Merge 98a65c76d3 into a7874af3d0

Replace channel links with links to discord invite link and remove
channel list.

Fixes: #1521

.github/workflows/test-integration.yaml

@@ -60,6 +60,7 @@ jobs:
           - TestEnableDisableAutoApprovedRoute
           - TestAutoApprovedSubRoute2068
           - TestSubnetRouteACL
+          - TestHASubnetRouterFailoverWhenNodeDisconnects2129
           - TestHeadscale
           - TestCreateTailscale
           - TestTailscaleNodesJoiningHeadcale

CODE_OF_CONDUCT.md

@@ -62,7 +62,7 @@ event.
 
 Instances of abusive, harassing, or otherwise unacceptable behavior
 may be reported to the community leaders responsible for enforcement
-at our Discord channel. All complaints
+on our [Discord server](https://discord.gg/c84AZQhmpx). All complaints
 will be reviewed and investigated promptly and fairly.
 
 All community leaders are obligated to respect the privacy and

README.md

@@ -4,7 +4,7 @@
 
 An open source, self-hosted implementation of the Tailscale control server.
 
-Join our [Discord](https://discord.gg/c84AZQhmpx) server for a chat.
+Join our [Discord server](https://discord.gg/c84AZQhmpx) for a chat.
 
 **Note:** Always select the same GitHub tag as the released version you use
 to ensure you have the correct example configuration and documentation.

docs/about/faq.md

@@ -41,13 +41,15 @@ In addition to that, you may use packages provided by the community or from dist
 [installation guide using community packages](../setup/install/community.md).
 
 For convenience, we also [build Docker images with headscale](../setup/install/container.md). But **please be aware that
-we don't officially support deploying headscale using Docker**. We have a [Discord
+we don't officially support deploying headscale using Docker**. On our [Discord server](https://discord.gg/c84AZQhmpx)
-channel](https://discord.com/channels/896711691637780480/1070619770942148618) where you can ask for Docker-specific help
+we have a "docker-issues" channel where you can ask for Docker-specific help to the community.
-to the community.
 
 ## Why is my reverse proxy not working with headscale?
 
-We don't know. We don't use reverse proxies with headscale ourselves, so we don't have any experience with them. We have [community documentation](../ref/integration/reverse-proxy.md) on how to configure various reverse proxies, and a dedicated [Discord channel](https://discord.com/channels/896711691637780480/1070619818346164324) where you can ask for help to the community.
+We don't know. We don't use reverse proxies with headscale ourselves, so we don't have any experience with them. We have
+[community documentation](../ref/integration/reverse-proxy.md) on how to configure various reverse proxies, and a
+dedicated "reverse-proxy-issues" channel on our [Discord server](https://discord.gg/c84AZQhmpx) where you can ask for
+help to the community.
 
 ## Can I use headscale and tailscale on the same machine?
 

docs/about/help.md

@@ -1,11 +1,5 @@
 # Getting help
 
-Join our Discord server for announcements and community support:
+Join our [Discord server](https://discord.gg/c84AZQhmpx) for announcements and community support.
-
-- [announcements](https://discord.com/channels/896711691637780480/896711692120129538)
-- [general](https://discord.com/channels/896711691637780480/896711692120129540)
-- [docker-issues](https://discord.com/channels/896711691637780480/1070619770942148618)
-- [reverse-proxy-issues](https://discord.com/channels/896711691637780480/1070619818346164324)
-- [web-interfaces](https://discord.com/channels/896711691637780480/1105842846386356294)
 
 Please report bugs via [GitHub issues](https://github.com/juanfont/headscale/issues)

docs/about/releases.md

@@ -6,5 +6,4 @@ code archives. Container images are available on [Docker Hub](https://hub.docker
 
 An Atom/RSS feed of headscale releases is available [here](https://github.com/juanfont/headscale/releases.atom).
 
-Join the ["announcements" channel on Discord](https://discord.com/channels/896711691637780480/896711692120129538) for
+See the "announcements" channel on our [Discord server](https://discord.gg/c84AZQhmpx) for news about headscale.
-news about headscale.

docs/index.md

@@ -10,7 +10,7 @@ Headscale is an open source, self-hosted implementation of the Tailscale control
 
 This page contains the documentation for the latest version of headscale. Please also check our [FAQ](./about/faq.md).
 
-Join our [Discord](https://discord.gg/c84AZQhmpx) server for a chat and community support.
+Join our [Discord server](https://discord.gg/c84AZQhmpx) for a chat and community support.
 
 ## Design goal
 

docs/ref/integration/web-ui.md

@@ -16,4 +16,4 @@ Headscale doesn't provide a built-in web interface but users may pick one from t
 | headscale-admin | [Github](https://github.com/GoodiesHQ/headscale-admin)  | Headscale-Admin is meant to be a simple, modern web interface for headscale         |
 | ouroboros       | [Github](https://github.com/yellowsink/ouroboros)       | Ouroboros is designed for users to manage their own devices, rather than for admins |
 
-You can ask for support on our dedicated [Discord channel](https://discord.com/channels/896711691637780480/1105842846386356294).
+You can ask for support on our [Discord server](https://discord.gg/c84AZQhmpx) in the "web-interfaces" channel.

hscontrol/poll.go

@@ -191,6 +191,7 @@ func (m *mapSession) serve() {
 //
 //nolint:gocyclo
 func (m *mapSession) serveLongPoll() {
+	start := time.Now()
 	m.beforeServeLongPoll()
 
 	// Clean up the session when the client disconnects
@@ -220,16 +221,6 @@ func (m *mapSession) serveLongPoll() {
 
 	m.pollFailoverRoutes("node connected", m.node)
 
-	// Upgrade the writer to a ResponseController
-	rc := http.NewResponseController(m.w)
-
-	// Longpolling will break if there is a write timeout,
-	// so it needs to be disabled.
-	rc.SetWriteDeadline(time.Time{})
-
-	ctx, cancel := context.WithCancel(context.WithValue(m.ctx, nodeNameContextKey, m.node.Hostname))
-	defer cancel()
-
 	m.keepAliveTicker = time.NewTicker(m.keepAlive)
 
 	m.h.nodeNotifier.AddNode(m.node.ID, m.ch)
@@ -243,12 +234,12 @@ func (m *mapSession) serveLongPoll() {
 		// consume channels with update, keep alives or "batch" blocking signals
 		select {
 		case <-m.cancelCh:
-			m.tracef("poll cancelled received")
+			m.tracef("poll cancelled received (%s)", time.Since(start).String())
 			mapResponseEnded.WithLabelValues("cancelled").Inc()
 			return
 
-		case <-ctx.Done():
+		case <-m.ctx.Done():
-			m.tracef("poll context done")
+			m.tracef("poll context done (%s): %s", time.Since(start).String(), m.ctx.Err().Error())
 			mapResponseEnded.WithLabelValues("done").Inc()
 			return
 
@@ -339,14 +330,7 @@ func (m *mapSession) serveLongPoll() {
 					m.errf(err, "could not write the map response(%s), for mapSession: %p", update.Type.String(), m)
 					return
 				}
-
+				m.w.(http.Flusher).Flush()
-				err = rc.Flush()
-				if err != nil {
-					mapResponseSent.WithLabelValues("error", updateType).Inc()
-					m.errf(err, "flushing the map response to client, for mapSession: %p", m)
-					return
-				}
-
 				log.Trace().Str("node", m.node.Hostname).TimeDiff("timeSpent", time.Now(), startWrite).Str("mkey", m.node.MachineKey.String()).Msg("finished writing mapresp to node")
 
 				if debugHighCardinalityMetrics {
@@ -360,22 +344,17 @@ func (m *mapSession) serveLongPoll() {
 		case <-m.keepAliveTicker.C:
 			data, err := m.mapper.KeepAliveResponse(m.req, m.node)
 			if err != nil {
-				m.errf(err, "Error generating the keep alive msg")
+				m.errf(err, "Error generating the keepalive msg")
 				mapResponseSent.WithLabelValues("error", "keepalive").Inc()
 				return
 			}
 			_, err = m.w.Write(data)
 			if err != nil {
-				m.errf(err, "Cannot write keep alive message")
+				m.errf(err, "Cannot write keepalive message")
-				mapResponseSent.WithLabelValues("error", "keepalive").Inc()
-				return
-			}
-			err = rc.Flush()
-			if err != nil {
-				m.errf(err, "flushing keep alive to client, for mapSession: %p", m)
 				mapResponseSent.WithLabelValues("error", "keepalive").Inc()
 				return
 			}
+			m.w.(http.Flusher).Flush()
 
 			if debugHighCardinalityMetrics {
 				mapResponseLastSentSeconds.WithLabelValues("keepalive", m.node.ID.String()).Set(float64(time.Now().Unix()))

integration/route_test.go

@@ -13,6 +13,7 @@ import (
 	"github.com/google/go-cmp/cmp/cmpopts"
 	v1 "github.com/juanfont/headscale/gen/go/headscale/v1"
 	"github.com/juanfont/headscale/hscontrol/policy"
+	"github.com/juanfont/headscale/hscontrol/types"
 	"github.com/juanfont/headscale/hscontrol/util"
 	"github.com/juanfont/headscale/integration/hsic"
 	"github.com/juanfont/headscale/integration/tsic"
@@ -1316,3 +1317,252 @@ func TestSubnetRouteACL(t *testing.T) {
 		t.Errorf("Subnet (%s) filter, unexpected result (-want +got):\n%s", subRouter1.Hostname(), diff)
 	}
 }
+
+func TestHASubnetRouterFailoverWhenNodeDisconnects2129(t *testing.T) {
+	IntegrationSkip(t)
+	t.Parallel()
+
+	user := "enable-routing"
+
+	scenario, err := NewScenario(dockertestMaxWait())
+	assertNoErrf(t, "failed to create scenario: %s", err)
+	// defer scenario.ShutdownAssertNoPanics(t)
+
+	spec := map[string]int{
+		user: 3,
+	}
+
+	err = scenario.CreateHeadscaleEnv(spec,
+		[]tsic.Option{},
+		hsic.WithTestName("clientdisc"),
+		hsic.WithEmbeddedDERPServerOnly(),
+		hsic.WithTLS(),
+		hsic.WithHostnameAsServerURL(),
+		hsic.WithIPAllocationStrategy(types.IPAllocationStrategyRandom),
+	)
+	assertNoErrHeadscaleEnv(t, err)
+
+	allClients, err := scenario.ListTailscaleClients()
+	assertNoErrListClients(t, err)
+
+	err = scenario.WaitForTailscaleSync()
+	assertNoErrSync(t, err)
+
+	headscale, err := scenario.Headscale()
+	assertNoErrGetHeadscale(t, err)
+
+	expectedRoutes := map[string]string{
+		"1": "10.0.0.0/24",
+		"2": "10.0.0.0/24",
+	}
+
+	// Sort nodes by ID
+	sort.SliceStable(allClients, func(i, j int) bool {
+		statusI, err := allClients[i].Status()
+		if err != nil {
+			return false
+		}
+
+		statusJ, err := allClients[j].Status()
+		if err != nil {
+			return false
+		}
+
+		return statusI.Self.ID < statusJ.Self.ID
+	})
+
+	subRouter1 := allClients[0]
+	subRouter2 := allClients[1]
+
+	t.Logf("Advertise route from r1 (%s) and r2 (%s), making it HA, n1 is primary", subRouter1.Hostname(), subRouter2.Hostname())
+	// advertise HA route on node 1 and 2
+	// ID 1 will be primary
+	// ID 2 will be secondary
+	for _, client := range allClients[:2] {
+		status, err := client.Status()
+		assertNoErr(t, err)
+
+		if route, ok := expectedRoutes[string(status.Self.ID)]; ok {
+			command := []string{
+				"tailscale",
+				"set",
+				"--advertise-routes=" + route,
+			}
+			_, _, err = client.Execute(command)
+			assertNoErrf(t, "failed to advertise route: %s", err)
+		} else {
+			t.Fatalf("failed to find route for Node %s (id: %s)", status.Self.HostName, status.Self.ID)
+		}
+	}
+
+	err = scenario.WaitForTailscaleSync()
+	assertNoErrSync(t, err)
+
+	var routes []*v1.Route
+	err = executeAndUnmarshal(
+		headscale,
+		[]string{
+			"headscale",
+			"routes",
+			"list",
+			"--output",
+			"json",
+		},
+		&routes,
+	)
+
+	assertNoErr(t, err)
+	assert.Len(t, routes, 2)
+
+	t.Logf("initial routes %#v", routes)
+
+	for _, route := range routes {
+		assert.Equal(t, true, route.GetAdvertised())
+		assert.Equal(t, false, route.GetEnabled())
+		assert.Equal(t, false, route.GetIsPrimary())
+	}
+
+	// Verify that no routes has been sent to the client,
+	// they are not yet enabled.
+	for _, client := range allClients {
+		status, err := client.Status()
+		assertNoErr(t, err)
+
+		for _, peerKey := range status.Peers() {
+			peerStatus := status.Peer[peerKey]
+
+			assert.Nil(t, peerStatus.PrimaryRoutes)
+		}
+	}
+
+	// Enable all routes
+	for _, route := range routes {
+		_, err = headscale.Execute(
+			[]string{
+				"headscale",
+				"routes",
+				"enable",
+				"--route",
+				strconv.Itoa(int(route.GetId())),
+			})
+		assertNoErr(t, err)
+
+		time.Sleep(time.Second)
+	}
+
+	var enablingRoutes []*v1.Route
+	err = executeAndUnmarshal(
+		headscale,
+		[]string{
+			"headscale",
+			"routes",
+			"list",
+			"--output",
+			"json",
+		},
+		&enablingRoutes,
+	)
+	assertNoErr(t, err)
+	assert.Len(t, enablingRoutes, 2)
+
+	// Node 1 is primary
+	assert.Equal(t, true, enablingRoutes[0].GetAdvertised())
+	assert.Equal(t, true, enablingRoutes[0].GetEnabled())
+	assert.Equal(t, true, enablingRoutes[0].GetIsPrimary(), "both subnet routers are up, expected r1 to be primary")
+
+	// Node 2 is not primary
+	assert.Equal(t, true, enablingRoutes[1].GetAdvertised())
+	assert.Equal(t, true, enablingRoutes[1].GetEnabled())
+	assert.Equal(t, false, enablingRoutes[1].GetIsPrimary(), "both subnet routers are up, expected r2 to be non-primary")
+
+	var nodeList []v1.Node
+	err = executeAndUnmarshal(
+		headscale,
+		[]string{
+			"headscale",
+			"nodes",
+			"list",
+			"--output",
+			"json",
+		},
+		&nodeList,
+	)
+	assert.Nil(t, err)
+	assert.Len(t, nodeList, 3)
+	assert.True(t, nodeList[0].Online)
+	assert.True(t, nodeList[1].Online)
+	assert.True(t, nodeList[2].Online)
+
+	// Kill off one of the docker containers to simulate a disconnect
+	err = scenario.DisconnectContainersFromScenario(subRouter1.Hostname())
+	assertNoErr(t, err)
+
+	time.Sleep(5 * time.Second)
+
+	var nodeListAfterDisconnect []v1.Node
+	err = executeAndUnmarshal(
+		headscale,
+		[]string{
+			"headscale",
+			"nodes",
+			"list",
+			"--output",
+			"json",
+		},
+		&nodeListAfterDisconnect,
+	)
+	assert.Nil(t, err)
+	assert.Len(t, nodeListAfterDisconnect, 3)
+	assert.False(t, nodeListAfterDisconnect[0].Online)
+	assert.True(t, nodeListAfterDisconnect[1].Online)
+	assert.True(t, nodeListAfterDisconnect[2].Online)
+
+	var routesAfterDisconnect []*v1.Route
+	err = executeAndUnmarshal(
+		headscale,
+		[]string{
+			"headscale",
+			"routes",
+			"list",
+			"--output",
+			"json",
+		},
+		&routesAfterDisconnect,
+	)
+	assertNoErr(t, err)
+	assert.Len(t, routesAfterDisconnect, 2)
+
+	// Node 1 is primary
+	assert.Equal(t, true, routesAfterDisconnect[0].GetAdvertised())
+	assert.Equal(t, true, routesAfterDisconnect[0].GetEnabled())
+	assert.Equal(t, false, routesAfterDisconnect[0].GetIsPrimary(), "both subnet routers are up, expected r1 to be non-primary")
+
+	// Node 2 is not primary
+	assert.Equal(t, true, routesAfterDisconnect[1].GetAdvertised())
+	assert.Equal(t, true, routesAfterDisconnect[1].GetEnabled())
+	assert.Equal(t, true, routesAfterDisconnect[1].GetIsPrimary(), "both subnet routers are up, expected r2 to be primary")
+
+	// // Ensure the node can reconncet as expected
+	// err = scenario.ConnectContainersToScenario(subRouter1.Hostname())
+	// assertNoErr(t, err)
+
+	// time.Sleep(5 * time.Second)
+
+	// var nodeListAfterReconnect []v1.Node
+	// err = executeAndUnmarshal(
+	// 	headscale,
+	// 	[]string{
+	// 		"headscale",
+	// 		"nodes",
+	// 		"list",
+	// 		"--output",
+	// 		"json",
+	// 	},
+	// 	&nodeListAfterReconnect,
+	// )
+	// assert.Nil(t, err)
+	// assert.Len(t, nodeListAfterReconnect, 3)
+	// assert.True(t, nodeListAfterReconnect[0].Online)
+	// assert.True(t, nodeListAfterReconnect[1].Online)
+	// assert.True(t, nodeListAfterReconnect[2].Online)
+}

integration/scenario.go

@@ -651,3 +651,31 @@ func (s *Scenario) WaitForTailscaleLogout() error {
 
 	return nil
 }
+
+// DisconnectContainersFromScenario disconnects a list of containers from the network.
+func (s *Scenario) DisconnectContainersFromScenario(containers ...string) error {
+	for _, container := range containers {
+		if ctr, ok := s.pool.ContainerByName(container); ok {
+			err := ctr.DisconnectFromNetwork(s.network)
+			if err != nil {
+				return err
+			}
+		}
+	}
+
+	return nil
+}
+
+// ConnectContainersToScenario disconnects a list of containers from the network.
+func (s *Scenario) ConnectContainersToScenario(containers ...string) error {
+	for _, container := range containers {
+		if ctr, ok := s.pool.ContainerByName(container); ok {
+			err := ctr.ConnectToNetwork(s.network)
+			if err != nil {
+				return err
+			}
+		}
+	}
+
+	return nil
+}