Merge pull request #251 from Arceliar/fastpeer

Tune peering

src/yggdrasil/peer.go

@@ -89,7 +89,7 @@ type peer struct {
 	firstSeen  time.Time       // To track uptime for getPeers
 	linkOut    (chan []byte)   // used for protocol traffic (to bypass queues)
 	doSend     (chan struct{}) // tell the linkLoop to send a switchMsg
-	dinfo      *dhtInfo        // used to keep the DHT working
+	dinfo      (chan *dhtInfo) // used to keep the DHT working
 	out        func([]byte)    // Set up by whatever created the peers struct, used to send packets to other nodes
 	close      func()          // Called when a peer is removed, to close the underlying connection, or via admin api
 }
@@ -104,6 +104,7 @@ func (ps *peers) newPeer(box *boxPubKey, sig *sigPubKey, linkShared *boxSharedKe
 		endpoint:   endpoint,
 		firstSeen:  now,
 		doSend:     make(chan struct{}, 1),
+		dinfo:      make(chan *dhtInfo, 1),
 		core:       ps.core}
 	ps.mutex.Lock()
 	defer ps.mutex.Unlock()
@@ -176,6 +177,8 @@ func (p *peer) doSendSwitchMsgs() {
 func (p *peer) linkLoop() {
 	tick := time.NewTicker(time.Second)
 	defer tick.Stop()
+	p.doSendSwitchMsgs()
+	var dinfo *dhtInfo
 	for {
 		select {
 		case _, ok := <-p.doSend:
@@ -183,12 +186,10 @@ func (p *peer) linkLoop() {
 				return
 			}
 			p.sendSwitchMsg()
+		case dinfo = <-p.dinfo:
 		case _ = <-tick.C:
-			//break             // FIXME disabled the below completely to test something
-			pdinfo := p.dinfo // FIXME this is a bad workarond NPE on the next line
-			if pdinfo != nil {
-				dinfo := *pdinfo
-				p.core.dht.peers <- &dinfo
+			if dinfo != nil {
+				p.core.dht.peers <- dinfo
 			}
 		}
 	}
@@ -218,8 +219,9 @@ func (p *peer) handlePacket(packet []byte) {
 // Called to handle traffic or protocolTraffic packets.
 // In either case, this reads from the coords of the packet header, does a switch lookup, and forwards to the next node.
 func (p *peer) handleTraffic(packet []byte, pTypeLen int) {
-	if p.port != 0 && p.dinfo == nil {
-		// Drop traffic until the peer manages to send us at least one good switchMsg
+	table := p.core.switchTable.getTable()
+	if _, isIn := table.elems[p.port]; !isIn && p.port != 0 {
+		// Drop traffic if the peer isn't in the switch
 		return
 	}
 	p.core.switchTable.packetIn <- packet
@@ -323,9 +325,7 @@ func (p *peer) handleSwitchMsg(packet []byte) {
 	p.core.switchTable.handleMsg(&msg, p.port)
 	if !p.core.switchTable.checkRoot(&msg) {
 		// Bad switch message
-		// Stop forwarding traffic from it
-		// Stop refreshing it in the DHT
-		p.dinfo = nil
+		p.dinfo <- nil
 		return
 	}
 	// Pass a mesage to the dht informing it that this peer (still) exists
@@ -334,8 +334,7 @@ func (p *peer) handleSwitchMsg(packet []byte) {
 		key:    p.box,
 		coords: loc.getCoords(),
 	}
-	//p.core.dht.peers <- &dinfo
-	p.dinfo = &dinfo
+	p.dinfo <- &dinfo
 }
 
 // This generates the bytes that we sign or check the signature of for a switchMsg.

src/yggdrasil/tcp.go

@@ -31,14 +31,6 @@ const tcp_msgSize = 2048 + 65535 // TODO figure out what makes sense
 const default_tcp_timeout = 6 * time.Second
 const tcp_ping_interval = (default_tcp_timeout * 2 / 3)
 
-// Wrapper function for non tcp/ip connections.
-func setNoDelay(c net.Conn, delay bool) {
-	tcp, ok := c.(*net.TCPConn)
-	if ok {
-		tcp.SetNoDelay(delay)
-	}
-}
-
 // The TCP listener and information about active TCP connections, to avoid duplication.
 type tcpInterface struct {
 	core        *Core
@@ -58,6 +50,18 @@ type tcpInfo struct {
 	remoteAddr string
 }
 
+// Wrapper function to set additional options for specific connection types.
+func (iface *tcpInterface) setExtraOptions(c net.Conn) {
+	switch sock := c.(type) {
+	case *net.TCPConn:
+		sock.SetNoDelay(true)
+		sock.SetKeepAlive(true)
+		sock.SetKeepAlivePeriod(iface.tcp_timeout)
+	// TODO something for socks5
+	default:
+	}
+}
+
 // Returns the address of the listener.
 func (iface *tcpInterface) getAddr() *net.TCPAddr {
 	return iface.serv.Addr().(*net.TCPAddr)
@@ -205,6 +209,7 @@ func (iface *tcpInterface) call(saddr string, socksaddr *string, sintf string) {
 // It defers a bunch of cleanup stuff to tear down all of these things when the reader exists (e.g. due to a closed connection or a timeout).
 func (iface *tcpInterface) handler(sock net.Conn, incoming bool) {
 	defer sock.Close()
+	iface.setExtraOptions(sock)
 	// Get our keys
 	myLinkPub, myLinkPriv := newBoxKeys() // ephemeral link keys
 	meta := version_getBaseMetadata()
@@ -342,7 +347,6 @@ func (iface *tcpInterface) handler(sock net.Conn, incoming bool) {
 		out <- msg
 	}
 	p.close = func() { sock.Close() }
-	setNoDelay(sock, true)
 	go p.linkLoop()
 	defer func() {
 		// Put all of our cleanup here...