/usr/share/gocode/src/github.com/opencontainers/runc/libcontainer/container_linux.go is in golang-github-opencontainers-runc-dev 0.0.8+dfsg-2.
This file is owned by root:root, with mode 0o644.
The actual contents of the file can be viewed below.
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 | // +build linux
package libcontainer
import (
"bytes"
"encoding/json"
"fmt"
"io"
"io/ioutil"
"os"
"os/exec"
"path/filepath"
"reflect"
"strings"
"sync"
"syscall"
"time"
"github.com/Sirupsen/logrus"
"github.com/golang/protobuf/proto"
"github.com/opencontainers/runc/libcontainer/cgroups"
"github.com/opencontainers/runc/libcontainer/configs"
"github.com/opencontainers/runc/libcontainer/criurpc"
"github.com/opencontainers/runc/libcontainer/utils"
"github.com/vishvananda/netlink/nl"
)
const stdioFdCount = 3
type linuxContainer struct {
id string
root string
config *configs.Config
cgroupManager cgroups.Manager
initPath string
initArgs []string
initProcess parentProcess
criuPath string
m sync.Mutex
criuVersion int
state containerState
created time.Time
}
// State represents a running container's state
type State struct {
BaseState
// Platform specific fields below here
// Path to all the cgroups setup for a container. Key is cgroup subsystem name
// with the value as the path.
CgroupPaths map[string]string `json:"cgroup_paths"`
// NamespacePaths are filepaths to the container's namespaces. Key is the namespace type
// with the value as the path.
NamespacePaths map[configs.NamespaceType]string `json:"namespace_paths"`
// Container's standard descriptors (std{in,out,err}), needed for checkpoint and restore
ExternalDescriptors []string `json:"external_descriptors,omitempty"`
}
// A libcontainer container object.
//
// Each container is thread-safe within the same process. Since a container can
// be destroyed by a separate process, any function may return that the container
// was not found.
type Container interface {
BaseContainer
// Methods below here are platform specific
// Checkpoint checkpoints the running container's state to disk using the criu(8) utility.
//
// errors:
// Systemerror - System error.
Checkpoint(criuOpts *CriuOpts) error
// Restore restores the checkpointed container to a running state using the criu(8) utiity.
//
// errors:
// Systemerror - System error.
Restore(process *Process, criuOpts *CriuOpts) error
// If the Container state is RUNNING or PAUSING, sets the Container state to PAUSING and pauses
// the execution of any user processes. Asynchronously, when the container finished being paused the
// state is changed to PAUSED.
// If the Container state is PAUSED, do nothing.
//
// errors:
// ContainerDestroyed - Container no longer exists,
// Systemerror - System error.
Pause() error
// If the Container state is PAUSED, resumes the execution of any user processes in the
// Container before setting the Container state to RUNNING.
// If the Container state is RUNNING, do nothing.
//
// errors:
// ContainerDestroyed - Container no longer exists,
// Systemerror - System error.
Resume() error
// NotifyOOM returns a read-only channel signaling when the container receives an OOM notification.
//
// errors:
// Systemerror - System error.
NotifyOOM() (<-chan struct{}, error)
// NotifyMemoryPressure returns a read-only channel signaling when the container reaches a given pressure level
//
// errors:
// Systemerror - System error.
NotifyMemoryPressure(level PressureLevel) (<-chan struct{}, error)
}
// ID returns the container's unique ID
func (c *linuxContainer) ID() string {
return c.id
}
// Config returns the container's configuration
func (c *linuxContainer) Config() configs.Config {
return *c.config
}
func (c *linuxContainer) Status() (Status, error) {
c.m.Lock()
defer c.m.Unlock()
return c.currentStatus()
}
func (c *linuxContainer) State() (*State, error) {
c.m.Lock()
defer c.m.Unlock()
return c.currentState()
}
func (c *linuxContainer) Processes() ([]int, error) {
pids, err := c.cgroupManager.GetAllPids()
if err != nil {
return nil, newSystemError(err)
}
return pids, nil
}
func (c *linuxContainer) Stats() (*Stats, error) {
var (
err error
stats = &Stats{}
)
if stats.CgroupStats, err = c.cgroupManager.GetStats(); err != nil {
return stats, newSystemError(err)
}
for _, iface := range c.config.Networks {
switch iface.Type {
case "veth":
istats, err := getNetworkInterfaceStats(iface.HostInterfaceName)
if err != nil {
return stats, newSystemError(err)
}
stats.Interfaces = append(stats.Interfaces, istats)
}
}
return stats, nil
}
func (c *linuxContainer) Set(config configs.Config) error {
c.m.Lock()
defer c.m.Unlock()
c.config = &config
return c.cgroupManager.Set(c.config)
}
func (c *linuxContainer) Start(process *Process) error {
c.m.Lock()
defer c.m.Unlock()
status, err := c.currentStatus()
if err != nil {
return err
}
doInit := status == Destroyed
parent, err := c.newParentProcess(process, doInit)
if err != nil {
return newSystemError(err)
}
if err := parent.start(); err != nil {
// terminate the process to ensure that it properly is reaped.
if err := parent.terminate(); err != nil {
logrus.Warn(err)
}
return newSystemError(err)
}
// generate a timestamp indicating when the container was started
c.created = time.Now().UTC()
c.state = &runningState{
c: c,
}
if doInit {
if err := c.updateState(parent); err != nil {
return err
}
if c.config.Hooks != nil {
s := configs.HookState{
Version: c.config.Version,
ID: c.id,
Pid: parent.pid(),
Root: c.config.Rootfs,
}
for _, hook := range c.config.Hooks.Poststart {
if err := hook.Run(s); err != nil {
if err := parent.terminate(); err != nil {
logrus.Warn(err)
}
return newSystemError(err)
}
}
}
}
return nil
}
func (c *linuxContainer) Signal(s os.Signal) error {
if err := c.initProcess.signal(s); err != nil {
return newSystemError(err)
}
return nil
}
func (c *linuxContainer) newParentProcess(p *Process, doInit bool) (parentProcess, error) {
parentPipe, childPipe, err := newPipe()
if err != nil {
return nil, newSystemError(err)
}
cmd, err := c.commandTemplate(p, childPipe)
if err != nil {
return nil, newSystemError(err)
}
if !doInit {
return c.newSetnsProcess(p, cmd, parentPipe, childPipe)
}
return c.newInitProcess(p, cmd, parentPipe, childPipe)
}
func (c *linuxContainer) commandTemplate(p *Process, childPipe *os.File) (*exec.Cmd, error) {
cmd := &exec.Cmd{
Path: c.initPath,
Args: c.initArgs,
}
cmd.Stdin = p.Stdin
cmd.Stdout = p.Stdout
cmd.Stderr = p.Stderr
cmd.Dir = c.config.Rootfs
if cmd.SysProcAttr == nil {
cmd.SysProcAttr = &syscall.SysProcAttr{}
}
cmd.ExtraFiles = append(p.ExtraFiles, childPipe)
cmd.Env = append(cmd.Env, fmt.Sprintf("_LIBCONTAINER_INITPIPE=%d", stdioFdCount+len(cmd.ExtraFiles)-1))
// NOTE: when running a container with no PID namespace and the parent process spawning the container is
// PID1 the pdeathsig is being delivered to the container's init process by the kernel for some reason
// even with the parent still running.
if c.config.ParentDeathSignal > 0 {
cmd.SysProcAttr.Pdeathsig = syscall.Signal(c.config.ParentDeathSignal)
}
return cmd, nil
}
func (c *linuxContainer) newInitProcess(p *Process, cmd *exec.Cmd, parentPipe, childPipe *os.File) (*initProcess, error) {
t := "_LIBCONTAINER_INITTYPE=" + string(initStandard)
cloneFlags := c.config.Namespaces.CloneFlags()
if cloneFlags&syscall.CLONE_NEWUSER != 0 {
if err := c.addUidGidMappings(cmd.SysProcAttr); err != nil {
// user mappings are not supported
return nil, err
}
enableSetgroups(cmd.SysProcAttr)
// Default to root user when user namespaces are enabled.
if cmd.SysProcAttr.Credential == nil {
cmd.SysProcAttr.Credential = &syscall.Credential{}
}
}
cmd.Env = append(cmd.Env, t)
cmd.SysProcAttr.Cloneflags = cloneFlags
return &initProcess{
cmd: cmd,
childPipe: childPipe,
parentPipe: parentPipe,
manager: c.cgroupManager,
config: c.newInitConfig(p),
container: c,
process: p,
}, nil
}
func (c *linuxContainer) newSetnsProcess(p *Process, cmd *exec.Cmd, parentPipe, childPipe *os.File) (*setnsProcess, error) {
cmd.Env = append(cmd.Env, "_LIBCONTAINER_INITTYPE="+string(initSetns))
// for setns process, we dont have to set cloneflags as the process namespaces
// will only be set via setns syscall
data, err := c.bootstrapData(0, c.initProcess.pid(), p.consolePath)
if err != nil {
return nil, err
}
// TODO: set on container for process management
return &setnsProcess{
cmd: cmd,
cgroupPaths: c.cgroupManager.GetPaths(),
childPipe: childPipe,
parentPipe: parentPipe,
config: c.newInitConfig(p),
process: p,
bootstrapData: data,
}, nil
}
func (c *linuxContainer) newInitConfig(process *Process) *initConfig {
return &initConfig{
Config: c.config,
Args: process.Args,
Env: process.Env,
User: process.User,
Cwd: process.Cwd,
Console: process.consolePath,
Capabilities: process.Capabilities,
PassedFilesCount: len(process.ExtraFiles),
}
}
func newPipe() (parent *os.File, child *os.File, err error) {
fds, err := syscall.Socketpair(syscall.AF_LOCAL, syscall.SOCK_STREAM|syscall.SOCK_CLOEXEC, 0)
if err != nil {
return nil, nil, err
}
return os.NewFile(uintptr(fds[1]), "parent"), os.NewFile(uintptr(fds[0]), "child"), nil
}
func (c *linuxContainer) Destroy() error {
c.m.Lock()
defer c.m.Unlock()
return c.state.destroy()
}
func (c *linuxContainer) Pause() error {
c.m.Lock()
defer c.m.Unlock()
status, err := c.currentStatus()
if err != nil {
return err
}
if status != Running {
return newGenericError(fmt.Errorf("container not running"), ContainerNotRunning)
}
if err := c.cgroupManager.Freeze(configs.Frozen); err != nil {
return err
}
return c.state.transition(&pausedState{
c: c,
})
}
func (c *linuxContainer) Resume() error {
c.m.Lock()
defer c.m.Unlock()
status, err := c.currentStatus()
if err != nil {
return err
}
if status != Paused {
return newGenericError(fmt.Errorf("container not paused"), ContainerNotPaused)
}
if err := c.cgroupManager.Freeze(configs.Thawed); err != nil {
return err
}
return c.state.transition(&runningState{
c: c,
})
}
func (c *linuxContainer) NotifyOOM() (<-chan struct{}, error) {
return notifyOnOOM(c.cgroupManager.GetPaths())
}
func (c *linuxContainer) NotifyMemoryPressure(level PressureLevel) (<-chan struct{}, error) {
return notifyMemoryPressure(c.cgroupManager.GetPaths(), level)
}
// XXX debug support, remove when debugging done.
func addArgsFromEnv(evar string, args *[]string) {
if e := os.Getenv(evar); e != "" {
for _, f := range strings.Fields(e) {
*args = append(*args, f)
}
}
fmt.Printf(">>> criu %v\n", *args)
}
// check Criu version greater than or equal to min_version
func (c *linuxContainer) checkCriuVersion(min_version string) error {
var x, y, z, versionReq int
_, err := fmt.Sscanf(min_version, "%d.%d.%d\n", &x, &y, &z) // 1.5.2
if err != nil {
_, err = fmt.Sscanf(min_version, "Version: %d.%d\n", &x, &y) // 1.6
}
versionReq = x*10000 + y*100 + z
out, err := exec.Command(c.criuPath, "-V").Output()
if err != nil {
return fmt.Errorf("Unable to execute CRIU command: %s", c.criuPath)
}
x = 0
y = 0
z = 0
if ep := strings.Index(string(out), "-"); ep >= 0 {
// criu Git version format
var version string
if sp := strings.Index(string(out), "GitID"); sp > 0 {
version = string(out)[sp:ep]
} else {
return fmt.Errorf("Unable to parse the CRIU version: %s", c.criuPath)
}
n, err := fmt.Sscanf(string(version), "GitID: v%d.%d.%d", &x, &y, &z) // 1.5.2
if err != nil {
n, err = fmt.Sscanf(string(version), "GitID: v%d.%d", &x, &y) // 1.6
y++
} else {
z++
}
if n < 2 || err != nil {
return fmt.Errorf("Unable to parse the CRIU version: %s %d %s", version, n, err)
}
} else {
// criu release version format
n, err := fmt.Sscanf(string(out), "Version: %d.%d.%d\n", &x, &y, &z) // 1.5.2
if err != nil {
n, err = fmt.Sscanf(string(out), "Version: %d.%d\n", &x, &y) // 1.6
}
if n < 2 || err != nil {
return fmt.Errorf("Unable to parse the CRIU version: %s %d %s", out, n, err)
}
}
c.criuVersion = x*10000 + y*100 + z
if c.criuVersion < versionReq {
return fmt.Errorf("CRIU version must be %s or higher", min_version)
}
return nil
}
const descriptorsFilename = "descriptors.json"
func (c *linuxContainer) addCriuDumpMount(req *criurpc.CriuReq, m *configs.Mount) {
mountDest := m.Destination
if strings.HasPrefix(mountDest, c.config.Rootfs) {
mountDest = mountDest[len(c.config.Rootfs):]
}
extMnt := &criurpc.ExtMountMap{
Key: proto.String(mountDest),
Val: proto.String(mountDest),
}
req.Opts.ExtMnt = append(req.Opts.ExtMnt, extMnt)
}
func (c *linuxContainer) Checkpoint(criuOpts *CriuOpts) error {
c.m.Lock()
defer c.m.Unlock()
if err := c.checkCriuVersion("1.5.2"); err != nil {
return err
}
if criuOpts.ImagesDirectory == "" {
return fmt.Errorf("invalid directory to save checkpoint")
}
// Since a container can be C/R'ed multiple times,
// the checkpoint directory may already exist.
if err := os.Mkdir(criuOpts.ImagesDirectory, 0755); err != nil && !os.IsExist(err) {
return err
}
if criuOpts.WorkDirectory == "" {
criuOpts.WorkDirectory = filepath.Join(c.root, "criu.work")
}
if err := os.Mkdir(criuOpts.WorkDirectory, 0755); err != nil && !os.IsExist(err) {
return err
}
workDir, err := os.Open(criuOpts.WorkDirectory)
if err != nil {
return err
}
defer workDir.Close()
imageDir, err := os.Open(criuOpts.ImagesDirectory)
if err != nil {
return err
}
defer imageDir.Close()
rpcOpts := criurpc.CriuOpts{
ImagesDirFd: proto.Int32(int32(imageDir.Fd())),
WorkDirFd: proto.Int32(int32(workDir.Fd())),
LogLevel: proto.Int32(4),
LogFile: proto.String("dump.log"),
Root: proto.String(c.config.Rootfs),
ManageCgroups: proto.Bool(true),
NotifyScripts: proto.Bool(true),
Pid: proto.Int32(int32(c.initProcess.pid())),
ShellJob: proto.Bool(criuOpts.ShellJob),
LeaveRunning: proto.Bool(criuOpts.LeaveRunning),
TcpEstablished: proto.Bool(criuOpts.TcpEstablished),
ExtUnixSk: proto.Bool(criuOpts.ExternalUnixConnections),
FileLocks: proto.Bool(criuOpts.FileLocks),
}
// append optional criu opts, e.g., page-server and port
if criuOpts.PageServer.Address != "" && criuOpts.PageServer.Port != 0 {
rpcOpts.Ps = &criurpc.CriuPageServerInfo{
Address: proto.String(criuOpts.PageServer.Address),
Port: proto.Int32(criuOpts.PageServer.Port),
}
}
// append optional manage cgroups mode
if criuOpts.ManageCgroupsMode != 0 {
if err := c.checkCriuVersion("1.7"); err != nil {
return err
}
rpcOpts.ManageCgroupsMode = proto.Uint32(uint32(criuOpts.ManageCgroupsMode))
}
t := criurpc.CriuReqType_DUMP
req := &criurpc.CriuReq{
Type: &t,
Opts: &rpcOpts,
}
for _, m := range c.config.Mounts {
switch m.Device {
case "bind":
c.addCriuDumpMount(req, m)
break
case "cgroup":
binds, err := getCgroupMounts(m)
if err != nil {
return err
}
for _, b := range binds {
c.addCriuDumpMount(req, b)
}
break
}
}
// Write the FD info to a file in the image directory
fdsJSON, err := json.Marshal(c.initProcess.externalDescriptors())
if err != nil {
return err
}
err = ioutil.WriteFile(filepath.Join(criuOpts.ImagesDirectory, descriptorsFilename), fdsJSON, 0655)
if err != nil {
return err
}
err = c.criuSwrk(nil, req, criuOpts, false)
if err != nil {
return err
}
return nil
}
func (c *linuxContainer) addCriuRestoreMount(req *criurpc.CriuReq, m *configs.Mount) {
mountDest := m.Destination
if strings.HasPrefix(mountDest, c.config.Rootfs) {
mountDest = mountDest[len(c.config.Rootfs):]
}
extMnt := &criurpc.ExtMountMap{
Key: proto.String(mountDest),
Val: proto.String(m.Source),
}
req.Opts.ExtMnt = append(req.Opts.ExtMnt, extMnt)
}
func (c *linuxContainer) Restore(process *Process, criuOpts *CriuOpts) error {
c.m.Lock()
defer c.m.Unlock()
if err := c.checkCriuVersion("1.5.2"); err != nil {
return err
}
if criuOpts.WorkDirectory == "" {
criuOpts.WorkDirectory = filepath.Join(c.root, "criu.work")
}
// Since a container can be C/R'ed multiple times,
// the work directory may already exist.
if err := os.Mkdir(criuOpts.WorkDirectory, 0655); err != nil && !os.IsExist(err) {
return err
}
workDir, err := os.Open(criuOpts.WorkDirectory)
if err != nil {
return err
}
defer workDir.Close()
if criuOpts.ImagesDirectory == "" {
return fmt.Errorf("invalid directory to restore checkpoint")
}
imageDir, err := os.Open(criuOpts.ImagesDirectory)
if err != nil {
return err
}
defer imageDir.Close()
// CRIU has a few requirements for a root directory:
// * it must be a mount point
// * its parent must not be overmounted
// c.config.Rootfs is bind-mounted to a temporary directory
// to satisfy these requirements.
root := filepath.Join(c.root, "criu-root")
if err := os.Mkdir(root, 0755); err != nil {
return err
}
defer os.Remove(root)
root, err = filepath.EvalSymlinks(root)
if err != nil {
return err
}
err = syscall.Mount(c.config.Rootfs, root, "", syscall.MS_BIND|syscall.MS_REC, "")
if err != nil {
return err
}
defer syscall.Unmount(root, syscall.MNT_DETACH)
t := criurpc.CriuReqType_RESTORE
req := &criurpc.CriuReq{
Type: &t,
Opts: &criurpc.CriuOpts{
ImagesDirFd: proto.Int32(int32(imageDir.Fd())),
WorkDirFd: proto.Int32(int32(workDir.Fd())),
EvasiveDevices: proto.Bool(true),
LogLevel: proto.Int32(4),
LogFile: proto.String("restore.log"),
RstSibling: proto.Bool(true),
Root: proto.String(root),
ManageCgroups: proto.Bool(true),
NotifyScripts: proto.Bool(true),
ShellJob: proto.Bool(criuOpts.ShellJob),
ExtUnixSk: proto.Bool(criuOpts.ExternalUnixConnections),
TcpEstablished: proto.Bool(criuOpts.TcpEstablished),
FileLocks: proto.Bool(criuOpts.FileLocks),
},
}
for _, m := range c.config.Mounts {
switch m.Device {
case "bind":
c.addCriuRestoreMount(req, m)
break
case "cgroup":
binds, err := getCgroupMounts(m)
if err != nil {
return err
}
for _, b := range binds {
c.addCriuRestoreMount(req, b)
}
break
}
}
for _, iface := range c.config.Networks {
switch iface.Type {
case "veth":
veth := new(criurpc.CriuVethPair)
veth.IfOut = proto.String(iface.HostInterfaceName)
veth.IfIn = proto.String(iface.Name)
req.Opts.Veths = append(req.Opts.Veths, veth)
break
case "loopback":
break
}
}
for _, i := range criuOpts.VethPairs {
veth := new(criurpc.CriuVethPair)
veth.IfOut = proto.String(i.HostInterfaceName)
veth.IfIn = proto.String(i.ContainerInterfaceName)
req.Opts.Veths = append(req.Opts.Veths, veth)
}
// append optional manage cgroups mode
if criuOpts.ManageCgroupsMode != 0 {
if err := c.checkCriuVersion("1.7"); err != nil {
return err
}
req.Opts.ManageCgroupsMode = proto.Uint32(uint32(criuOpts.ManageCgroupsMode))
}
var (
fds []string
fdJSON []byte
)
if fdJSON, err = ioutil.ReadFile(filepath.Join(criuOpts.ImagesDirectory, descriptorsFilename)); err != nil {
return err
}
if err := json.Unmarshal(fdJSON, &fds); err != nil {
return err
}
for i := range fds {
if s := fds[i]; strings.Contains(s, "pipe:") {
inheritFd := new(criurpc.InheritFd)
inheritFd.Key = proto.String(s)
inheritFd.Fd = proto.Int32(int32(i))
req.Opts.InheritFd = append(req.Opts.InheritFd, inheritFd)
}
}
return c.criuSwrk(process, req, criuOpts, true)
}
func (c *linuxContainer) criuApplyCgroups(pid int, req *criurpc.CriuReq) error {
if err := c.cgroupManager.Apply(pid); err != nil {
return err
}
path := fmt.Sprintf("/proc/%d/cgroup", pid)
cgroupsPaths, err := cgroups.ParseCgroupFile(path)
if err != nil {
return err
}
for c, p := range cgroupsPaths {
cgroupRoot := &criurpc.CgroupRoot{
Ctrl: proto.String(c),
Path: proto.String(p),
}
req.Opts.CgRoot = append(req.Opts.CgRoot, cgroupRoot)
}
return nil
}
func (c *linuxContainer) criuSwrk(process *Process, req *criurpc.CriuReq, opts *CriuOpts, applyCgroups bool) error {
fds, err := syscall.Socketpair(syscall.AF_LOCAL, syscall.SOCK_SEQPACKET|syscall.SOCK_CLOEXEC, 0)
if err != nil {
return err
}
logPath := filepath.Join(opts.WorkDirectory, req.GetOpts().GetLogFile())
criuClient := os.NewFile(uintptr(fds[0]), "criu-transport-client")
criuServer := os.NewFile(uintptr(fds[1]), "criu-transport-server")
defer criuClient.Close()
defer criuServer.Close()
args := []string{"swrk", "3"}
logrus.Debugf("Using CRIU %d at: %s", c.criuVersion, c.criuPath)
logrus.Debugf("Using CRIU with following args: %s", args)
cmd := exec.Command(c.criuPath, args...)
if process != nil {
cmd.Stdin = process.Stdin
cmd.Stdout = process.Stdout
cmd.Stderr = process.Stderr
}
cmd.ExtraFiles = append(cmd.ExtraFiles, criuServer)
if err := cmd.Start(); err != nil {
return err
}
criuServer.Close()
defer func() {
criuClient.Close()
_, err := cmd.Process.Wait()
if err != nil {
return
}
}()
if applyCgroups {
err := c.criuApplyCgroups(cmd.Process.Pid, req)
if err != nil {
return err
}
}
var extFds []string
if process != nil {
extFds, err = getPipeFds(cmd.Process.Pid)
if err != nil {
return err
}
}
logrus.Debugf("Using CRIU in %s mode", req.GetType().String())
val := reflect.ValueOf(req.GetOpts())
v := reflect.Indirect(val)
for i := 0; i < v.NumField(); i++ {
st := v.Type()
name := st.Field(i).Name
if strings.HasPrefix(name, "XXX_") {
continue
}
value := val.MethodByName("Get" + name).Call([]reflect.Value{})
logrus.Debugf("CRIU option %s with value %v", name, value[0])
}
data, err := proto.Marshal(req)
if err != nil {
return err
}
_, err = criuClient.Write(data)
if err != nil {
return err
}
buf := make([]byte, 10*4096)
for true {
n, err := criuClient.Read(buf)
if err != nil {
return err
}
if n == 0 {
return fmt.Errorf("unexpected EOF")
}
if n == len(buf) {
return fmt.Errorf("buffer is too small")
}
resp := new(criurpc.CriuResp)
err = proto.Unmarshal(buf[:n], resp)
if err != nil {
return err
}
if !resp.GetSuccess() {
typeString := req.GetType().String()
return fmt.Errorf("criu failed: type %s errno %d\nlog file: %s", typeString, resp.GetCrErrno(), logPath)
}
t := resp.GetType()
switch {
case t == criurpc.CriuReqType_NOTIFY:
if err := c.criuNotifications(resp, process, opts, extFds); err != nil {
return err
}
t = criurpc.CriuReqType_NOTIFY
req = &criurpc.CriuReq{
Type: &t,
NotifySuccess: proto.Bool(true),
}
data, err = proto.Marshal(req)
if err != nil {
return err
}
n, err = criuClient.Write(data)
if err != nil {
return err
}
continue
case t == criurpc.CriuReqType_RESTORE:
case t == criurpc.CriuReqType_DUMP:
break
default:
return fmt.Errorf("unable to parse the response %s", resp.String())
}
break
}
// cmd.Wait() waits cmd.goroutines which are used for proxying file descriptors.
// Here we want to wait only the CRIU process.
st, err := cmd.Process.Wait()
if err != nil {
return err
}
if !st.Success() {
return fmt.Errorf("criu failed: %s\nlog file: %s", st.String(), logPath)
}
return nil
}
// block any external network activity
func lockNetwork(config *configs.Config) error {
for _, config := range config.Networks {
strategy, err := getStrategy(config.Type)
if err != nil {
return err
}
if err := strategy.detach(config); err != nil {
return err
}
}
return nil
}
func unlockNetwork(config *configs.Config) error {
for _, config := range config.Networks {
strategy, err := getStrategy(config.Type)
if err != nil {
return err
}
if err = strategy.attach(config); err != nil {
return err
}
}
return nil
}
func (c *linuxContainer) criuNotifications(resp *criurpc.CriuResp, process *Process, opts *CriuOpts, fds []string) error {
notify := resp.GetNotify()
if notify == nil {
return fmt.Errorf("invalid response: %s", resp.String())
}
switch {
case notify.GetScript() == "post-dump":
f, err := os.Create(filepath.Join(c.root, "checkpoint"))
if err != nil {
return err
}
f.Close()
case notify.GetScript() == "network-unlock":
if err := unlockNetwork(c.config); err != nil {
return err
}
case notify.GetScript() == "network-lock":
if err := lockNetwork(c.config); err != nil {
return err
}
case notify.GetScript() == "post-restore":
pid := notify.GetPid()
r, err := newRestoredProcess(int(pid), fds)
if err != nil {
return err
}
process.ops = r
if err := c.state.transition(&restoredState{
imageDir: opts.ImagesDirectory,
c: c,
}); err != nil {
return err
}
if err := c.updateState(r); err != nil {
return err
}
if err := os.Remove(filepath.Join(c.root, "checkpoint")); err != nil {
if !os.IsNotExist(err) {
logrus.Error(err)
}
}
}
return nil
}
func (c *linuxContainer) updateState(process parentProcess) error {
c.initProcess = process
state, err := c.currentState()
if err != nil {
return err
}
return c.saveState(state)
}
func (c *linuxContainer) saveState(s *State) error {
f, err := os.Create(filepath.Join(c.root, stateFilename))
if err != nil {
return err
}
defer f.Close()
return utils.WriteJSON(f, s)
}
func (c *linuxContainer) deleteState() error {
return os.Remove(filepath.Join(c.root, stateFilename))
}
func (c *linuxContainer) currentStatus() (Status, error) {
if err := c.refreshState(); err != nil {
return -1, err
}
return c.state.status(), nil
}
// refreshState needs to be called to verify that the current state on the
// container is what is true. Because consumers of libcontainer can use it
// out of process we need to verify the container's status based on runtime
// information and not rely on our in process info.
func (c *linuxContainer) refreshState() error {
paused, err := c.isPaused()
if err != nil {
return err
}
if paused {
return c.state.transition(&pausedState{c: c})
}
running, err := c.isRunning()
if err != nil {
return err
}
if running {
return c.state.transition(&runningState{c: c})
}
return c.state.transition(&stoppedState{c: c})
}
func (c *linuxContainer) isRunning() (bool, error) {
if c.initProcess == nil {
return false, nil
}
// return Running if the init process is alive
if err := syscall.Kill(c.initProcess.pid(), 0); err != nil {
if err == syscall.ESRCH {
return false, nil
}
return false, newSystemError(err)
}
return true, nil
}
func (c *linuxContainer) isPaused() (bool, error) {
data, err := ioutil.ReadFile(filepath.Join(c.cgroupManager.GetPaths()["freezer"], "freezer.state"))
if err != nil {
if os.IsNotExist(err) {
return false, nil
}
return false, newSystemError(err)
}
return bytes.Equal(bytes.TrimSpace(data), []byte("FROZEN")), nil
}
func (c *linuxContainer) currentState() (*State, error) {
var (
startTime string
externalDescriptors []string
pid = -1
)
if c.initProcess != nil {
pid = c.initProcess.pid()
startTime, _ = c.initProcess.startTime()
externalDescriptors = c.initProcess.externalDescriptors()
}
state := &State{
BaseState: BaseState{
ID: c.ID(),
Config: *c.config,
InitProcessPid: pid,
InitProcessStartTime: startTime,
Created: c.created,
},
CgroupPaths: c.cgroupManager.GetPaths(),
NamespacePaths: make(map[configs.NamespaceType]string),
ExternalDescriptors: externalDescriptors,
}
if pid > 0 {
for _, ns := range c.config.Namespaces {
state.NamespacePaths[ns.Type] = ns.GetPath(pid)
}
for _, nsType := range configs.NamespaceTypes() {
if _, ok := state.NamespacePaths[nsType]; !ok {
ns := configs.Namespace{Type: nsType}
state.NamespacePaths[ns.Type] = ns.GetPath(pid)
}
}
}
return state, nil
}
// bootstrapData encodes the necessary data in netlink binary format as a io.Reader.
// Consumer can write the data to a bootstrap program such as one that uses
// nsenter package to bootstrap the container's init process correctly, i.e. with
// correct namespaces, uid/gid mapping etc.
func (c *linuxContainer) bootstrapData(cloneFlags uintptr, pid int, consolePath string) (io.Reader, error) {
// create the netlink message
r := nl.NewNetlinkRequest(int(InitMsg), 0)
// write pid
r.AddData(&Int32msg{
Type: PidAttr,
Value: uint32(pid),
})
// write console path
if consolePath != "" {
r.AddData(&Bytemsg{
Type: ConsolePathAttr,
Value: []byte(consolePath),
})
}
return bytes.NewReader(r.Serialize()), nil
}
|