Can't connect any more
under review
Nissanka Fernando
Rene Rosendahl adam@labfoo.dev what port are you connecting to? It's not shown on the error screen shot, we would like to recreate in dev. Thanks
adam@labfoo.dev
Nissanka Fernando We are running
2025.07
community edition and connecting on the protocol bolt://
on port 7687
.I am able to connect with the CLI tool libneo4j-omni with the following:
neo4j-client -u neo4j --insecure bolt://<REDACTED_HOSTNAME>:7687
I suspect the issue started to happen with the release of
neo4j-javascript
v6 being used in Neo4j Desktop.I ran the application through a local proxy and can see binary communication on the websocket. I've redacted some of this since it contains my credentials but you can see that the user agent and credentials are passed and then almost immedaitely we get a closed signal. This could just be part of the exchange.
I can confirm this issue has only started happening recently with no infra changes on our side or issues connecting from our application or CLI tools (only Neo4j Desktop for both me and Rene). I can't pin it to a specific version as I not aware of what version I was running before 2.1.0.
Nissanka Fernando
I'm escalating now, would you say this only started happening in v2.1.0 (from 3 Dec) ?
Nissanka Fernando
marked this post as
under review
Waiting for engineering to review
Nissanka Fernando
Hi Rene, please could you share your Neo4j Desktop main log files from the Developer menu and any general info about the remote instance you are trying to connect to. For example, is it SSO configured, is an AuraDB instance or self-managed etc. Thanks
Rene Rosendahl
Nissanka Fernando
Rene Rosendahl
Nissanka Fernando this is a self-managed instance (no SSO) in AWS.
Rene Rosendahl
Nissanka Fernando Another colleague of mine just started having the exact same issue, so it's probably not something related to my machine or networking.
Any update?
adam@labfoo.dev
Nissanka Fernando I am also getting this same issue from Neo4j Desktop (Version 2.1.0).
We've not had this issue in the past I am able to connect to the server through the CLI tool libneo4j-omni without any issues. The error I am getting is exactly the same as Rene Rosendahl reports.
Code: 50N42
error: general processing exception - unexpected error. Connection acquisition timed out in 60000 ms. Pool status: Active conn count = 0, Idle conn count = 0.
I believe this issue has been introduced by an update. One other thing to note is that we run OpenVPN to allow us to connect to our instance and thus rely on Neo4j desktop to use the system proxy for connections.
Please let me know if I can be any more help in diagnosing.
Rene Rosendahl
Nissanka Fernando Just saw 2.1.1 is available, but it's still happening with the latest version. Any update on this issue?
Nissanka Fernando
Rene Rosendahl adam@labfoo.dev
Response today from one of our driver engineers below. It's unlikely we'll be able to resolve or provide support until the new year but there are some things to try below:
"
I started up a 2025.07 community database using docker and connected with the latest driver with no issue.
If I could see the driver logs there maybe more clarity to be gleaned.
The error message is essentially the driver saying "there's no server there to connect to", which, given the setup, makes me suspect there's something going on with the proxy here.
It'd be interesting to see if they could connect with a plain driver with debug logging turned on.
If the normal driver can connect without issue then they could follow the guidelines in the driver README to import the driver in a browser and see if the same query succeeds there.
"