Neo4j-Python connection problem: Neo4j on GCP

Hello, I 've been trying to deploy a Neo4J database on Google Cloud but I ‘ve been facing connection problems.

I am using the neo4j:5.15.0-enterprise Docker image and I am only interested in exposing the database port (no browser).

Locally everything works fine. On Google Cloud, the idea is to deploy it using Cloud Run (right now we don’t care about data persistence - we do mount to GCS buckets though). Cloud Run allows only one port to be exposed and this is served at :443, plus it automatically adds SSL certificates.

Therefore, I have a Neo4J database in Docker, deployed as a Cloud Run instance, exposing 7687 at 443.

I have also configured the Neo4j environment variables by following the official instructions.

The problem:

The deployment is successful and I am also able to use my Neo4J Desktop app to connect remotely to the database and interact with it, BUT I cannot connect to it using neo4j-python, py2neo or cypher-shell.

Here is the log output from neo4j-python:

[#0000] _: <POOL> created, direct address IPv4Address((‘<database-uri>’, 443))
[#0000] _: <POOL> acquire direct connection, access_mode='WRITE', database='neo4j'
[#0000] _: <POOL> trying to hand out new connection
[#0000] _: <RESOLVE> in: <database-uri>:443
[#0000] _: <RESOLVE> dns resolver out: <google-cloud-ip>:443
[#0000] C: <OPEN> <google-cloud-ip>:443
[#FB7A] C: <SECURE> <database-uri>
[#FB7A] C: <MAGIC> 0x6060B017
[#FB7A] C: <HANDSHAKE> 0x00040405 0x00020404 0x00000104 0x00000003
[#FB7A] S: <CONNECTION FAILED> ServiceUnavailable: Failed to read any data from server ResolvedIPv4Address((‘<google-cloud-ip>’, 443)) after connected (deadline Deadline(timeout=60.0))
[#0000] _: <RESOLVE> dns resolver out: <google-cloud-ip>:443
[#0000] C: <OPEN> <google-cloud-ip>:443
[#FB85] C: <SECURE> <database-uri>
[#FB85] C: <MAGIC> 0x6060B017
[#FB85] C: <HANDSHAKE> 0x00040405 0x00020404 0x00000104 0x00000003
[#FB85] S: <CONNECTION FAILED> ServiceUnavailable: Failed to read any data from server ResolvedIPv4Address((‘’, 443)) after connected (deadline Deadline(timeout=60.0))
[#0000] _: <RESOLVE> dns resolver out: :443
[#0000] C: <OPEN> <google-cloud-ip>:443
[#FB86] C: <SECURE> <database-uri>
[#FB86] C: <MAGIC> 0x6060B017
[#FB86] C: <HANDSHAKE> 0x00040405 0x00020404 0x00000104 0x00000003
[#FB86] S: <CONNECTION FAILED> ServiceUnavailable: Failed to read any data from server ResolvedIPv4Address((‘<google-cloud-ip>’, 443)) after connected (deadline Deadline(timeout=60.0))
[#0000] _: <RESOLVE> dns resolver out: <google-cloud-ip>:443
[#0000] C: <OPEN> <google-cloud-ip>:443
[#FB87] C: <SECURE> <database-uri>
[#FB87] C: <MAGIC> 0x6060B017
[#FB87] C: <HANDSHAKE> 0x00040405 0x00020404 0x00000104 0x00000003
[#FB87] S: <CONNECTION FAILED> ServiceUnavailable: Failed to read any data from server ResolvedIPv4Address((‘<google-cloud-ip>’, 443)) after connected (deadline Deadline(timeout=60.0))
[#0000] _: <POOL> close

Basically, the script times out and cannot connect. On Cloud Run I don’t see any message in the logs, like there is no attempt from a client to connect. I ‘ve checked everywhere online and have tried many different things but have not been able to come up with a solution.

I don’t see how is it possible to connect using Neo4j Desktop, but not neo4j-python or cypher-shell.
Does this mean that I should try to figure out the problem only on the client-side and the server is working fine?
If so, what is Neo4j Desktop doing internally that makes it able to connect and how to replicate it using neo4j-python?

Any help would be much appreciated!
Thanks