forked from moby/moby
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathnetwork.go
413 lines (357 loc) · 12 KB
/
network.go
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
package docker
import (
"fmt"
"github.com/dotcloud/docker/networkdriver"
"github.com/dotcloud/docker/networkdriver/ipallocator"
"github.com/dotcloud/docker/networkdriver/portallocator"
"github.com/dotcloud/docker/networkdriver/portmapper"
"github.com/dotcloud/docker/pkg/iptables"
"github.com/dotcloud/docker/pkg/netlink"
"github.com/dotcloud/docker/utils"
"io/ioutil"
"log"
"net"
"strconv"
"syscall"
"unsafe"
)
const (
DefaultNetworkBridge = "docker0"
DisableNetworkBridge = "none"
DefaultNetworkMtu = 1500
siocBRADDBR = 0x89a0
)
// CreateBridgeIface creates a network bridge interface on the host system with the name `ifaceName`,
// and attempts to configure it with an address which doesn't conflict with any other interface on the host.
// If it can't find an address which doesn't conflict, it will return an error.
func CreateBridgeIface(config *DaemonConfig) error {
addrs := []string{
// Here we don't follow the convention of using the 1st IP of the range for the gateway.
// This is to use the same gateway IPs as the /24 ranges, which predate the /16 ranges.
// In theory this shouldn't matter - in practice there's bound to be a few scripts relying
// on the internal addressing or other stupid things like that.
// The shouldn't, but hey, let's not break them unless we really have to.
"172.17.42.1/16", // Don't use 172.16.0.0/16, it conflicts with EC2 DNS 172.16.0.23
"10.0.42.1/16", // Don't even try using the entire /8, that's too intrusive
"10.1.42.1/16",
"10.42.42.1/16",
"172.16.42.1/24",
"172.16.43.1/24",
"172.16.44.1/24",
"10.0.42.1/24",
"10.0.43.1/24",
"192.168.42.1/24",
"192.168.43.1/24",
"192.168.44.1/24",
}
nameservers := []string{}
resolvConf, _ := utils.GetResolvConf()
// we don't check for an error here, because we don't really care
// if we can't read /etc/resolv.conf. So instead we skip the append
// if resolvConf is nil. It either doesn't exist, or we can't read it
// for some reason.
if resolvConf != nil {
nameservers = append(nameservers, utils.GetNameserversAsCIDR(resolvConf)...)
}
var ifaceAddr string
if len(config.BridgeIp) != 0 {
_, _, err := net.ParseCIDR(config.BridgeIp)
if err != nil {
return err
}
ifaceAddr = config.BridgeIp
} else {
for _, addr := range addrs {
_, dockerNetwork, err := net.ParseCIDR(addr)
if err != nil {
return err
}
if err := networkdriver.CheckNameserverOverlaps(nameservers, dockerNetwork); err == nil {
if err := networkdriver.CheckRouteOverlaps(dockerNetwork); err == nil {
ifaceAddr = addr
break
} else {
utils.Debugf("%s %s", addr, err)
}
}
}
}
if ifaceAddr == "" {
return fmt.Errorf("Could not find a free IP address range for interface '%s'. Please configure its address manually and run 'docker -b %s'", config.BridgeIface, config.BridgeIface)
}
utils.Debugf("Creating bridge %s with network %s", config.BridgeIface, ifaceAddr)
if err := createBridgeIface(config.BridgeIface); err != nil {
return err
}
iface, err := net.InterfaceByName(config.BridgeIface)
if err != nil {
return err
}
ipAddr, ipNet, err := net.ParseCIDR(ifaceAddr)
if err != nil {
return err
}
if netlink.NetworkLinkAddIp(iface, ipAddr, ipNet); err != nil {
return fmt.Errorf("Unable to add private network: %s", err)
}
if err := netlink.NetworkLinkUp(iface); err != nil {
return fmt.Errorf("Unable to start network bridge: %s", err)
}
return nil
}
// Create the actual bridge device. This is more backward-compatible than
// netlink.NetworkLinkAdd and works on RHEL 6.
func createBridgeIface(name string) error {
s, err := syscall.Socket(syscall.AF_INET6, syscall.SOCK_STREAM, syscall.IPPROTO_IP)
if err != nil {
utils.Debugf("Bridge socket creation failed IPv6 probably not enabled: %v", err)
s, err = syscall.Socket(syscall.AF_INET, syscall.SOCK_STREAM, syscall.IPPROTO_IP)
if err != nil {
return fmt.Errorf("Error creating bridge creation socket: %s", err)
}
}
defer syscall.Close(s)
nameBytePtr, err := syscall.BytePtrFromString(name)
if err != nil {
return fmt.Errorf("Error converting bridge name %s to byte array: %s", name, err)
}
if _, _, err := syscall.Syscall(syscall.SYS_IOCTL, uintptr(s), siocBRADDBR, uintptr(unsafe.Pointer(nameBytePtr))); err != 0 {
return fmt.Errorf("Error creating bridge: %s", err)
}
return nil
}
// Return the IPv4 address of a network interface
func getIfaceAddr(name string) (net.Addr, error) {
iface, err := net.InterfaceByName(name)
if err != nil {
return nil, err
}
addrs, err := iface.Addrs()
if err != nil {
return nil, err
}
var addrs4 []net.Addr
for _, addr := range addrs {
ip := (addr.(*net.IPNet)).IP
if ip4 := ip.To4(); len(ip4) == net.IPv4len {
addrs4 = append(addrs4, addr)
}
}
switch {
case len(addrs4) == 0:
return nil, fmt.Errorf("Interface %v has no IP addresses", name)
case len(addrs4) > 1:
fmt.Printf("Interface %v has more than 1 IPv4 address. Defaulting to using %v\n",
name, (addrs4[0].(*net.IPNet)).IP)
}
return addrs4[0], nil
}
// Network interface represents the networking stack of a container
type NetworkInterface struct {
IPNet net.IPNet
Gateway net.IP
manager *NetworkManager
extPorts []*Nat
disabled bool
}
// Allocate an external port and map it to the interface
func (iface *NetworkInterface) AllocatePort(port Port, binding PortBinding) (*Nat, error) {
if iface.disabled {
return nil, fmt.Errorf("Trying to allocate port for interface %v, which is disabled", iface) // FIXME
}
ip := iface.manager.defaultBindingIP
if binding.HostIp != "" {
ip = net.ParseIP(binding.HostIp)
} else {
binding.HostIp = ip.String()
}
nat := &Nat{
Port: port,
Binding: binding,
}
containerPort, err := parsePort(port.Port())
if err != nil {
return nil, err
}
hostPort, _ := parsePort(nat.Binding.HostPort)
extPort, err := portallocator.RequestPort(ip, nat.Port.Proto(), hostPort)
if err != nil {
return nil, err
}
var backend net.Addr
if nat.Port.Proto() == "tcp" {
backend = &net.TCPAddr{IP: iface.IPNet.IP, Port: containerPort}
} else {
backend = &net.UDPAddr{IP: iface.IPNet.IP, Port: containerPort}
}
if err := portmapper.Map(backend, ip, extPort); err != nil {
portallocator.ReleasePort(ip, nat.Port.Proto(), extPort)
return nil, err
}
nat.Binding.HostPort = strconv.Itoa(extPort)
iface.extPorts = append(iface.extPorts, nat)
return nat, nil
}
type Nat struct {
Port Port
Binding PortBinding
}
func (n *Nat) String() string {
return fmt.Sprintf("%s:%s:%s/%s", n.Binding.HostIp, n.Binding.HostPort, n.Port.Port(), n.Port.Proto())
}
// Release: Network cleanup - release all resources
func (iface *NetworkInterface) Release() {
if iface.disabled {
return
}
for _, nat := range iface.extPorts {
hostPort, err := parsePort(nat.Binding.HostPort)
if err != nil {
log.Printf("Unable to get host port: %s", err)
continue
}
ip := net.ParseIP(nat.Binding.HostIp)
utils.Debugf("Unmaping %s/%s:%s", nat.Port.Proto, ip.String(), nat.Binding.HostPort)
var host net.Addr
if nat.Port.Proto() == "tcp" {
host = &net.TCPAddr{IP: ip, Port: hostPort}
} else {
host = &net.UDPAddr{IP: ip, Port: hostPort}
}
if err := portmapper.Unmap(host); err != nil {
log.Printf("Unable to unmap port %s: %s", nat, err)
}
if err := portallocator.ReleasePort(ip, nat.Port.Proto(), hostPort); err != nil {
log.Printf("Unable to release port %s", nat)
}
}
if err := ipallocator.ReleaseIP(iface.manager.bridgeNetwork, &iface.IPNet.IP); err != nil {
log.Printf("Unable to release ip %s\n", err)
}
}
// Network Manager manages a set of network interfaces
// Only *one* manager per host machine should be used
type NetworkManager struct {
bridgeIface string
bridgeNetwork *net.IPNet
defaultBindingIP net.IP
disabled bool
}
// Allocate a network interface
func (manager *NetworkManager) Allocate() (*NetworkInterface, error) {
if manager.disabled {
return &NetworkInterface{disabled: true}, nil
}
var ip *net.IP
var err error
ip, err = ipallocator.RequestIP(manager.bridgeNetwork, nil)
if err != nil {
return nil, err
}
iface := &NetworkInterface{
IPNet: net.IPNet{IP: *ip, Mask: manager.bridgeNetwork.Mask},
Gateway: manager.bridgeNetwork.IP,
manager: manager,
}
return iface, nil
}
func newNetworkManager(config *DaemonConfig) (*NetworkManager, error) {
if config.BridgeIface == DisableNetworkBridge {
manager := &NetworkManager{
disabled: true,
}
return manager, nil
}
var network *net.IPNet
addr, err := getIfaceAddr(config.BridgeIface)
if err != nil {
// If the iface is not found, try to create it
if err := CreateBridgeIface(config); err != nil {
return nil, err
}
addr, err = getIfaceAddr(config.BridgeIface)
if err != nil {
return nil, err
}
network = addr.(*net.IPNet)
} else {
network = addr.(*net.IPNet)
}
// Configure iptables for link support
if config.EnableIptables {
// Enable NAT
natArgs := []string{"POSTROUTING", "-t", "nat", "-s", addr.String(), "!", "-d", addr.String(), "-j", "MASQUERADE"}
if !iptables.Exists(natArgs...) {
if output, err := iptables.Raw(append([]string{"-A"}, natArgs...)...); err != nil {
return nil, fmt.Errorf("Unable to enable network bridge NAT: %s", err)
} else if len(output) != 0 {
return nil, fmt.Errorf("Error iptables postrouting: %s", output)
}
}
args := []string{"FORWARD", "-i", config.BridgeIface, "-o", config.BridgeIface, "-j"}
acceptArgs := append(args, "ACCEPT")
dropArgs := append(args, "DROP")
if !config.InterContainerCommunication {
iptables.Raw(append([]string{"-D"}, acceptArgs...)...)
if !iptables.Exists(dropArgs...) {
utils.Debugf("Disable inter-container communication")
if output, err := iptables.Raw(append([]string{"-I"}, dropArgs...)...); err != nil {
return nil, fmt.Errorf("Unable to prevent intercontainer communication: %s", err)
} else if len(output) != 0 {
return nil, fmt.Errorf("Error disabling intercontainer communication: %s", output)
}
}
} else {
iptables.Raw(append([]string{"-D"}, dropArgs...)...)
if !iptables.Exists(acceptArgs...) {
utils.Debugf("Enable inter-container communication")
if output, err := iptables.Raw(append([]string{"-I"}, acceptArgs...)...); err != nil {
return nil, fmt.Errorf("Unable to allow intercontainer communication: %s", err)
} else if len(output) != 0 {
return nil, fmt.Errorf("Error enabling intercontainer communication: %s", output)
}
}
}
// Accept all non-intercontainer outgoing packets
outgoingArgs := []string{"FORWARD", "-i", config.BridgeIface, "!", "-o", config.BridgeIface, "-j", "ACCEPT"}
if !iptables.Exists(outgoingArgs...) {
if output, err := iptables.Raw(append([]string{"-I"}, outgoingArgs...)...); err != nil {
return nil, fmt.Errorf("Unable to allow outgoing packets: %s", err)
} else if len(output) != 0 {
return nil, fmt.Errorf("Error iptables allow outgoing: %s", output)
}
}
// Accept incoming packets for existing connections
existingArgs := []string{"FORWARD", "-o", config.BridgeIface, "-m", "conntrack", "--ctstate", "RELATED,ESTABLISHED", "-j", "ACCEPT"}
if !iptables.Exists(existingArgs...) {
if output, err := iptables.Raw(append([]string{"-I"}, existingArgs...)...); err != nil {
return nil, fmt.Errorf("Unable to allow incoming packets: %s", err)
} else if len(output) != 0 {
return nil, fmt.Errorf("Error iptables allow incoming: %s", output)
}
}
}
if config.EnableIpForward {
// Enable IPv4 forwarding
if err := ioutil.WriteFile("/proc/sys/net/ipv4/ip_forward", []byte{'1', '\n'}, 0644); err != nil {
log.Printf("WARNING: unable to enable IPv4 forwarding: %s\n", err)
}
}
// We can always try removing the iptables
if err := iptables.RemoveExistingChain("DOCKER"); err != nil {
return nil, err
}
if config.EnableIptables {
chain, err := iptables.NewChain("DOCKER", config.BridgeIface)
if err != nil {
return nil, err
}
portmapper.SetIptablesChain(chain)
}
manager := &NetworkManager{
bridgeIface: config.BridgeIface,
bridgeNetwork: network,
defaultBindingIP: config.DefaultIp,
}
return manager, nil
}