tailcfg,ipn/ipnlocal: regulate netmap caching via a node attribute (#19117)

Add a new tailcfg.NodeCapability (NodeAttrCacheNetworkMaps) to control whether a node with support for caching network maps will attempt to do so. Update the capability version to reflect this change (mainly as a safety measure, as the control plane does not currently need to know about it). Use the presence (or absence) of the node attribute to decide whether to create and update a netmap cache for each profile. If caching is disabled, discard the cached data; this allows us to use the presence of a cached netmap as an indicator it should be used (unless explicitly overridden). Add a test that verifies the attribute is respected. Reverse the sense of the environment knob to be true by default, with an override to disable caching at the client regardless what the node attribute says. Move the creation/update of the netmap cache (when enabled) until after successfully applying the network map, to reduce the possibility that we will cache (and thus reuse after a restart) a network map that fails to correctly configure the client. Updates #12639 Change-Id: I1df4dd791fdb485c6472a9f741037db6ed20c47e Signed-off-by: M. J. Fromberger <fromberger@tailscale.com>
2026-04-01 15:02:53 -07:00
parent c76113ac75
commit 211ef67222
6 changed files with 189 additions and 9 deletions
@@ -710,6 +710,118 @@ func TestLoadCachedNetMap(t *testing.T) {
 	}
 }

+func TestUpdateNetMapCache(t *testing.T) {
+	t.Setenv("TS_USE_CACHED_NETMAP", "1")
+
+	// Set up a cache directory so we can check what happens to it, in response
+	// to netmap updates.
+	varRoot := t.TempDir()
+	cacheDir := filepath.Join(varRoot, "profile-data", "id0", "netmap-cache")
+
+	testMap := &netmap.NetworkMap{
+		SelfNode: (&tailcfg.Node{
+			Name: "example.ts.net",
+			User: tailcfg.UserID(1),
+			Addresses: []netip.Prefix{
+				netip.MustParsePrefix("100.2.3.4/32"),
+			},
+		}).View(),
+		UserProfiles: map[tailcfg.UserID]tailcfg.UserProfileView{
+			tailcfg.UserID(1): (&tailcfg.UserProfile{
+				ID:          1,
+				LoginName:   "amelie@example.com",
+				DisplayName: "Amelie du Pangoline",
+			}).View(),
+		},
+		Peers: []tailcfg.NodeView{
+			(&tailcfg.Node{
+				ID:           601,
+				StableID:     "n601FAKE",
+				ComputedName: "some-peer",
+				User:         tailcfg.UserID(1),
+				Key:          makeNodeKeyFromID(601),
+				Addresses: []netip.Prefix{
+					netip.MustParsePrefix("100.2.3.5/32"),
+				},
+			}).View(),
+		},
+	}
+
+	// Make a new backend to which we can send network maps to test that
+	// netmap caching decisions are made appropriately.
+	sys := tsd.NewSystem()
+	e, err := wgengine.NewFakeUserspaceEngine(logger.Discard,
+		sys.Set,
+		sys.HealthTracker.Get(),
+		sys.UserMetricsRegistry(),
+		sys.Bus.Get(),
+	)
+	if err != nil {
+		t.Fatalf("Make userspace engine: %v", err)
+	}
+	t.Cleanup(e.Close)
+	sys.Set(e)
+	sys.Set(new(mem.Store))
+
+	logf := tstest.WhileTestRunningLogger(t)
+	clb, err := NewLocalBackend(logf, logid.PublicID{}, sys, 0)
+	if err != nil {
+		t.Fatalf("Make local backend: %v", err)
+	}
+	t.Cleanup(clb.Shutdown)
+	clb.SetVarRoot(varRoot)
+
+	pm := must.Get(newProfileManager(new(mem.Store), logf, health.NewTracker(sys.Bus.Get())))
+	pm.currentProfile = (&ipn.LoginProfile{ID: "id0"}).View()
+	clb.pm = pm
+	if err := clb.Start(ipn.Options{}); err != nil {
+		t.Fatalf("Start local backend: %v", err)
+	}
+
+	wantCacheEmpty := func() {
+		// The cache directory should be empty, as caching is not enabled.
+		if des, err := os.ReadDir(cacheDir); err != nil {
+			t.Errorf("List cache directory: %v", err)
+		} else if len(des) != 0 {
+			t.Errorf("Cache directory has %d items, want 0\n%+v", len(des), des)
+		}
+	}
+
+	// Send the initial network map to the backend. Because the map does not
+	// include the cache attribute, no cache should be written.
+	clb.mu.Lock()
+	clb.setNetMapLocked(testMap)
+	clb.mu.Unlock()
+
+	wantCacheEmpty()
+
+	// Now enable the netmap caching attribute, and send another update.
+	// After doing so, the cache should have real data in it.
+	testMap.AllCaps = set.Of(tailcfg.NodeAttrCacheNetworkMaps)
+
+	clb.mu.Lock()
+	clb.setNetMapLocked(testMap)
+	clb.mu.Unlock()
+
+	if des, err := os.ReadDir(cacheDir); err != nil {
+		t.Errorf("List cache directory: %v", err)
+	} else if len(des) == 0 {
+		t.Error("Cache is unexpectedly empty")
+	} else {
+		t.Logf("Cache directory has %d entries (OK)", len(des))
+	}
+
+	// Now disable the node attribute again, send another update, and verify
+	// that the cache got cleaned up.
+	testMap.AllCaps = nil
+
+	clb.mu.Lock()
+	clb.setNetMapLocked(testMap)
+	clb.mu.Unlock()
+
+	wantCacheEmpty()
+}
+
 func TestConfigureExitNode(t *testing.T) {
 	controlURL := "https://localhost:1/"
 	exitNode1 := makeExitNode(1, withName("node-1"), withDERP(1), withAddresses(netip.MustParsePrefix("100.64.1.1/32")))