Host Restarts in Azure Function Triggered by EventHub, Causing Data Loss

Hi Ryan,

thank you for the response. Yes, I have tried the options that are relevant. To be more specific:

• A regression from third-party packages - the only libraries in my requirements.txt are pyodbc and azure-functions , which are not pinned to a specific version, so the latest version should be installed. pyodbc has not had a newer version come out in the time since February, so it cannot be the issue.
• Unpickling from a malformed .pkl file - this is not relevant to my case since there is no unpickling or reading of any files
• Pyodbc connection collision - this is the only one that could be relevant in my opinion, and I did try this suggestion to implement singleton pattern, but it still crashes during increased traffic of incoming events. It never happened during single, isolated events. This is how it is implemented:

    class DBService:
        def __init__(self, db_host=None, db_name=None, db_user=None, db_password=None):
            self.db_host = f"tcp:{db_host}" if db_host else None
            self.db_name = db_name
            self.db_user = db_user
            self.db_password = db_password
            self.connection = None
       
        def connect(self, driver_version=17):
            if self.connection is None:
                connection_str = (f"DRIVER={{ODBC Driver {driver_version} for SQL Server}};"
                                  f"SERVER={self.db_host};"
                                  f"DATABASE={self.db_name};"
                                  f"UID={self.db_user};"
                                  f"PWD={self.db_password}")
                self.connection = pyodbc.connect(connection_str,autocommit=True)
     
        def close(self):
            if self.connection is not None:
                self.connection.close()
                self.connection = None
     
        def execute_query(self, query_str, params=None):
            self.connect()
            try:
    			with self.connection.cursor() as cursor:
    	            if params is not None:
    	                self.cursor.execute(query_str, params)	         
    	            else:
    	                self.cursor.execute(query_str)  
    				
    				while cursor.nextset():
    					logger.info(cursor.messages)	             
            except pyodbc.Error as e:
                logger.error(f"Pyodbc error {type(e).__name__} occured: {str(e)}.")
                self.connection.rollback()
                raise
            finally:
                self.close()
      
    class SingletonDBService(DBService):
        _instance = None
     
        def __new__(cls, *args, **kwargs):
            if not cls._instance:
                cls._instance = super().__new__(cls)
            return cls._instance
  

In the function app code, I will receive events, process them and construct SQL query and parameters, instantiate SingletonDBService, and then run execute_query() method. Sometimes, it will fail after I get the messages from the database that the query has run successfully (from the line logger.info(cursor.messages)), in which case there would be no data loss, however the problem is the other case when it happens before.