02-03-2021 11:28 PM
Team,
We are seeing this message in the Pipeline Execute Snap “Detected a stalled snap thread: Pipeline Execute”. And the child pipeline did not process any document but shows that the execution time as 14 hours and still executing. The first snap in the child pipeline is Oracle Execute Snap where we truncate the table but it did not start or finish the execution, instead it shows the execution time as 887 mins.
Can you help us to understand what this message means we see in the runtime_logs message.
Thanks,
Dhivya Aroma.
11-15-2022 10:20 AM
Did you finally know why it was happening? I am looking at a similar error.
11-15-2022 10:42 AM
It’s a warning, not an error, and sometimes it can be ignored, such as when the snap is an event-polling snap like the Salesforce Subscriber or Kafka Consumer. Which snap are you seeing this warning for?
11-15-2022 11:07 AM
I see this for unions and pipeline executes
11-15-2022 12:48 PM
I did some research on this. This warning from the platform is shown for snap threads that have not consumed any CPU for a while, since in some cases those threads might actually be stalled, not making progress when they should. However, many snaps naturally wait as part of their design, and for these, the warning is a false positive. Some such snaps like the Salesforce Subscriber are annotated to suppress the platform warning, but many other snaps that should be annotated in the same way are not yet, so the warnings that you see for those snaps can be ignored.