Skip to content
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

899 test for concurrent gpus #913

Merged
merged 7 commits into from
Oct 19, 2022
Merged

899 test for concurrent gpus #913

merged 7 commits into from
Oct 19, 2022

Conversation

binocarlos
Copy link
Contributor

This does not close #899 (because we still need to add 5 GPUs to the network) but it adds a test for concurrently scheduling GPU jobs to a network of varying latency and we now log when CalculateJobNodeDistanceDelay drops a job so we can catch when GPUs are not being scheduled to because of it

Additional features:

  • we now have a multi-node noop stack setup (we only ever had a single node noop stack)
  • we can introduce arbitrary network delays on the transports on any node
  • this actually gives us a fast (it's noop) stack for testing distributed networks

@binocarlos binocarlos merged commit 0043888 into main Oct 19, 2022
@philwinder philwinder deleted the 899-concurrent-gpus branch October 21, 2022 08:46
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

add 5 GPUs to production network
1 participant