resolver: never resend a refused query asking for recursion (closes #206)

queryDNS resent a query that a server refused, this time asking for
recursion, so on a network that intercepts DNS the answers could come
from a recursive resolver without anyone knowing. A refusal is now only
a refusal, and the server is passed over for the next.

When every server of a zone refuses, the error says so. When every root
server refuses, the error is ErrIntercepted: root servers refuse no
query, so something on the network is answering in their place.
FindAuthoritativeNameservers stops at that error instead of trying each
parent name, so the watcher's log line says it.

A live test asks Quad9, which refuses a query not asking for recursion,
so that the resend cannot come back unnoticed.

Model: opus-5-5

TODO.md

@@ -19,6 +19,8 @@ trial run of the finished image: https://git.eeqj.de/sneak/dnswatcher/issues/149
 
 # Completed Steps
 
+- 2026-10-02: a query a server refuses is not resent asking for recursion, and
+  every root server refusing is reported as DNS interception (closes #206).
 - 2026-10-02: the resolver tries root servers, and every other server list it
   walks, in a random order each time, not always from the top (closes #138).
 - 2026-10-02: a name listed more than once in `DNSWATCHER_TARGETS`, in any

internal/resolver/errors.go

@@ -22,6 +22,11 @@ var (
 		"reply is an error or a referral that leads no closer",
 	)
 
+	// ErrIntercepted is returned when every root server refused a
+	// query. Root servers refuse no query, so the refusals came from
+	// something on the network answering in their place.
+	ErrIntercepted = errors.New("this network intercepts DNS queries")
+
 	// ErrCNAMEDepthExceeded is returned when a CNAME chain
 	// exceeds MaxCNAMEDepth.
 	ErrCNAMEDepthExceeded = errors.New(

internal/resolver/export_test.go

@@ -28,6 +28,17 @@ func CollectIPs(
 	return collectIPs(results)
 }
 
+// QueryServers exports queryServers for testing.
+func (r *Resolver) QueryServers(
+	ctx context.Context,
+	servers []string,
+	zone string,
+	name string,
+	qtype uint16,
+) (*dns.Msg, error) {
+	return r.queryServers(ctx, servers, zone, name, qtype)
+}
+
 // QueryEachNS exports queryEachNS for testing.
 func (r *Resolver) QueryEachNS(
 	ctx context.Context,

internal/resolver/iterative.go

@@ -107,9 +107,8 @@ func (r *Resolver) retryTCP(
 	return resp
 }
 
-// queryDNS sends a DNS query to a specific server IP.
+// queryDNS sends a DNS query to a specific server IP, never asking it
-// Tries non-recursive first, falls back to recursive on
+// for recursion. A reply of REFUSED is returned as ErrRefused.
-// REFUSED (handles DNS interception environments).
 func (r *Resolver) queryDNS(
 	ctx context.Context,
 	serverIP string,
@@ -132,26 +131,13 @@ func (r *Resolver) queryDNS(
 		return nil, fmt.Errorf("query %s @%s: %w", name, serverIP, err)
 	}
 
-	if resp.Rcode == dns.RcodeRefused {
-		msg.RecursionDesired = true
-
-		resp, err = r.tryExchange(ctx, msg, addr)
-		if err != nil {
-			return nil, fmt.Errorf(
-				"query %s @%s: %w", name, serverIP, err,
-			)
-		}
-
 	if resp.Rcode == dns.RcodeRefused {
 		return nil, fmt.Errorf(
 			"query %s @%s: %w", name, serverIP, ErrRefused,
 		)
 	}
-	}
 
-	resp = r.retryTCP(ctx, msg, addr, resp)
+	return r.retryTCP(ctx, msg, addr, resp), nil
-
-	return resp, nil
 }
 
 func extractNSSet(rrs []dns.RR) []string {
@@ -279,7 +265,9 @@ func shuffled(
 
 // queryServers asks servers, the servers of zone, about name in a random
 // order until one gives a usable reply. A server that times out, refuses
-// or gives a reply that is not usable is passed over for the next.
+// or gives a reply that is not usable is passed over for the next. When
+// every server refused, the error says so, and when they are the root
+// servers it is ErrIntercepted.
 func (r *Resolver) queryServers(
 	ctx context.Context,
 	servers []string,
@@ -289,6 +277,8 @@ func (r *Resolver) queryServers(
 ) (*dns.Msg, error) {
 	var lastErr error
 
+	refused := 0
+
 	for _, ip := range shuffled(servers, rand.Shuffle) {
 		if checkCtx(ctx) != nil {
 			return nil, ErrContextCanceled
@@ -305,9 +295,27 @@ func (r *Resolver) queryServers(
 			return resp, nil
 		}
 
+		if errors.Is(err, ErrRefused) {
+			refused++
+		}
+
 		lastErr = err
 	}
 
+	if refused == len(servers) && zone == "." {
+		return nil, fmt.Errorf(
+			"every root server refused a query for %s: %w",
+			name, ErrIntercepted,
+		)
+	}
+
+	if refused == len(servers) {
+		return nil, fmt.Errorf(
+			"every server of %s refused a query for %s: %w",
+			zone, name, ErrRefused,
+		)
+	}
+
 	return nil, fmt.Errorf("all servers failed: %w", lastErr)
 }
 
@@ -528,6 +536,11 @@ func (r *Resolver) FindAuthoritativeNameservers(
 
 			return nsNames, nil
 		}
+
+		// The root servers would refuse every parent name too.
+		if errors.Is(err, ErrIntercepted) {
+			return nil, err
+		}
 	}
 
 	return nil, ErrNoNameservers

internal/resolver/resolver_test.go

@@ -11,6 +11,7 @@ import (
 	"testing"
 	"time"
 
+	"github.com/miekg/dns"
 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
 
@@ -295,6 +296,154 @@ func TestQueryNameserver_Refused(t *testing.T) {
 	assert.Equal(t, "server returned REFUSED", resp.Error)
 }
 
+// TestQueryNameserverIP_RecursiveResolverRefused asks Quad9, a public
+// recursive resolver, about google.com at both of its addresses. Quad9
+// refuses a query that does not ask for recursion and answers one that
+// does. The resolver never asks for recursion, so it must be reported
+// as refusing, never as answering.
+func TestQueryNameserverIP_RecursiveResolverRefused(t *testing.T) {
+	t.Parallel()
+
+	r := newTestResolver(t)
+
+	for _, ip := range []string{"9.9.9.9", "149.112.112.112"} {
+		var resp *resolver.NameserverResponse
+
+		livednstest.Retry(
+			t,
+			"QueryNameserverIP("+ip+", google.com)",
+			func(ctx context.Context) error {
+				var err error
+
+				resp, err = r.QueryNameserverIP(
+					ctx, ip, ip, "google.com",
+				)
+				if err != nil {
+					return err
+				}
+
+				// A timeout or a network error is no reply at all.
+				if resp.Status == resolver.StatusTimeout ||
+					strings.HasPrefix(resp.Error, "network error") {
+					return fmt.Errorf(
+						"%w: %s: %s",
+						livednstest.ErrNoAnswer, ip, resp.Error,
+					)
+				}
+
+				return nil
+			},
+		)
+
+		assert.Equal(t, resolver.StatusError, resp.Status, ip)
+		assert.Equal(t, "server returned REFUSED", resp.Error, ip)
+	}
+}
+
+// googleNameserverIPv4s returns the IPv4 addresses of google.com's
+// nameservers. The resolver asks servers only at their IPv4 addresses.
+func googleNameserverIPv4s(t *testing.T, r *resolver.Resolver) []string {
+	t.Helper()
+
+	var servers []string
+
+	for _, ns := range liveFindAuthoritative(t, r, "google.com") {
+		for _, ip := range liveResolveIPs(t, r, ns) {
+			if net.ParseIP(ip).To4() != nil {
+				servers = append(servers, ip)
+			}
+		}
+	}
+
+	return servers
+}
+
+// TestQueryServers_EveryServerRefused asks all of google.com's
+// nameservers about cloudflare.com, a zone they do not serve, which
+// they all refuse. The error says every server refused; it is not
+// ErrIntercepted, which only the root servers refusing shows.
+func TestQueryServers_EveryServerRefused(t *testing.T) {
+	t.Parallel()
+
+	r := newTestResolver(t)
+	servers := googleNameserverIPv4s(t, r)
+
+	var err error
+
+	livednstest.Retry(
+		t,
+		"QueryServers(google.com servers, cloudflare.com)",
+		func(ctx context.Context) error {
+			_, err = r.QueryServers(
+				ctx, servers, "google.com.", "cloudflare.com.",
+				dns.TypeNS,
+			)
+
+			// When not every server refused, one may have given no
+			// reply at all, so the attempt is tried again.
+			if err != nil &&
+				!strings.HasPrefix(err.Error(), "every server of") {
+				return fmt.Errorf(
+					"%w: %w", livednstest.ErrNoAnswer, err,
+				)
+			}
+
+			return nil
+		},
+	)
+
+	require.ErrorIs(t, err, resolver.ErrRefused)
+	require.NotErrorIs(t, err, resolver.ErrIntercepted)
+	require.EqualError(
+		t, err,
+		"every server of google.com. refused a query for "+
+			"cloudflare.com.: dns query refused",
+	)
+}
+
+// TestQueryServers_EveryRootServerRefused passes google.com's
+// nameservers to QueryServers as the servers of the root zone. They
+// refuse a query about cloudflare.com, as root servers would if
+// something on the network answered in their place, so the error is
+// ErrIntercepted.
+func TestQueryServers_EveryRootServerRefused(t *testing.T) {
+	t.Parallel()
+
+	r := newTestResolver(t)
+	servers := googleNameserverIPv4s(t, r)
+
+	var err error
+
+	livednstest.Retry(
+		t,
+		"QueryServers(google.com servers as root servers, cloudflare.com)",
+		func(ctx context.Context) error {
+			_, err = r.QueryServers(
+				ctx, servers, ".", "cloudflare.com.", dns.TypeNS,
+			)
+
+			// When not every server refused, one may have given no
+			// reply at all, so the attempt is tried again. Both errors
+			// for every server refusing say "refused a query for".
+			if err != nil &&
+				!strings.Contains(err.Error(), "refused a query for") {
+				return fmt.Errorf(
+					"%w: %w", livednstest.ErrNoAnswer, err,
+				)
+			}
+
+			return nil
+		},
+	)
+
+	require.ErrorIs(t, err, resolver.ErrIntercepted)
+	require.EqualError(
+		t, err,
+		"every root server refused a query for cloudflare.com.: "+
+			"this network intercepts DNS queries",
+	)
+}
+
 func TestQueryNameserver_RecordsSorted(t *testing.T) {
 	t.Parallel()