Notebook cell gets hung up but code completes

[tim-mcwilliams]

Have been running into an issue when running a pymc-marketing model in a Databricks notebook. The cell that fits the model gets hung up and the progress bar stops moving, however the code completes and dumps all needed output into a folder. After the code completes I have to then detach the notebook since hitting Interrupt doesn't respond. I took a peek at the cluster logs and can confirm everything runs as expected (see screenshot!).

Any ideas the issue here or have you run into the same issue??

[Kaniz]

Hi @tim-mcwilliams,

It sounds like you’re encountering a situation where the notebook cell appears to hang while running a pymc-marketing model in Databricks, but the code was eventually completed successfully.

Let’s explore some potential reasons for this behaviour:

1. Resource Constraints:

   • Check if your Databricks cluster has sufficient resources (CPU, memory, and disk space) to handle the model-fitting process. If the cluster is under-provisioned, it might cause the progress bar to stall even though the code continues executing.
   • Consider increasing the cluster resources or using a larger instance type.

2. Concurrency and Parallelism:

   • Databricks Notebooks execute cells in parallel by default. If other cells are running concurrently, they might compete for resources and cause the progress bar to hang.
   • Try running the model in an isolated notebook or at a time when other cells are not executing.

3. Interrupt Signal Handling:

   • The fact that hitting “Interrupt” doesn’t respond suggests that the notebook might not be handling the interrupt signal properly.
   • Check if there are any custom signal handlers or other code that interfere with the default behaviour of interrupting a cell.
   • You can also try restarting the kernel or detaching the notebook as you’ve been doing.

4. Code Execution and Output:

   • Since the code completes successfully and dumps the output into a folder, it seems that the actual computation is working as expected.
   • Verify that the output files are correct and contain the expected results.

5. Databricks Environment and Dependencies:

   • Ensure that all necessary dependencies (including pymc-marketing) are correctly installed in your Databricks environment.
   • Check for any conflicting libraries or versions that might cause unexpected behavior.

6. Cluster Logs and Monitoring:

   • Continue monitoring the cluster logs to see if any specific errors or warnings occur during the execution of the model.
   • Look for any patterns or clues that might help identify the issue.

Hopefully, you’ll find a solution soon! 😊🚀.