topic Re: Fatal error: The Python kernel is unresponsive. in Data Engineering

Fatal error: The Python kernel is unresponsive.

Data_Analytics1 — Wed, 08 Feb 2023 05:59:37 GMT

I am using MultiThread in this job which creates 8 parallel jobs. It fails for few times in a day and sometimes stuck in any of the Python notebook cell process. Here

The Python process exited with an unknown exit code.

The last 10 KB of the process's stderr and stdout can be found below. See driver logs for full logs.

---------------------------------------------------------------------------

Last messages on stderr:

Tue Feb 7 17:10:18 2023 Connection to spark from PID 24461

Tue Feb 7 17:10:18 2023 Initialized gateway on port 34499

Tue Feb 7 17:10:19 2023 Connected to spark.

---------------------------------------------------------------------------

Last messages on stdout:

NOTE: When using the `ipython kernel` entry point, Ctrl-C will not work.

To exit, you will have to explicitly quit this process, by either sending

"quit" from a client, or using Ctrl-\ in UNIX-like environments.

To read more about this, see https://github.com/ipython/ipython/issues/2049

Re: Fatal error: The Python kernel is unresponsive.

Lakshay — Wed, 08 Feb 2023 13:31:03 GMT

Hi @Mahesh Chahare , What is the DBR version you are using?

Re: Fatal error: The Python kernel is unresponsive.

Data_Analytics1 — Wed, 08 Feb 2023 13:33:22 GMT

Hi @Lakshay Goel , 11.3 LTS (includes Apache Spark 3.3.0, Scala 2.12)

Re: Fatal error: The Python kernel is unresponsive.

Lakshay — Wed, 08 Feb 2023 13:39:28 GMT

please try using this configuration:

spark.databricks.python.defaultPythonRepl pythonshell

Re: Fatal error: The Python kernel is unresponsive.

Data_Analytics1 — Wed, 08 Feb 2023 13:54:04 GMT

@Lakshay Goel , I have updated it on the cluster. I'll update you whether it works for me or not. Thanks !

Re: Fatal error: The Python kernel is unresponsive.

Lakshay — Wed, 08 Feb 2023 14:02:43 GMT

sure

Re: Fatal error: The Python kernel is unresponsive.

Manoj12421 — Thu, 09 Feb 2023 19:31:26 GMT

@Lakshay Goel

Am also facing same issues.

Re: Fatal error: The Python kernel is unresponsive.

Lakshay — Fri, 10 Feb 2023 10:52:04 GMT

Hi @Manoj Kumar , If you are facing this issue on DBR 11.x, you can enable the same configuration as above i.e. spark.databricks.python.defaultPythonRepl pythonshell

Re: Fatal error: The Python kernel is unresponsive.

308633 — Mon, 13 Feb 2023 06:22:10 GMT

Hello We are also facing the same issue, did check by enabling the spark.databricks.python.defaultPythonRepl pythonshell , the issue has not been solved yet

Re: Fatal error: The Python kernel is unresponsive.

Lakshay — Mon, 13 Feb 2023 12:58:44 GMT

Could you please try running the job on DBR 10.4?

Re: Fatal error: The Python kernel is unresponsive.

Manoj12421 — Fri, 17 Feb 2023 04:58:31 GMT

Yeah it's working thanks @Lakshay Goel

Re: Fatal error: The Python kernel is unresponsive.

Lakshay — Fri, 17 Feb 2023 11:10:01 GMT

Happy to help!

Re: Fatal error: The Python kernel is unresponsive.

youssefmrini — Fri, 24 Feb 2023 13:44:24 GMT

If you do face cluster issues. Make sure to first, restart the cluster if this one was running for a long time to get the patches, then if you can use the latest runtime. If the issue is still persistent open a ticket.

Re: Fatal error: The Python kernel is unresponsive.

Anonymous — Sat, 08 Apr 2023 07:33:20 GMT

Hi @Mahesh Chahare

Hope everything is going great.

Just wanted to check in if you were able to resolve your issue. If yes, would you be happy to mark an answer as best so that other members can find the solution more quickly? If not, please tell us so we can help you.

Cheers!

Re: Fatal error: The Python kernel is unresponsive.

rnby_mike — Mon, 24 Apr 2023 15:52:42 GMT

Hello,

I have the same problem. After upgrading from 10.5 to 12.2 we started to face this issue. Setting

spark.databricks.python.defaultPythonRepl pythonshell didn't help

Re: Fatal error: The Python kernel is unresponsive.

deedstoke — Tue, 25 Apr 2023 09:35:44 GMT

We are also facing the same issue.

Re: Fatal error: The Python kernel is unresponsive.

luis_herrera — Thu, 27 Apr 2023 20:21:45 GMT

Unresponsive Python notebooks or cancelled commands could be the cause of "Fatal error: The Python kernel is unresponsive." This can be caused by a number of problems, such as metastore connectivity issues or conflicting libraries. To troubleshoot this issue, you can check for metastore connectivity and review the Cluster cancels Python command execution due to library conflict KB article for more information:

https://kb.databricks.com/python-command-cancelled

Re: Fatal error: The Python kernel is unresponsive.

luis_herrera — Fri, 28 Apr 2023 15:48:37 GMT

Hey, it seems that the issue is related to the driver undergoing a memory bottleneck, which causes it to crash with an out of memory (OOM) condition and gets restarted or becomes unresponsive due to frequent full garbage collection. The reason for the memory bottleneck can be, among others that 1) The driver instance type is not optimal for the load executed on the driver, 2) there are memory-intensive operations executed on the driver, or many notebooks or jobs are running in parallel on the same cluster. The solution varies from case to case, but the easiest way to resolve the issue in the absence of specific details is to increase the driver's memory. Other points to consider include avoiding memory-intensive operations like collect() operator, which brings a large amount of data to the driver, conversion of a large DataFrame to Pandas, and running batch jobs on a shared interactive cluster. It is also recommended to distribute the workloads into different clusters. (https://kb.databricks.com/jobs/driver-unavailable)

For more information on troubleshooting unresponsive Python notebooks or canceled commands, please refer to the Troubleshooting unresponsive Python notebooks or canceled commands article in the Databricks documentation.

PS: Check #DAIS2023 talks as well