Enforce upstream attempt and overall query deadlines
build / build-push (push) Successful in 35s

The initial upstream exchange ignored its supplied timeout and used hardcoded client limits. Bound each attempt with a context, copy the shared client for its timeout, and share the attempt budget with UDP-to-TCP fallback. Close private connections on cancellation to interrupt reads, and exclude caller cancellation from upstream failures so parallel losers remain healthy.

The normal engine pipeline never applied server.timeouts.query, allowing failover to outlive the configured query budget. Apply a hot-reloaded overall context deadline at Handle while retaining earlier caller deadlines. Add silent UDP/TCP/TLS, cancellation, fallback, failover, and reload regressions.

The server inflight test assumed receiving a TCP reply meant its sent counter had already been incremented. Wait for both completion counters before asserting them, preserving the existing assertions while removing the scheduling race.
This commit is contained in:
mats committed 2026-10-03 19:06:57 +02:00
1 parent 32b214fac0
commit 340b4eaf80
6 files changed
+233 -10

No files matched your search

+106
View File
@@ -0,0 +1,106 @@
package resolver
import (
"context"
"errors"
"net"
"testing"
"time"
"git.k8s-git-k6n.net/mats/fancy-dns/internal/config"
)
func TestAttemptTimeoutSilentUDP(t *testing.T) {
pc, err := net.ListenPacket("udp", "127.0.0.1:0")
if err != nil {
t.Fatal(err)
}
defer pc.Close()
p := NewPool(poolCfg(udpServer(pc.LocalAddr().String())))
start := time.Now()
_, _, err = p.Exchange(context.Background(), query("silent.test."), nil, 50*time.Millisecond)
if err == nil || time.Since(start) > 500*time.Millisecond {
t.Fatalf("elapsed %v, error %v", time.Since(start), err)
}
}
func TestCallerCancellationInterruptsRead(t *testing.T) {
pc, err := net.ListenPacket("udp", "127.0.0.1:0")
if err != nil {
t.Fatal(err)
}
defer pc.Close()
p := NewPool(poolCfg(udpServer(pc.LocalAddr().String())))
ctx, cancel := context.WithCancel(context.Background())
defer cancel()
result := make(chan error, 1)
go func() { _, _, err := p.Exchange(ctx, query("cancel.test."), nil, 5*time.Second); result <- err }()
_ = pc.SetReadDeadline(time.Now().Add(time.Second))
if _, _, err := pc.ReadFrom(make([]byte, 4096)); err != nil {
t.Fatal(err)
}
cancel()
select {
case err := <-result:
if p.All()[0].Stats().Failures != 0 {
t.Fatal("caller cancellation counted as an upstream failure")
}
if !errors.Is(err, context.Canceled) {
t.Fatal(err)
}
case <-time.After(500 * time.Millisecond):
t.Fatal("cancellation did not interrupt read")
}
}
func TestFallbackSharesAttemptDeadline(t *testing.T) {
// Each leg takes 150ms: a fresh budget for TCP would incorrectly succeed.
up := newFakeUpstream(t, fakeOpt{delay: 150 * time.Millisecond, truncateUDP: true})
p := NewPool(poolCfg(udpServer(up.addr)))
start := time.Now()
_, _, err := p.Exchange(context.Background(), query("fallback.test."), nil, 250*time.Millisecond)
if err == nil || time.Since(start) > 450*time.Millisecond || up.tcpHits.Load() != 1 {
t.Fatalf("elapsed %v, TCP hits %d, error %v", time.Since(start), up.tcpHits.Load(), err)
}
}
func TestFailoverAfterAttemptTimeout(t *testing.T) {
pc, err := net.ListenPacket("udp", "127.0.0.1:0")
if err != nil {
t.Fatal(err)
}
defer pc.Close()
up := newFakeUpstream(t, fakeOpt{})
cfg := poolCfg(udpServer(pc.LocalAddr().String()), udpServer(up.addr))
cfg.Policy = "failover"
p := NewPool(cfg)
start := time.Now()
_, addr, err := p.Exchange(context.Background(), query("failover.test."), nil, 50*time.Millisecond)
if err != nil || addr != up.addr || time.Since(start) > 500*time.Millisecond {
t.Fatalf("elapsed %v, addr %s, error %v", time.Since(start), addr, err)
}
}
func TestAttemptTimeoutSilentStream(t *testing.T) {
for _, proto := range []string{"tcp", "tls"} {
t.Run(proto, func(t *testing.T) {
lc, err := net.Listen("tcp", "127.0.0.1:0")
if err != nil {
t.Fatal(err)
}
defer lc.Close()
accepted := make(chan net.Conn, 1)
go func() { conn, _ := lc.Accept(); accepted <- conn }()
p := NewPool(poolCfg(config.UpstreamServer{Addr: lc.Addr().String(), Proto: proto, InsecureSkipVerify: true}))
start := time.Now()
_, _, err = p.Exchange(context.Background(), query("silent.test."), nil, 50*time.Millisecond)
if err == nil || time.Since(start) > 500*time.Millisecond {
t.Errorf("elapsed %v, error %v", time.Since(start), err)
}
_ = lc.Close()
if conn := <-accepted; conn != nil {
_ = conn.Close()
}
})
}
}