-
Notifications
You must be signed in to change notification settings - Fork 3.6k
Allow resource overcommitting when running functions in Kubernetes #4829
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
Conversation
rerun cpp tests |
1 similar comment
rerun cpp tests |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
overall looks good to me. Can you add validation to make sure people don't set zero or negative numbers for these settings?
@sijie currently we don't have any validation logic for any function worker configs. We definitely should add that validation logic. Preferably using the annotation based validation framework I introduced before. Lets create an issue for this to add validation framework for worker configs, since currently there is nothing of there. For this PR, lets not add a one off check just for that config. |
I don't think there is validation for any of the broker config as well :/ |
Please create the issue and link it here and add a TODO item with the link to the newly added fields. So we don't forget about it. |
…pache#4829) Currently, when running Pulsar Functions, Sources, and Sinks in Kubernetes. The resources requests and resource limits are set to the same values. While this is ok and everything will run as it should, actual resource utilization in the cluster might be low. To increase actual resource utilization, we need to be able to overcommit a certain amount in our clusters
Motivation
Currently, when running Pulsar Functions, Sources, and Sinks in Kubernetes. The resources requests and resource limits are set to the same values. While this is ok and everything will run as it should, actual resource utilization in the cluster might be low. To increase actual resource utilization, we need to be able to overcommit a certain amount in our clusters