moby--moby/network_proxy.go

259 lines
7.0 KiB
Go
Raw Normal View History

Add support for UDP (closes #33) API Changes ----------- The port notation is extended to support "/udp" or "/tcp" at the *end* of the specifier string (and defaults to tcp if "/tcp" or "/udp" are missing) `docker ps` now shows UDP ports as "frontend->backend/udp". Nothing changes for TCP ports. `docker inspect` now displays two sub-dictionaries: "Tcp" and "Udp", under "PortMapping" in "NetworkSettings". Theses changes stand true for the values returned by the HTTP API too. This changeset will definitely break tools built upon the API (or upon `docker inspect`). A less intrusive way to add UDP ports in `docker inspect` would be to simply add "/udp" for UDP ports but it will still break existing applications which tries to convert the whole field to an integer. I believe that having two TCP/UDP sub-dictionaries is better because it makes the whole thing more clear and more easy to parse right away (i.e: you don't have to check the format of the string, split it and convert the right part to an integer) Code Changes ------------ Significant changes in network.go: - A second PortAllocator is instantiated for the UDP range; - PortMapper maintains separate mapping for TCP and UDP; - The extPorts array in NetworkInterface is now an array of Nat objects (so we can know on which protocol a given port was mapped when NetworkInterface.Release() is called); - TCP proxying on localhost has been moved away in network_proxy.go. localhost proxy code rewrite in network_proxy.go: We have to proxy the traffic between localhost:frontend-port and container:backend-port because Netfilter doesn't work properly on the loopback interface and DNAT iptable rules aren't applied there. - Goroutines in the TCP proxying code are now explicitly stopped when the proxy is stopped; - UDP connection tracking using a map (more infos in [1]); - Support for IPv6 (to be more accurate, the code is transparent to the Go net package, so you can use, tcp/tcp4/tcp6/udp/udp4/udp6); - Single Proxy interface for both UDP and TCP proxying; - Full test suite. [1] https://github.com/dotcloud/docker/issues/33#issuecomment-20010400
2013-06-11 22:46:23 +00:00
package docker
import (
"encoding/binary"
"fmt"
"github.com/dotcloud/docker/utils"
"io"
"log"
"net"
"sync"
"syscall"
"time"
)
const (
UDPConnTrackTimeout = 90 * time.Second
UDPBufSize = 2048
)
type Proxy interface {
// Start forwarding traffic back and forth the front and back-end
// addresses.
Run()
// Stop forwarding traffic and close both ends of the Proxy.
Close()
// Return the address on which the proxy is listening.
FrontendAddr() net.Addr
// Return the proxied address.
BackendAddr() net.Addr
}
type TCPProxy struct {
listener *net.TCPListener
frontendAddr *net.TCPAddr
backendAddr *net.TCPAddr
}
func NewTCPProxy(frontendAddr, backendAddr *net.TCPAddr) (*TCPProxy, error) {
listener, err := net.ListenTCP("tcp", frontendAddr)
if err != nil {
return nil, err
}
// If the port in frontendAddr was 0 then ListenTCP will have a picked
// a port to listen on, hence the call to Addr to get that actual port:
return &TCPProxy{
listener: listener,
frontendAddr: listener.Addr().(*net.TCPAddr),
backendAddr: backendAddr,
}, nil
}
func (proxy *TCPProxy) clientLoop(client *net.TCPConn, quit chan bool) {
backend, err := net.DialTCP("tcp", nil, proxy.backendAddr)
if err != nil {
log.Printf("Can't forward traffic to backend tcp/%v: %v\n", proxy.backendAddr, err.Error())
client.Close()
return
}
event := make(chan int64)
var broker = func(to, from *net.TCPConn) {
written, err := io.Copy(to, from)
if err != nil {
err, ok := err.(*net.OpError)
// If the socket we are writing to is shutdown with
// SHUT_WR, forward it to the other end of the pipe:
if ok && err.Err == syscall.EPIPE {
from.CloseWrite()
}
}
to.CloseRead()
Add support for UDP (closes #33) API Changes ----------- The port notation is extended to support "/udp" or "/tcp" at the *end* of the specifier string (and defaults to tcp if "/tcp" or "/udp" are missing) `docker ps` now shows UDP ports as "frontend->backend/udp". Nothing changes for TCP ports. `docker inspect` now displays two sub-dictionaries: "Tcp" and "Udp", under "PortMapping" in "NetworkSettings". Theses changes stand true for the values returned by the HTTP API too. This changeset will definitely break tools built upon the API (or upon `docker inspect`). A less intrusive way to add UDP ports in `docker inspect` would be to simply add "/udp" for UDP ports but it will still break existing applications which tries to convert the whole field to an integer. I believe that having two TCP/UDP sub-dictionaries is better because it makes the whole thing more clear and more easy to parse right away (i.e: you don't have to check the format of the string, split it and convert the right part to an integer) Code Changes ------------ Significant changes in network.go: - A second PortAllocator is instantiated for the UDP range; - PortMapper maintains separate mapping for TCP and UDP; - The extPorts array in NetworkInterface is now an array of Nat objects (so we can know on which protocol a given port was mapped when NetworkInterface.Release() is called); - TCP proxying on localhost has been moved away in network_proxy.go. localhost proxy code rewrite in network_proxy.go: We have to proxy the traffic between localhost:frontend-port and container:backend-port because Netfilter doesn't work properly on the loopback interface and DNAT iptable rules aren't applied there. - Goroutines in the TCP proxying code are now explicitly stopped when the proxy is stopped; - UDP connection tracking using a map (more infos in [1]); - Support for IPv6 (to be more accurate, the code is transparent to the Go net package, so you can use, tcp/tcp4/tcp6/udp/udp4/udp6); - Single Proxy interface for both UDP and TCP proxying; - Full test suite. [1] https://github.com/dotcloud/docker/issues/33#issuecomment-20010400
2013-06-11 22:46:23 +00:00
event <- written
}
utils.Debugf("Forwarding traffic between tcp/%v and tcp/%v", client.RemoteAddr(), backend.RemoteAddr())
go broker(client, backend)
go broker(backend, client)
var transferred int64 = 0
for i := 0; i < 2; i++ {
select {
case written := <-event:
transferred += written
case <-quit:
// Interrupt the two brokers and "join" them.
client.Close()
backend.Close()
for ; i < 2; i++ {
transferred += <-event
}
goto done
}
}
client.Close()
backend.Close()
done:
utils.Debugf("%v bytes transferred between tcp/%v and tcp/%v", transferred, client.RemoteAddr(), backend.RemoteAddr())
}
func (proxy *TCPProxy) Run() {
quit := make(chan bool)
defer close(quit)
utils.Debugf("Starting proxy on tcp/%v for tcp/%v", proxy.frontendAddr, proxy.backendAddr)
for {
client, err := proxy.listener.Accept()
if err != nil {
utils.Debugf("Stopping proxy on tcp/%v for tcp/%v (%v)", proxy.frontendAddr, proxy.backendAddr, err.Error())
return
}
go proxy.clientLoop(client.(*net.TCPConn), quit)
}
}
func (proxy *TCPProxy) Close() { proxy.listener.Close() }
func (proxy *TCPProxy) FrontendAddr() net.Addr { return proxy.frontendAddr }
func (proxy *TCPProxy) BackendAddr() net.Addr { return proxy.backendAddr }
// A net.Addr where the IP is split into two fields so you can use it as a key
// in a map:
type connTrackKey struct {
IPHigh uint64
IPLow uint64
Port int
}
func newConnTrackKey(addr *net.UDPAddr) *connTrackKey {
if len(addr.IP) == net.IPv4len {
return &connTrackKey{
IPHigh: 0,
IPLow: uint64(binary.BigEndian.Uint32(addr.IP)),
Port: addr.Port,
}
}
return &connTrackKey{
IPHigh: binary.BigEndian.Uint64(addr.IP[:8]),
IPLow: binary.BigEndian.Uint64(addr.IP[8:]),
Port: addr.Port,
}
}
type connTrackMap map[connTrackKey]*net.UDPConn
type UDPProxy struct {
listener *net.UDPConn
frontendAddr *net.UDPAddr
backendAddr *net.UDPAddr
connTrackTable connTrackMap
connTrackLock sync.Mutex
}
func NewUDPProxy(frontendAddr, backendAddr *net.UDPAddr) (*UDPProxy, error) {
listener, err := net.ListenUDP("udp", frontendAddr)
if err != nil {
return nil, err
}
return &UDPProxy{
listener: listener,
frontendAddr: listener.LocalAddr().(*net.UDPAddr),
backendAddr: backendAddr,
connTrackTable: make(connTrackMap),
}, nil
}
func (proxy *UDPProxy) replyLoop(proxyConn *net.UDPConn, clientAddr *net.UDPAddr, clientKey *connTrackKey) {
defer func() {
proxy.connTrackLock.Lock()
delete(proxy.connTrackTable, *clientKey)
proxy.connTrackLock.Unlock()
utils.Debugf("Done proxying between udp/%v and udp/%v", clientAddr.String(), proxy.backendAddr.String())
proxyConn.Close()
}()
readBuf := make([]byte, UDPBufSize)
for {
proxyConn.SetReadDeadline(time.Now().Add(UDPConnTrackTimeout))
again:
read, err := proxyConn.Read(readBuf)
if err != nil {
if err, ok := err.(*net.OpError); ok && err.Err == syscall.ECONNREFUSED {
// This will happen if the last write failed
// (e.g: nothing is actually listening on the
// proxied port on the container), ignore it
// and continue until UDPConnTrackTimeout
// expires:
goto again
}
return
}
for i := 0; i != read; {
written, err := proxy.listener.WriteToUDP(readBuf[i:read], clientAddr)
if err != nil {
return
}
i += written
utils.Debugf("Forwarded %v/%v bytes to udp/%v", i, read, clientAddr.String())
}
}
}
func (proxy *UDPProxy) Run() {
readBuf := make([]byte, UDPBufSize)
utils.Debugf("Starting proxy on udp/%v for udp/%v", proxy.frontendAddr, proxy.backendAddr)
for {
read, from, err := proxy.listener.ReadFromUDP(readBuf)
if err != nil {
// NOTE: Apparently ReadFrom doesn't return
// ECONNREFUSED like Read do (see comment in
// UDPProxy.replyLoop)
utils.Debugf("Stopping proxy on udp/%v for udp/%v (%v)", proxy.frontendAddr, proxy.backendAddr, err.Error())
break
}
fromKey := newConnTrackKey(from)
proxy.connTrackLock.Lock()
proxyConn, hit := proxy.connTrackTable[*fromKey]
if !hit {
proxyConn, err = net.DialUDP("udp", nil, proxy.backendAddr)
if err != nil {
log.Printf("Can't proxy a datagram to udp/%s: %v\n", proxy.backendAddr.String(), err)
continue
}
proxy.connTrackTable[*fromKey] = proxyConn
go proxy.replyLoop(proxyConn, from, fromKey)
}
proxy.connTrackLock.Unlock()
for i := 0; i != read; {
written, err := proxyConn.Write(readBuf[i:read])
if err != nil {
log.Printf("Can't proxy a datagram to udp/%s: %v\n", proxy.backendAddr.String(), err)
break
}
i += written
utils.Debugf("Forwarded %v/%v bytes to udp/%v", i, read, proxy.backendAddr.String())
}
}
}
func (proxy *UDPProxy) Close() {
proxy.listener.Close()
proxy.connTrackLock.Lock()
defer proxy.connTrackLock.Unlock()
for _, conn := range proxy.connTrackTable {
conn.Close()
}
}
func (proxy *UDPProxy) FrontendAddr() net.Addr { return proxy.frontendAddr }
func (proxy *UDPProxy) BackendAddr() net.Addr { return proxy.backendAddr }
func NewProxy(frontendAddr, backendAddr net.Addr) (Proxy, error) {
switch frontendAddr.(type) {
case *net.UDPAddr:
return NewUDPProxy(frontendAddr.(*net.UDPAddr), backendAddr.(*net.UDPAddr))
case *net.TCPAddr:
return NewTCPProxy(frontendAddr.(*net.TCPAddr), backendAddr.(*net.TCPAddr))
default:
panic(fmt.Errorf("Unsupported protocol"))
}
}