Bring your towel and join one of the fastest growing data communities. Welcome to our second-generation open source orchestration platform, a completely rethought approach to dataflow automation.

Prefect Community

Hello everyone! I built a simple ETL and it is working perfectly. When I run the agent in the cloud I get an error saying "cannot pickle psycopg2", any clues?

Hi <@U01LQSQV3ST> - it sounds like you've baked in part of the `psycopg2` module as part of your flow's metadata, perhaps by referencing it outside of a task block. Since a flow's metadata (right now) needs to be pickleable so that it can be sent to the API, you'll need to make sure you're not inadvertently introducing non-pickleable objects.

One question though, the thing is our pipeline sues a OOP design using a lof of composition and sharing attributes among classes. how would you go about this issue in the cloud since i have my db object constantly querying the db all along the pipeline.

Ah, good question -&gt; I personally use OOP by default when writing flows for similar reasons; for instance you can write a task following normal class inheritance models like this:
```class CheckInstance(Task):
    def list_instances(self, service, project, zone):
        result = service.instances().list(project=project, zone=zone).execute()
        return result["items"] if "items" in result else None

    def run(self, credentials, instance, project, zone):
        service = googleapiclient.discovery.build(
            "compute", "v1", credentials=credentials
        )

        instances = self.list_instances(service, project, zone)

        super(CheckInstance, self).run()
        return any(i["name"] == instance for i in instances)```

In a case like that, there's nothing preventing you from creating instances of your object that also inherit the `Task` class, and calling `super(Instance, self).run()` to call the parent class run method

and thereby letting you share methods and attributes

One trick that solved everything was to avoid declaring a sef.conn and self.cursor, for the database class, so i am pretty much creating and destroying the connection every time I query th DB and that solved the issue