Merge d117486 into 88ef5fd

Author: wphyojpl

cumulus_lambda_functions/daac_archiver/daac_archiver_logic.py

@@ -43,20 +43,37 @@ def get_cnm_response_json_file(self, potential_file, granule_id):
                 return None
         if len(cnm_response_keys) > 1:
             LOGGER.warning(f'more than 1 cnm response file: {cnm_response_keys}')
-        cnm_response_keys = cnm_response_keys[0]
+        # assuming the names are the same, and it has processing date in the filename, it is easier to reverse it
+        cnm_response_keys = sorted(cnm_response_keys)[-1]  # sort and get the last one which is supposed to be the most recent one.
         LOGGER.debug(f'cnm_response_keys: {cnm_response_keys}')
         local_file = self.__s3.set_s3_url(f's3://{self.__s3.target_bucket}/{cnm_response_keys}').download('/tmp')
         cnm_response_json = FileUtils.read_json(local_file)
         FileUtils.remove_if_exists(local_file)
         return cnm_response_json
 
+    @staticmethod
+    def revert_to_s3_url(input_url):
+        if input_url.startswith("s3://"):
+            return input_url
+        if input_url.startswith("http://") or input_url.startswith("https://"):
+            parts = input_url.split('/', 3)
+            if len(parts) < 4:
+                ValueError(f'invalid url: {input_url}')
+            path_parts = parts[3].split('/', 1)
+            if len(path_parts) != 2:
+                ValueError(f'invalid url: {input_url}')
+            bucket, key = path_parts
+            return f"s3://{bucket}/{key}"
+        raise ValueError(f'unknown schema: {input_url}')
+
     def __extract_files(self, uds_cnm_json: dict, daac_config: dict):
         granule_files = uds_cnm_json['product']['files']
         if 'archiving_types' not in daac_config or len(daac_config['archiving_types']) < 1:
             return granule_files  # TODO remove missing md5?
         archiving_types = {k['data_type']: [] if 'file_extension' not in k else k['file_extension'] for k in daac_config['archiving_types']}
         result_files = []
         for each_file in granule_files:
+            LOGGER.debug(f'each_file: {each_file}')
             """
             {
                 "type": "data",
@@ -71,6 +88,7 @@ def __extract_files(self, uds_cnm_json: dict, daac_config: dict):
             if each_file['type'] not in archiving_types:
                 continue
             file_extensions = archiving_types[each_file['type']]
+            each_file['uri'] = self.revert_to_s3_url(each_file['uri'])
             if len(file_extensions) < 1:
                 result_files.append(each_file)  # TODO remove missing md5?
             temp_filename = each_file['name'].upper().strip()
@@ -79,28 +97,36 @@ def __extract_files(self, uds_cnm_json: dict, daac_config: dict):
         return result_files
 
     def send_to_daac_internal(self, uds_cnm_json: dict):
+        LOGGER.debug(f'uds_cnm_json: {uds_cnm_json}')
         granule_identifier = UdsCollections.decode_identifier(uds_cnm_json['identifier'])  # This is normally meant to be for collection. Since our granule ID also has collection id prefix. we can use this.
         self.__archive_index_logic.set_tenant_venue(granule_identifier.tenant, granule_identifier.venue)
         daac_config = self.__archive_index_logic.percolate_document(uds_cnm_json['identifier'])
         if daac_config is None or len(daac_config) < 1:
             LOGGER.debug(f'uds_cnm_json is not configured for archival. uds_cnm_json: {uds_cnm_json}')
             return
         daac_config = daac_config[0]  # TODO This is currently not supporting more than 1 daac.
+        result = JsonValidator(UdsArchiveConfigIndex.basic_schema).validate(daac_config)
+        if result is not None:
+            raise ValueError(f'daac_config does not have valid schema. Pls re-add the daac config: {result} for {daac_config}')
         try:
             self.__sns.set_topic_arn(daac_config['daac_sns_topic_arn'])
             daac_cnm_message = {
-                "collection": daac_config['daac_collection_name'],
+                "collection": {
+                    'name': daac_config['daac_collection_name'],
+                    'version': daac_config['daac_data_version'],
+                },
                 "identifier": uds_cnm_json['identifier'],
                 "submissionTime": f'{TimeUtils.get_current_time()}Z',
                 "provider": granule_identifier.tenant,
                 "version": "1.6.0",  # TODO this is hardcoded?
                 "product": {
                     "name": granule_identifier.id,
-                    "dataVersion": daac_config['daac_data_version'],
+                    # "dataVersion": daac_config['daac_data_version'],
                     'files': self.__extract_files(uds_cnm_json, daac_config),
                 }
             }
-            self.__sns.publish_message(json.dumps(daac_cnm_message))
+            LOGGER.debug(f'daac_cnm_message: {daac_cnm_message}')
+            self.__sns.set_external_role(daac_config['daac_role_arn'], daac_config['daac_role_session_name']).publish_message(json.dumps(daac_cnm_message), True)
             self.__granules_index.update_entry(granule_identifier.tenant, granule_identifier.venue, {
                 'archive_status': 'cnm_s_success',
                 'archive_error_message': '',

cumulus_lambda_functions/lib/uds_db/archive_index.py

@@ -15,11 +15,14 @@ class UdsArchiveConfigIndex:
     basic_schema = {
         'type': 'object',
         "additionalProperties": False,
-        'required': ['daac_collection_id', 'daac_sns_topic_arn', 'daac_data_version', 'collection', 'ss_username', 'archiving_types'],
+        'required': ['daac_collection_id', 'daac_sns_topic_arn', 'daac_data_version', 'daac_role_arn', 'daac_role_session_name',
+                     'collection', 'ss_username', 'archiving_types'],
         'properties': {
             'daac_collection_id': {'type': 'string'},
             'daac_sns_topic_arn': {'type': 'string'},
             'daac_data_version': {'type': 'string'},
+            'daac_role_arn': {'type': 'string'},
+            'daac_role_session_name': {'type': 'string'},
             'collection': {'type': 'string'},
             'ss_username': {'type': 'string'},
             'archiving_types': {'type': 'array', 'items': {'type': 'object'}},

cumulus_lambda_functions/uds_api/dapa/daac_archive_crud.py

@@ -19,13 +19,17 @@ class DaacUpdateModel(BaseModel):
     daac_collection_id: str
     daac_data_version: Optional[str] = None
     daac_sns_topic_arn: Optional[str] = None
+    daac_role_arn: Optional[str] = None
+    daac_role_session_name: Optional[str] = None
     archiving_types: Optional[list[ArchivingTypesModel]] = None
 
 
 class DaacAddModel(BaseModel):
     daac_collection_id: str
     daac_data_version: str
     daac_sns_topic_arn: str
+    daac_role_arn: str
+    daac_role_session_name: str
     archiving_types: Optional[list[ArchivingTypesModel]] = []