topic Re: Connect Timeout - Error when trying to run a cell in Data Engineering

Connect Timeout - Error when trying to run a cell

HelloDatabricks — Wed, 27 Mar 2024 16:35:06 GMT

Hello everybody.

Whenever I am trying to run a simple cell I receive the following error message now:
Notebook detached. Exception when creating expectation context:
java.net.SocketTimeoutException: Connect Timeout.

After that error message the cluster is disconnected from the notebook. When I reconnect (apparently successfully) and run a cell again the same error appears after some time of waiting. Also when I navigate to my profile the values of "Display Name" and "Groups" are infinitely loading and they are never shown.

Where is the problem? I have tried many things like using multiple different browsers and creating many new clusters with different runtime versions (instead of only focusing on 12.2LTS).

I appreciate every help. Thank you very much. It did work for some time.

Re: Connect Timeout - Error when trying to run a cell

MarijaS — Wed, 27 Mar 2024 16:44:29 GMT

I have the same problem this whole day. Maybe it is a databricks issue?

Re: Connect Timeout - Error when trying to run a cell

toolhater — Wed, 27 Mar 2024 18:07:01 GMT

Same thing. Even just a simple %fs ls times out. Tried new notebooks and computes. Any idea what is going on?

Re: Connect Timeout - Error when trying to run a cell

StanleyCruvinel — Wed, 27 Mar 2024 18:28:37 GMT

I have no ideia what's happening.
Exception when creating execution context:
java.net.SocketTimeoutException: Connect Timeout

Re: Connect Timeout - Error when trying to run a cell

Wolverine — Wed, 27 Mar 2024 20:28:51 GMT

I am also facing same issue from yesterday. not sure what's going on

Re: Connect Timeout - Error when trying to run a cell

MarijaS — Thu, 28 Mar 2024 14:11:23 GMT

today is ok

Re: Connect Timeout - Error when trying to run a cell

LokeshManne — Wed, 14 May 2025 09:16:44 GMT

@HelloDatabricks

The error you are facing is due to not detaching the old cluster, which is terminated/auto-terminated and when you run a cluster that connected to new cluster with same name and config. Which notebook tries to connect to old cluster which is still not deleted and just terminated with new cluster under same name, which makes the notebook run under work cluster which is terminated.

Solution: detach old cluster from notebook and reload the web_page, you can see only new cluster but not both at a time as below. Now attach the new cluster and run your notebook.

detach old cluster

reload the page, you can see only new cluster

Now attach the new cluster and run your notebook.