You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Commit 34ad211
Browse filesBrowse the repository at this point in the historyBrowse files
Share the header check between the routed attempt and its retry, keep the
first-frame EOF rule in one bidi helper, reuse Get's request for the owner
lookup, and document the header contract next to the replica routing docs.
Part of #4223
Signed-off-by: divesh <dgude@nvidia.com>
The chart renders one Service per replica and a GRPCRoute rule that matches the `x-openshell-replica` request header. The CLI and the Rust SDK set that header from the owner the gateway reports and retry without it if the named replica is unavailable. Replica routing requires `workload.kind=statefulset`, `workload.allowMultiReplicaStatefulSet=true`, and at most 15 replicas.
260
+
The chart renders one Service per replica and a GRPCRoute rule that matches the `x-openshell-replica` request header. The CLI and the Rust and Go SDKs send `x-openshell-want-owner: 1` on `GetSandbox`, read the owner from the `x-openshell-owner` response header, and set it as `x-openshell-replica`. If the proxy answers `UNAVAILABLE` with an empty message, `no healthy upstream`, or `connection failure`, they retry once without it. Replica routing requires `workload.kind=statefulset`, `workload.allowMultiReplicaStatefulSet=true`, and at most 15 replicas.
261
261
262
262
If the release uses `workload.kind=deployment`, this upgrade replaces the Deployment with a StatefulSet, so every gateway pod restarts at once. The StatefulSet also requests a 1Gi PersistentVolumeClaim per replica, which stays `Pending` on clusters without a default StorageClass.
0 commit comments