feat(server): device-authorization for P2P gRPC (SA token + fabric-device check)#1
Draft
wseaton wants to merge 2 commits into
Draft
feat(server): device-authorization for P2P gRPC (SA token + fabric-device check)#1wseaton wants to merge 2 commits into
wseaton wants to merge 2 commits into
Conversation
…ice check Gates the P2pService route on a bound projected SA token (verified via TokenReview) and the caller pod requesting a fabric device. Modes: off / permissive / enforce. Helm + ClusterRBAC + Python client interceptor + docs. See the security RFC for design rationale. Assisted-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com> Signed-off-by: Will Eaton <weaton@redhat.com>
Signed-off-by: Will Eaton <weaton@redhat.com>
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Draft PR for visibility against my fork's main. Tracks the upstream RFC:
ai-dynamo#298 — [RFC] Coordinator gRPC authorization using device assertions.
Summary
P2pServicegRPC: verifies the caller's bound projected ServiceAccount token via KubernetesTokenReview, then confirms the caller's pod actually holds a fabric device (RDMA/IB, RoCE, EFA via device-plugin resources, or DRAResourceClaimby device class).off/permissive(verify+log, never block) /enforce(reject).p2pAuth.*plus cluster-scoped RBAC template; Python client attaches a projected token on each RPC and re-reads on rotation.ci/k8s/auth/cluster-e2e.sh) and a real-transfer GPU e2e (ci/k8s/auth/gpu-e2e.sh).Notes
p2pAuth.*) kept stable; user-facing prose/log lines renamed to "device authorization".Test plan
cargo build -p modelexpress-servercleancargo test -p modelexpress-server auth::passesci/k8s/auth/cluster-e2e.shagainst a real cluster: enforce path asserts no-token Unauthenticated, device-holding pod OK, device-less pod PermissionDeniedci/k8s/auth/gpu-e2e.shend-to-end: target logsRDMA transfer complete