Issue connecting to neo4j db through jupyter notebook

Drivers & Stacks Python
1

Hi all,

looking to get some guidance here, I have a jupyter notebook that i'm using to try to query the neo4j cluster directly. I get the following error when I attempt to run it.

Passing user/pw/uri as a variable below.

ResultConsumedError: The result is out of scope. The associated transaction has been closed. Results can only be used while the transaction is open.

SS 2023-07-18 at 09.52.59@2x
SS 2023-07-18 at 09.52.59@2x2022×148 12.9 KB

neo4j version: 4.4.23
os: Ventura 13.4.1 (22F82)
misc: dataspell version 2023.1.3, neo4j running on local machine.

Any help would be appreciated!

code below:

class NeoDB(object):
    def __init__(self):
        self._parse_config()
        self._connect()

    def _parse_config(self):
        self._neo4j_uri = os.environ['NEO4J_URI']
        self._neo4j_user = os.environ['NEO4J_USER']
        self._neo4j_password = os.environ['NEO4J_SECRETS_PASSWORD']

    def _connect(self):
        self._driver = GraphDatabase.driver(self._neo4j_uri,
                                            auth=(self._neo4j_user, self._neo4j_password))
        print('Neo4J Client instantiated: {}'.format(self._neo4j_uri))

    @staticmethod
2

Can you share the query and how you consume the results?
That is where the error is happening.

3

See below for query:

removed
4

Thanks for that information.
Do you know where that error is thrown? Or can you help us narrow that down?

5

David,

the error is thrown after I attempt to run this part of the code:

r = ReportBuilder()
r.report_aws_inventory()

I have a query.json file that has a preset built report that I can call, and it calls that report and attempts to run it based on tags assigned to the queries. I can provide here if required but heres a sample snippet of one of the queries:

  {
      "name": "aws_accounts",
      "tags": [
          "inventory",
          "security",
          "cloud",
          "aws"
      ],
      "description": "List of AWSAccounts currently analysed",
      "query": "MATCH (a:AWSAccount) RETURN ",
      "return": "a.id, a.name",
      "result_headers": [
          "account_id",
          "account_name"
      ]
  },
6

Hi :wave:
The problem lies within _exec_query. You execute a query within a transaction tx.run(...) which gives you back a Result. This Result is really just a cursor and doesn't contain the actual query results yet. That cursor becomes invalid the moment the transaction ends, which happens when returning from _exec_query.

To fix your code, you need to consume the result within the transaction function. There are plenty of options. Here is one suggestion:

    def _exec_query(tx, query, kwargs):
        if kwargs:
            result = tx.run(query, **kwargs)
        else:
            result = tx.run(query)
        return list(result)

Other options are result.to_eager_result, result.consume, result.data and more.

1 Like
7

Rouven,

Thank you, that resolved my issue for me, appreciate the assist! I'll sift through more of the documentation to see if I want to use any of the other suggested options.

1 Like