First idea where to start the VM inside of actinia

src/actinia_core/rest/base/resource_base.py

@@ -352,6 +352,14 @@ def preprocess(
         ]
         self.job_timeout = int(process_time_limit * process_num_limit * 20)
 
+        # IDEA: maybe here we can also ask for the user the (
+        # self.user_credentials["vm_definition"]):
+        #  * VM size
+        #  * number of worker per VM
+        #  * mounts
+        #  * ...
+        # return directly an error message (create_response_from_model)
+
         # Create the resource URL base and use a placeholder for the file name
         # The placeholder __None__ must be replaced by the resource URL
         # generator

src/actinia_core/rest/ephemeral_processing.py

@@ -172,6 +172,16 @@ def post(self, location_name):
         rdc = self.preprocess(location_name=location_name)
 
         if rdc:
+            # IDEA: here VM start (asyncron - subprocess?)
+            # QUESTION:
+            # * set status to accepted or to startingVM?
+            # * consider from "queue_type": only if "per_job", "per_user" the VM should be started?
+            # prepare_actinia(vm_size, queue_name (created in enqueue_job))
+            #    * check if VM already run (if only one VM should be started per user)
+            #    * start VM
+            #    * install actinia
+            #    * start worker for queue_name (execute_actinia)
+            # vm_size, ... in preprocess (attributes of rdc?)
             enqueue_job(self.job_timeout, start_job, rdc)
 
         html_code, response_model = pickle.loads(self.response_data)