
           "Synchronous":False
        }))
            
     if changeType == orthanc.ChangeType.JOB_FAILURE:
	    print(resource) # this is failed job id

     if changeType == orthanc.ChangeType.STABLE_STUDY:
     	# logic to resend all previous failed instances

but JOB_FAILURE callback gives failed job id, not instance id. I can make request to get job details and grab instance id, but at that time job can be deleted according to JobsHistorySize.
For example If i set JobsHistorySize to 1000, and 5 instances failed and 1000 instances succeed after it, my first 5 failed jobs will be deleted and i can’t resume it or get instance id in job details.

May be any other better ways?

2

We use dedicated Orthanc instances for that with this kind of lua script.

The same kind of OrthancForwarder scripts also exists in python as part of this library.