仓库指标
- Star
- (27,997 star)
- PR 合并指标
- (平均合并 8天) (30 天内合并 378 个 PR)
描述
Title: The alt-port in Healthcheck needs to be implemented. I would like to confirm it from any leads who might be aware of this.
Description:
We have a manifest as below. We are using protos from the go-control plane. " default: interval: 1s timeout: 1s unhealthyThreshold: 1 healthyThreshold: 1 port: 8081 # alt port http: host: microsoft.com path: /dynamic/app match: statusCodes: - start: 200 end: 201 "
The service to which this is applied is exposing 2 ports - 8080 and 8081. The primary port is 8080. If the port is 8081 in the above config, the request is receiving 200s. But if the port is 8080(which is the serving port) we are receiving 503s. We are also receiving 503 when we set the port to some random value other than 8080 and 8081.
I would like to know if this field is even being supported or not as we are seeing varying behavior or if my understanding is wrong
Below is the proto of alt-port in health check in envoy
// [#not-implemented-hide:] Non-serving port for health checking. AltPort *wrappers.UInt32Value ``protobuf:"bytes,6,opt,name=alt_port,json=altPort,proto3" json:"alt_port,omitempty"
[optional Relevant Links:]