-
Notifications
You must be signed in to change notification settings - Fork 152
-
Notifications
You must be signed in to change notification settings - Fork 152
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Dynamic DAG failed after changing parallelism for many times #1777
Comments
Not reproducible on my machine, postpone this issue. |
Can you reproduce this? It is possible because of the timeout setting of launching a JVM. Under certain circumstance, like very limited memory and very high CPU usage, it may takes a lot of time to start a new JVM. It may upto tens of seconds. |
Can you also attach the AppMaster log next time? So that we can compare the timestamp of both. |
the second log is AppMaster log |
@clockfly Yes, I can reproduce this in my environment. I agree with your guess because there are a bunch of dead executors still taking up memories after several rounds of DAG changes. |
To work-around this, I will change the JVM start timeout setting to a bigger value. |
fix #1777, Dynamic DAG failed after changing parallelism for many times
How to reproduce:
Submit SOL through Web UI and continuously changing parallelism SOLStreamProducer from 1 to 2, 3, 4, 5. After 4 or 5 times, SOL will fail itself.
executor log:
application master log:
The text was updated successfully, but these errors were encountered: