Repository navigation
Tolerate the AMD GPU and CPU pool taints when counting schedulable nodes - #2
Merged
Merged
Conversation
Each controller's Setup took a logger positionally and built its Reconciler's options itself, so configuring a Reconciler any other way meant adding another positional argument to Setup. Setup now takes the same ReconcilerOptions NewReconciler does, and passes them through. Signed-off-by: Nic Cope <nicc@rk0n.org>
The agent counted a node as schedulable only if every NoSchedule or NoExecute taint on it was nvidia.com/gpu with NoSchedule. Modelplane's engine pods also tolerate amd.com/gpu and modelplane.ai/cpu, so the agent would report AMD GPU and CPU pools as having no schedulable nodes. To avoid that Modelplane only asks the agent to observe NVIDIA pools, and falls back to whole-node scheduling on the others (modelplaneai/modelplane#601). The agent now tolerates the taints its --tolerated-taints flag names, which default to those three. A Modelplane that installs it can observe every pool. Signed-off-by: Nic Cope <nicc@rk0n.org>
Signed-off-by: Nic Cope <nicc@rk0n.org>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Part of modelplaneai/modelplane#601
The agent counted a node as schedulable only if every NoSchedule or NoExecute taint on it was
nvidia.com/gpu. Modelplane's engine pods also tolerateamd.com/gpuandmodelplane.ai/cpu, so it wouldn't count any node in an AMD GPU or CPU pool as schedulable.The agent now tolerates the taints named by a
--tolerated-taintsflag (the chart'stoleratedTaintsvalue), writtenkey[=value]:effect. A taint without a value matches any value, like theExiststolerations on Modelplane's pods. Both default to:The first commit makes each controller's
SetuptakeReconcilerOptions rather than a positional logger, somainpasses the configuredpool.ObserverwithWithPoolObserver.pool.NewObservertakes aWithTolerationsoption, and tolerates nothing without it.Once we release this, Modelplane needs to bump the agent and drop the NVIDIA-only filter in compose-inference-cluster.