Error while proxying request - [http2: Transport received Server's graceful shutdown GOAWAY]

Bug Report

What did you do?

Observe Operator logs

What did you expect to see?

No error messages

What did you see instead? Under which circumstances?

{
   "level":"error",
   "ts":"2025-11-09T14:44:53Z",
   "logger":"proxy",
   "msg":"Error while proxying request",
   "error":"http2: Transport: cannot retry err [http2: Transport received Server's graceful shutdown GOAWAY] after Request.Body was written; define Request.GetBody to avoid this error",
   "stacktrace":"
	github.com/operator-framework/ansible-operator-plugins/internal/ansible/proxy.(*responder).Error
		ansible-operator-plugins/internal/ansible/proxy/kubectl.go:164
	net/http/httputil.(*ReverseProxy).ServeHTTP
		/opt/hostedtoolcache/go/1.24.3/x64/src/net/http/httputil/reverseproxy.go:486
	k8s.io/apimachinery/pkg/util/proxy.(*UpgradeAwareHandler).ServeHTTP
		/home/runner/go/pkg/mod/k8s.io/apimachinery@v0.33.2/pkg/util/proxy/upgradeaware.go:293
	github.com/operator-framework/ansible-operator-plugins/internal/ansible/proxy.newServer.stripLeaveSlash.func1
		ansible-operator-plugins/internal/ansible/proxy/kubectl.go:276
	net/http.HandlerFunc.ServeHTTP
		/opt/hostedtoolcache/go/1.24.3/x64/src/net/http/server.go:2294
	net/http.(*ServeMux).ServeHTTP
		/opt/hostedtoolcache/go/1.24.3/x64/src/net/http/server.go:2822
	github.com/operator-framework/ansible-operator-plugins/internal/ansible/proxy.Run.removeAuthorizationHeader.func3
		ansible-operator-plugins/internal/ansible/proxy/proxy.go:282
	net/http.HandlerFunc.ServeHTTP
		/opt/hostedtoolcache/go/1.24.3/x64/src/net/http/server.go:2294
	github.com/operator-framework/ansible-operator-plugins/internal/ansible/proxy.(*cacheResponseHandler).ServeHTTP
		ansible-operator-plugins/internal/ansible/proxy/cache_response.go:150
	net/http.serverHandler.ServeHTTP
		/opt/hostedtoolcache/go/1.24.3/x64/src/net/http/server.go:3301
	net/http.(*conn).serve
		/opt/hostedtoolcache/go/1.24.3/x64/src/net/http/server.go:2102"
}

Environment

Kubernetes cluster type:

OpenShift 4.18.25

$ operator-sdk version

ansible-operator version: "v1.41.0", commit: "8e923f25256c2b64c85a2c9dfabd157e8f880ea8", kubernetes version: "v1.32.0", go version: "go1.24.4", GOOS: "linux", GOARCH: "amd64"

$ go version (if language is Go)

1.24.4

$ kubectl version

1.29.7

Possible Solution

The client should possibly handle GOAWAY more gracefully.

Additional context

This appears as part of

• https://access.redhat.com/errata/RHBA-2025:15714
• https://issues.redhat.com/browse/OCPBUGS-61039

(Set goaway chance to 0.001)

Start by tracing the proxy error through internal/ansible/proxy/kubectl.go, especially responder.Error and newServer, then inspect the surrounding ReverseProxy and cache response handling shown in the stack trace. Reproduce the HTTP/2 GOAWAY scenario if possible and define completion as handling the graceful shutdown without emitting the observed proxy error.