Fixing post-ingestion message error

Sending an ingestion-complete.vlba message to AAT when an observation ingestion was successful. The default img-ingestion-complete message required a requestId parameter which ingest_obs workflows don't have and they would've kicked off the wrong indexer.

More work will be done on the AAT side in the future which may warrant some changes here.

shared/workspaces/workspaces/workflow/message_architect.py

@@ -120,6 +120,16 @@ archive_msg_templates = DictView(
                 "status": "complete",
             },
         },
+        "obs_ingestion_complete": {
+            "application": "ingestor",
+            "message": "Observation Ingestion Complete",
+            "request": "observation",
+            "logData": {
+                "project_code": None,
+                "ingestion_type": "vlba",
+                "status": "complete",
+            },
+        },
         "download_complete": {
             "application": "workflow",
             "download_url": None,
@@ -281,7 +291,7 @@ class ArchiveMessageArchitect(MessageArchitectIF):
         # image indexing is based on project not request id, use different subject
         template["subject"] = (
             f"WS workflow triggering indexer for {self.project}"
-            if self.key == "ingestor.images"
+            if self.key in ["ingestor.images", "ingestion-complete.vlba"]
             else f"WS workflow request #{self.request.workflow_request_id}"
         )
         template["routing_key"] = self.key

[Daniel Nemergut]

Wasn't sure if the message and request values matter. The examples in the archive have them as 'message': 'fitsaips import' and 'request': 're-index please', so I thought maybe they don't matter.

[Jim Sheckard]

If they're in the other templates, it's probably best to follow suit. They were either necessary to have not-blank or set as a preventive measure.

[Jim Sheckard]

As we refine more of the AAT systems, the templates can be refactored to something a bit closer to WS.