Abhishek Malvankar

icon company Red Hat, IBM icon location Interested in distributed computing, resource management, and performance for AI workloads

Results 77 comments of Abhishek Malvankar

mcad-controller image has existing vulnerabilities

@jbusche do we think we can close this issue?

[metrics] Expose MCAD reasons for not dispatching an AppWrapper

FYI @dimakis

MCAD Quota Dashboard

@mamy-CS Can you please comment on the issue thread?

[CI] Automation of Multi-Cluster E2E Travis Testing

@dmatch01 we have moved to github actions. can we create an ADR for your proposal please, here is the ADR repo: https://github.com/project-codeflare/adr

[CI] Automation of Multi-Cluster E2E Travis Testing

@dmatch01 I think we should start with documentation PR on how to deploy MCAD in multi-cluster mode and provide sample AWs as a starting point.

[Feature request] Expose cluster available capacity for deployment sizing

FYI @dimakis @anishasthana another metrics requirement

[testing] Add test case to put jobs back into the queue after MCAD restart

FYI @anishasthana @dimakis

Kepler integration with MCAD

@metalcycling @eklee15 did we make progress on this issue?

[core] Gang termination in MCAD

With recent simplifications, low startup logic may have been removed. we need to verify and create a test case for it.

[quotamanagement] Handle case where tree has duplicate node names

@metalcycling Can you please comment here based on your experience with quota management?