forked from adamlaska/moby
-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathdaemon_unix.go
More file actions
1738 lines (1568 loc) · 62.7 KB
/
daemon_unix.go
File metadata and controls
1738 lines (1568 loc) · 62.7 KB
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
//go:build linux || freebsd
// +build linux freebsd
package daemon // import "github.com/docker/docker/daemon"
import (
"bufio"
"context"
"fmt"
"net"
"os"
"path/filepath"
"runtime"
"runtime/debug"
"strconv"
"strings"
"sync"
"time"
"github.com/containerd/cgroups"
statsV1 "github.com/containerd/cgroups/stats/v1"
statsV2 "github.com/containerd/cgroups/v2/stats"
"github.com/containerd/containerd/pkg/userns"
"github.com/docker/docker/api/types"
"github.com/docker/docker/api/types/blkiodev"
pblkiodev "github.com/docker/docker/api/types/blkiodev"
containertypes "github.com/docker/docker/api/types/container"
"github.com/docker/docker/container"
"github.com/docker/docker/daemon/config"
"github.com/docker/docker/daemon/initlayer"
"github.com/docker/docker/errdefs"
"github.com/docker/docker/libcontainerd/remote"
"github.com/docker/docker/libnetwork"
nwconfig "github.com/docker/docker/libnetwork/config"
"github.com/docker/docker/libnetwork/drivers/bridge"
"github.com/docker/docker/libnetwork/netlabel"
"github.com/docker/docker/libnetwork/netutils"
"github.com/docker/docker/libnetwork/options"
lntypes "github.com/docker/docker/libnetwork/types"
"github.com/docker/docker/opts"
"github.com/docker/docker/pkg/containerfs"
"github.com/docker/docker/pkg/idtools"
"github.com/docker/docker/pkg/parsers"
"github.com/docker/docker/pkg/parsers/kernel"
"github.com/docker/docker/pkg/sysinfo"
"github.com/docker/docker/runconfig"
volumemounts "github.com/docker/docker/volume/mounts"
"github.com/moby/sys/mount"
specs "github.com/opencontainers/runtime-spec/specs-go"
"github.com/opencontainers/selinux/go-selinux"
"github.com/opencontainers/selinux/go-selinux/label"
"github.com/pkg/errors"
"github.com/sirupsen/logrus"
"github.com/vishvananda/netlink"
"golang.org/x/sys/unix"
)
const (
isWindows = false
// See https://git.kernel.org/cgit/linux/kernel/git/tip/tip.git/tree/kernel/sched/sched.h?id=8cd9234c64c584432f6992fe944ca9e46ca8ea76#n269
linuxMinCPUShares = 2
linuxMaxCPUShares = 262144
platformSupported = true
// It's not kernel limit, we want this 6M limit to account for overhead during startup, and to supply a reasonable functional container
linuxMinMemory = 6291456
// constants for remapped root settings
defaultIDSpecifier = "default"
defaultRemappedID = "dockremap"
// constant for cgroup drivers
cgroupFsDriver = "cgroupfs"
cgroupSystemdDriver = "systemd"
cgroupNoneDriver = "none"
)
type containerGetter interface {
GetContainer(string) (*container.Container, error)
}
func getMemoryResources(config containertypes.Resources) *specs.LinuxMemory {
memory := specs.LinuxMemory{}
if config.Memory > 0 {
memory.Limit = &config.Memory
}
if config.MemoryReservation > 0 {
memory.Reservation = &config.MemoryReservation
}
if config.MemorySwap > 0 {
memory.Swap = &config.MemorySwap
}
if config.MemorySwappiness != nil {
swappiness := uint64(*config.MemorySwappiness)
memory.Swappiness = &swappiness
}
if config.OomKillDisable != nil {
memory.DisableOOMKiller = config.OomKillDisable
}
if config.KernelMemory != 0 {
memory.Kernel = &config.KernelMemory
}
if config.KernelMemoryTCP != 0 {
memory.KernelTCP = &config.KernelMemoryTCP
}
return &memory
}
func getPidsLimit(config containertypes.Resources) *specs.LinuxPids {
if config.PidsLimit == nil {
return nil
}
if *config.PidsLimit <= 0 {
// docker API allows 0 and negative values to unset this to be consistent
// with default values. When updating values, runc requires -1 to unset
// the previous limit.
return &specs.LinuxPids{Limit: -1}
}
return &specs.LinuxPids{Limit: *config.PidsLimit}
}
func getCPUResources(config containertypes.Resources) (*specs.LinuxCPU, error) {
cpu := specs.LinuxCPU{}
if config.CPUShares < 0 {
return nil, fmt.Errorf("shares: invalid argument")
}
if config.CPUShares >= 0 {
shares := uint64(config.CPUShares)
cpu.Shares = &shares
}
if config.CpusetCpus != "" {
cpu.Cpus = config.CpusetCpus
}
if config.CpusetMems != "" {
cpu.Mems = config.CpusetMems
}
if config.NanoCPUs > 0 {
// https://www.kernel.org/doc/Documentation/scheduler/sched-bwc.txt
period := uint64(100 * time.Millisecond / time.Microsecond)
quota := config.NanoCPUs * int64(period) / 1e9
cpu.Period = &period
cpu.Quota = "a
}
if config.CPUPeriod != 0 {
period := uint64(config.CPUPeriod)
cpu.Period = &period
}
if config.CPUQuota != 0 {
q := config.CPUQuota
cpu.Quota = &q
}
if config.CPURealtimePeriod != 0 {
period := uint64(config.CPURealtimePeriod)
cpu.RealtimePeriod = &period
}
if config.CPURealtimeRuntime != 0 {
c := config.CPURealtimeRuntime
cpu.RealtimeRuntime = &c
}
return &cpu, nil
}
func getBlkioWeightDevices(config containertypes.Resources) ([]specs.LinuxWeightDevice, error) {
var stat unix.Stat_t
var blkioWeightDevices []specs.LinuxWeightDevice
for _, weightDevice := range config.BlkioWeightDevice {
if err := unix.Stat(weightDevice.Path, &stat); err != nil {
return nil, errors.WithStack(&os.PathError{Op: "stat", Path: weightDevice.Path, Err: err})
}
weight := weightDevice.Weight
d := specs.LinuxWeightDevice{Weight: &weight}
// The type is 32bit on mips.
d.Major = int64(unix.Major(uint64(stat.Rdev))) //nolint: unconvert
d.Minor = int64(unix.Minor(uint64(stat.Rdev))) //nolint: unconvert
blkioWeightDevices = append(blkioWeightDevices, d)
}
return blkioWeightDevices, nil
}
func (daemon *Daemon) parseSecurityOpt(container *container.Container, hostConfig *containertypes.HostConfig) error {
container.NoNewPrivileges = daemon.configStore.NoNewPrivileges
return parseSecurityOpt(container, hostConfig)
}
func parseSecurityOpt(container *container.Container, config *containertypes.HostConfig) error {
var (
labelOpts []string
err error
)
for _, opt := range config.SecurityOpt {
if opt == "no-new-privileges" {
container.NoNewPrivileges = true
continue
}
if opt == "disable" {
labelOpts = append(labelOpts, "disable")
continue
}
var con []string
if strings.Contains(opt, "=") {
con = strings.SplitN(opt, "=", 2)
} else if strings.Contains(opt, ":") {
con = strings.SplitN(opt, ":", 2)
logrus.Warn("Security options with `:` as a separator are deprecated and will be completely unsupported in 17.04, use `=` instead.")
}
if len(con) != 2 {
return fmt.Errorf("invalid --security-opt 1: %q", opt)
}
switch con[0] {
case "label":
labelOpts = append(labelOpts, con[1])
case "apparmor":
container.AppArmorProfile = con[1]
case "seccomp":
container.SeccompProfile = con[1]
case "no-new-privileges":
noNewPrivileges, err := strconv.ParseBool(con[1])
if err != nil {
return fmt.Errorf("invalid --security-opt 2: %q", opt)
}
container.NoNewPrivileges = noNewPrivileges
default:
return fmt.Errorf("invalid --security-opt 2: %q", opt)
}
}
container.ProcessLabel, container.MountLabel, err = label.InitLabels(labelOpts)
return err
}
func getBlkioThrottleDevices(devs []*blkiodev.ThrottleDevice) ([]specs.LinuxThrottleDevice, error) {
var throttleDevices []specs.LinuxThrottleDevice
var stat unix.Stat_t
for _, d := range devs {
if err := unix.Stat(d.Path, &stat); err != nil {
return nil, errors.WithStack(&os.PathError{Op: "stat", Path: d.Path, Err: err})
}
d := specs.LinuxThrottleDevice{Rate: d.Rate}
// the type is 32bit on mips
d.Major = int64(unix.Major(uint64(stat.Rdev))) //nolint: unconvert
d.Minor = int64(unix.Minor(uint64(stat.Rdev))) //nolint: unconvert
throttleDevices = append(throttleDevices, d)
}
return throttleDevices, nil
}
// adjustParallelLimit takes a number of objects and a proposed limit and
// figures out if it's reasonable (and adjusts it accordingly). This is only
// used for daemon startup, which does a lot of parallel loading of containers
// (and if we exceed RLIMIT_NOFILE then we're in trouble).
func adjustParallelLimit(n int, limit int) int {
// Rule-of-thumb overhead factor (how many files will each goroutine open
// simultaneously). Yes, this is ugly but to be frank this whole thing is
// ugly.
const overhead = 2
// On Linux, we need to ensure that parallelStartupJobs doesn't cause us to
// exceed RLIMIT_NOFILE. If parallelStartupJobs is too large, we reduce it
// and give a warning (since in theory the user should increase their
// ulimits to the largest possible value for dockerd).
var rlim unix.Rlimit
if err := unix.Getrlimit(unix.RLIMIT_NOFILE, &rlim); err != nil {
logrus.Warnf("Couldn't find dockerd's RLIMIT_NOFILE to double-check startup parallelism factor: %v", err)
return limit
}
softRlimit := int(rlim.Cur)
// Much fewer containers than RLIMIT_NOFILE. No need to adjust anything.
if softRlimit > overhead*n {
return limit
}
// RLIMIT_NOFILE big enough, no need to adjust anything.
if softRlimit > overhead*limit {
return limit
}
logrus.Warnf("Found dockerd's open file ulimit (%v) is far too small -- consider increasing it significantly (at least %v)", softRlimit, overhead*limit)
return softRlimit / overhead
}
// adaptContainerSettings is called during container creation to modify any
// settings necessary in the HostConfig structure.
func (daemon *Daemon) adaptContainerSettings(hostConfig *containertypes.HostConfig, adjustCPUShares bool) error {
if adjustCPUShares && hostConfig.CPUShares > 0 {
// Handle unsupported CPUShares
if hostConfig.CPUShares < linuxMinCPUShares {
logrus.Warnf("Changing requested CPUShares of %d to minimum allowed of %d", hostConfig.CPUShares, linuxMinCPUShares)
hostConfig.CPUShares = linuxMinCPUShares
} else if hostConfig.CPUShares > linuxMaxCPUShares {
logrus.Warnf("Changing requested CPUShares of %d to maximum allowed of %d", hostConfig.CPUShares, linuxMaxCPUShares)
hostConfig.CPUShares = linuxMaxCPUShares
}
}
if hostConfig.Memory > 0 && hostConfig.MemorySwap == 0 {
// By default, MemorySwap is set to twice the size of Memory.
hostConfig.MemorySwap = hostConfig.Memory * 2
}
if hostConfig.ShmSize == 0 {
hostConfig.ShmSize = config.DefaultShmSize
if daemon.configStore != nil {
hostConfig.ShmSize = int64(daemon.configStore.ShmSize)
}
}
// Set default IPC mode, if unset for container
if hostConfig.IpcMode.IsEmpty() {
m := config.DefaultIpcMode
if daemon.configStore != nil {
m = containertypes.IpcMode(daemon.configStore.IpcMode)
}
hostConfig.IpcMode = m
}
// Set default cgroup namespace mode, if unset for container
if hostConfig.CgroupnsMode.IsEmpty() {
// for cgroup v2: unshare cgroupns even for privileged containers
// https://github.com/containers/libpod/pull/4374#issuecomment-549776387
if hostConfig.Privileged && cgroups.Mode() != cgroups.Unified {
hostConfig.CgroupnsMode = containertypes.CgroupnsModeHost
} else {
m := containertypes.CgroupnsModeHost
if cgroups.Mode() == cgroups.Unified {
m = containertypes.CgroupnsModePrivate
}
if daemon.configStore != nil {
m = containertypes.CgroupnsMode(daemon.configStore.CgroupNamespaceMode)
}
hostConfig.CgroupnsMode = m
}
}
adaptSharedNamespaceContainer(daemon, hostConfig)
var err error
secOpts, err := daemon.generateSecurityOpt(hostConfig)
if err != nil {
return err
}
hostConfig.SecurityOpt = append(hostConfig.SecurityOpt, secOpts...)
if hostConfig.OomKillDisable == nil {
defaultOomKillDisable := false
hostConfig.OomKillDisable = &defaultOomKillDisable
}
return nil
}
// adaptSharedNamespaceContainer replaces container name with its ID in hostConfig.
// To be more precisely, it modifies `container:name` to `container:ID` of PidMode, IpcMode
// and NetworkMode.
//
// When a container shares its namespace with another container, use ID can keep the namespace
// sharing connection between the two containers even the another container is renamed.
func adaptSharedNamespaceContainer(daemon containerGetter, hostConfig *containertypes.HostConfig) {
containerPrefix := "container:"
if hostConfig.PidMode.IsContainer() {
pidContainer := hostConfig.PidMode.Container()
// if there is any error returned here, we just ignore it and leave it to be
// handled in the following logic
if c, err := daemon.GetContainer(pidContainer); err == nil {
hostConfig.PidMode = containertypes.PidMode(containerPrefix + c.ID)
}
}
if hostConfig.IpcMode.IsContainer() {
ipcContainer := hostConfig.IpcMode.Container()
if c, err := daemon.GetContainer(ipcContainer); err == nil {
hostConfig.IpcMode = containertypes.IpcMode(containerPrefix + c.ID)
}
}
if hostConfig.NetworkMode.IsContainer() {
netContainer := hostConfig.NetworkMode.ConnectedContainer()
if c, err := daemon.GetContainer(netContainer); err == nil {
hostConfig.NetworkMode = containertypes.NetworkMode(containerPrefix + c.ID)
}
}
}
// verifyPlatformContainerResources performs platform-specific validation of the container's resource-configuration
func verifyPlatformContainerResources(resources *containertypes.Resources, sysInfo *sysinfo.SysInfo, update bool) (warnings []string, err error) {
fixMemorySwappiness(resources)
// memory subsystem checks and adjustments
if resources.Memory != 0 && resources.Memory < linuxMinMemory {
return warnings, fmt.Errorf("Minimum memory limit allowed is 6MB")
}
if resources.Memory > 0 && !sysInfo.MemoryLimit {
warnings = append(warnings, "Your kernel does not support memory limit capabilities or the cgroup is not mounted. Limitation discarded.")
resources.Memory = 0
resources.MemorySwap = -1
}
if resources.Memory > 0 && resources.MemorySwap != -1 && !sysInfo.SwapLimit {
warnings = append(warnings, "Your kernel does not support swap limit capabilities or the cgroup is not mounted. Memory limited without swap.")
resources.MemorySwap = -1
}
if resources.Memory > 0 && resources.MemorySwap > 0 && resources.MemorySwap < resources.Memory {
return warnings, fmt.Errorf("Minimum memoryswap limit should be larger than memory limit, see usage")
}
if resources.Memory == 0 && resources.MemorySwap > 0 && !update {
return warnings, fmt.Errorf("You should always set the Memory limit when using Memoryswap limit, see usage")
}
if resources.MemorySwappiness != nil && !sysInfo.MemorySwappiness {
warnings = append(warnings, "Your kernel does not support memory swappiness capabilities or the cgroup is not mounted. Memory swappiness discarded.")
resources.MemorySwappiness = nil
}
if resources.MemorySwappiness != nil {
swappiness := *resources.MemorySwappiness
if swappiness < 0 || swappiness > 100 {
return warnings, fmt.Errorf("Invalid value: %v, valid memory swappiness range is 0-100", swappiness)
}
}
if resources.MemoryReservation > 0 && !sysInfo.MemoryReservation {
warnings = append(warnings, "Your kernel does not support memory soft limit capabilities or the cgroup is not mounted. Limitation discarded.")
resources.MemoryReservation = 0
}
if resources.MemoryReservation > 0 && resources.MemoryReservation < linuxMinMemory {
return warnings, fmt.Errorf("Minimum memory reservation allowed is 6MB")
}
if resources.Memory > 0 && resources.MemoryReservation > 0 && resources.Memory < resources.MemoryReservation {
return warnings, fmt.Errorf("Minimum memory limit can not be less than memory reservation limit, see usage")
}
if resources.KernelMemory > 0 {
// Kernel memory limit is not supported on cgroup v2.
// Even on cgroup v1, kernel memory limit (`kmem.limit_in_bytes`) has been deprecated since kernel 5.4.
// https://github.com/torvalds/linux/commit/0158115f702b0ba208ab0b5adf44cae99b3ebcc7
if !sysInfo.KernelMemory {
warnings = append(warnings, "Your kernel does not support kernel memory limit capabilities or the cgroup is not mounted. Limitation discarded.")
resources.KernelMemory = 0
}
if resources.KernelMemory > 0 && resources.KernelMemory < linuxMinMemory {
return warnings, fmt.Errorf("Minimum kernel memory limit allowed is 6MB")
}
if !kernel.CheckKernelVersion(4, 0, 0) {
warnings = append(warnings, "You specified a kernel memory limit on a kernel older than 4.0. Kernel memory limits are experimental on older kernels, it won't work as expected and can cause your system to be unstable.")
}
}
if resources.OomKillDisable != nil && !sysInfo.OomKillDisable {
// only produce warnings if the setting wasn't to *disable* the OOM Kill; no point
// warning the caller if they already wanted the feature to be off
if *resources.OomKillDisable {
warnings = append(warnings, "Your kernel does not support OomKillDisable. OomKillDisable discarded.")
}
resources.OomKillDisable = nil
}
if resources.OomKillDisable != nil && *resources.OomKillDisable && resources.Memory == 0 {
warnings = append(warnings, "OOM killer is disabled for the container, but no memory limit is set, this can result in the system running out of resources.")
}
if resources.PidsLimit != nil && !sysInfo.PidsLimit {
if *resources.PidsLimit > 0 {
warnings = append(warnings, "Your kernel does not support PIDs limit capabilities or the cgroup is not mounted. PIDs limit discarded.")
}
resources.PidsLimit = nil
}
// cpu subsystem checks and adjustments
if resources.NanoCPUs > 0 && resources.CPUPeriod > 0 {
return warnings, fmt.Errorf("Conflicting options: Nano CPUs and CPU Period cannot both be set")
}
if resources.NanoCPUs > 0 && resources.CPUQuota > 0 {
return warnings, fmt.Errorf("Conflicting options: Nano CPUs and CPU Quota cannot both be set")
}
if resources.NanoCPUs > 0 && !sysInfo.CPUCfs {
return warnings, fmt.Errorf("NanoCPUs can not be set, as your kernel does not support CPU CFS scheduler or the cgroup is not mounted")
}
// The highest precision we could get on Linux is 0.001, by setting
// cpu.cfs_period_us=1000ms
// cpu.cfs_quota=1ms
// See the following link for details:
// https://www.kernel.org/doc/Documentation/scheduler/sched-bwc.txt
// Here we don't set the lower limit and it is up to the underlying platform (e.g., Linux) to return an error.
// The error message is 0.01 so that this is consistent with Windows
if resources.NanoCPUs < 0 || resources.NanoCPUs > int64(sysinfo.NumCPU())*1e9 {
return warnings, fmt.Errorf("Range of CPUs is from 0.01 to %d.00, as there are only %d CPUs available", sysinfo.NumCPU(), sysinfo.NumCPU())
}
if resources.CPUShares > 0 && !sysInfo.CPUShares {
warnings = append(warnings, "Your kernel does not support CPU shares or the cgroup is not mounted. Shares discarded.")
resources.CPUShares = 0
}
if (resources.CPUPeriod != 0 || resources.CPUQuota != 0) && !sysInfo.CPUCfs {
warnings = append(warnings, "Your kernel does not support CPU CFS scheduler. CPU period/quota discarded.")
resources.CPUPeriod = 0
resources.CPUQuota = 0
}
if resources.CPUPeriod != 0 && (resources.CPUPeriod < 1000 || resources.CPUPeriod > 1000000) {
return warnings, fmt.Errorf("CPU cfs period can not be less than 1ms (i.e. 1000) or larger than 1s (i.e. 1000000)")
}
if resources.CPUQuota > 0 && resources.CPUQuota < 1000 {
return warnings, fmt.Errorf("CPU cfs quota can not be less than 1ms (i.e. 1000)")
}
if resources.CPUPercent > 0 {
warnings = append(warnings, fmt.Sprintf("%s does not support CPU percent. Percent discarded.", runtime.GOOS))
resources.CPUPercent = 0
}
// cpuset subsystem checks and adjustments
if (resources.CpusetCpus != "" || resources.CpusetMems != "") && !sysInfo.Cpuset {
warnings = append(warnings, "Your kernel does not support cpuset or the cgroup is not mounted. Cpuset discarded.")
resources.CpusetCpus = ""
resources.CpusetMems = ""
}
cpusAvailable, err := sysInfo.IsCpusetCpusAvailable(resources.CpusetCpus)
if err != nil {
return warnings, errors.Wrapf(err, "Invalid value %s for cpuset cpus", resources.CpusetCpus)
}
if !cpusAvailable {
return warnings, fmt.Errorf("Requested CPUs are not available - requested %s, available: %s", resources.CpusetCpus, sysInfo.Cpus)
}
memsAvailable, err := sysInfo.IsCpusetMemsAvailable(resources.CpusetMems)
if err != nil {
return warnings, errors.Wrapf(err, "Invalid value %s for cpuset mems", resources.CpusetMems)
}
if !memsAvailable {
return warnings, fmt.Errorf("Requested memory nodes are not available - requested %s, available: %s", resources.CpusetMems, sysInfo.Mems)
}
// blkio subsystem checks and adjustments
if resources.BlkioWeight > 0 && !sysInfo.BlkioWeight {
warnings = append(warnings, "Your kernel does not support Block I/O weight or the cgroup is not mounted. Weight discarded.")
resources.BlkioWeight = 0
}
if resources.BlkioWeight > 0 && (resources.BlkioWeight < 10 || resources.BlkioWeight > 1000) {
return warnings, fmt.Errorf("Range of blkio weight is from 10 to 1000")
}
if resources.IOMaximumBandwidth != 0 || resources.IOMaximumIOps != 0 {
return warnings, fmt.Errorf("Invalid QoS settings: %s does not support Maximum IO Bandwidth or Maximum IO IOps", runtime.GOOS)
}
if len(resources.BlkioWeightDevice) > 0 && !sysInfo.BlkioWeightDevice {
warnings = append(warnings, "Your kernel does not support Block I/O weight_device or the cgroup is not mounted. Weight-device discarded.")
resources.BlkioWeightDevice = []*pblkiodev.WeightDevice{}
}
if len(resources.BlkioDeviceReadBps) > 0 && !sysInfo.BlkioReadBpsDevice {
warnings = append(warnings, "Your kernel does not support BPS Block I/O read limit or the cgroup is not mounted. Block I/O BPS read limit discarded.")
resources.BlkioDeviceReadBps = []*pblkiodev.ThrottleDevice{}
}
if len(resources.BlkioDeviceWriteBps) > 0 && !sysInfo.BlkioWriteBpsDevice {
warnings = append(warnings, "Your kernel does not support BPS Block I/O write limit or the cgroup is not mounted. Block I/O BPS write limit discarded.")
resources.BlkioDeviceWriteBps = []*pblkiodev.ThrottleDevice{}
}
if len(resources.BlkioDeviceReadIOps) > 0 && !sysInfo.BlkioReadIOpsDevice {
warnings = append(warnings, "Your kernel does not support IOPS Block read limit or the cgroup is not mounted. Block I/O IOPS read limit discarded.")
resources.BlkioDeviceReadIOps = []*pblkiodev.ThrottleDevice{}
}
if len(resources.BlkioDeviceWriteIOps) > 0 && !sysInfo.BlkioWriteIOpsDevice {
warnings = append(warnings, "Your kernel does not support IOPS Block write limit or the cgroup is not mounted. Block I/O IOPS write limit discarded.")
resources.BlkioDeviceWriteIOps = []*pblkiodev.ThrottleDevice{}
}
return warnings, nil
}
func (daemon *Daemon) getCgroupDriver() string {
if UsingSystemd(daemon.configStore) {
return cgroupSystemdDriver
}
if daemon.Rootless() {
return cgroupNoneDriver
}
return cgroupFsDriver
}
// getCD gets the raw value of the native.cgroupdriver option, if set.
func getCD(config *config.Config) string {
for _, option := range config.ExecOptions {
key, val, err := parsers.ParseKeyValueOpt(option)
if err != nil || !strings.EqualFold(key, "native.cgroupdriver") {
continue
}
return val
}
return ""
}
// verifyCgroupDriver validates native.cgroupdriver
func verifyCgroupDriver(config *config.Config) error {
cd := getCD(config)
if cd == "" || cd == cgroupFsDriver || cd == cgroupSystemdDriver {
return nil
}
if cd == cgroupNoneDriver {
return fmt.Errorf("native.cgroupdriver option %s is internally used and cannot be specified manually", cd)
}
return fmt.Errorf("native.cgroupdriver option %s not supported", cd)
}
// UsingSystemd returns true if cli option includes native.cgroupdriver=systemd
func UsingSystemd(config *config.Config) bool {
cd := getCD(config)
if cd == cgroupSystemdDriver {
return true
}
// On cgroup v2 hosts, default to systemd driver
if cd == "" && cgroups.Mode() == cgroups.Unified && isRunningSystemd() {
return true
}
return false
}
var (
runningSystemd bool
detectSystemd sync.Once
)
// isRunningSystemd checks whether the host was booted with systemd as its init
// system. This functions similarly to systemd's `sd_booted(3)`: internally, it
// checks whether /run/systemd/system/ exists and is a directory.
// http://www.freedesktop.org/software/systemd/man/sd_booted.html
//
// NOTE: This function comes from package github.com/coreos/go-systemd/util
// It was borrowed here to avoid a dependency on cgo.
func isRunningSystemd() bool {
detectSystemd.Do(func() {
fi, err := os.Lstat("/run/systemd/system")
if err != nil {
return
}
runningSystemd = fi.IsDir()
})
return runningSystemd
}
// verifyPlatformContainerSettings performs platform-specific validation of the
// hostconfig and config structures.
func verifyPlatformContainerSettings(daemon *Daemon, hostConfig *containertypes.HostConfig, update bool) (warnings []string, err error) {
if hostConfig == nil {
return nil, nil
}
sysInfo := daemon.RawSysInfo()
w, err := verifyPlatformContainerResources(&hostConfig.Resources, sysInfo, update)
// no matter err is nil or not, w could have data in itself.
warnings = append(warnings, w...)
if err != nil {
return warnings, err
}
if hostConfig.ShmSize < 0 {
return warnings, fmt.Errorf("SHM size can not be less than 0")
}
if hostConfig.OomScoreAdj < -1000 || hostConfig.OomScoreAdj > 1000 {
return warnings, fmt.Errorf("Invalid value %d, range for oom score adj is [-1000, 1000]", hostConfig.OomScoreAdj)
}
// ip-forwarding does not affect container with '--net=host' (or '--net=none')
if sysInfo.IPv4ForwardingDisabled && !(hostConfig.NetworkMode.IsHost() || hostConfig.NetworkMode.IsNone()) {
warnings = append(warnings, "IPv4 forwarding is disabled. Networking will not work.")
}
if hostConfig.NetworkMode.IsHost() && len(hostConfig.PortBindings) > 0 {
warnings = append(warnings, "Published ports are discarded when using host network mode")
}
// check for various conflicting options with user namespaces
if daemon.configStore.RemappedRoot != "" && hostConfig.UsernsMode.IsPrivate() {
if hostConfig.Privileged {
return warnings, fmt.Errorf("privileged mode is incompatible with user namespaces. You must run the container in the host namespace when running privileged mode")
}
if hostConfig.NetworkMode.IsHost() && !hostConfig.UsernsMode.IsHost() {
return warnings, fmt.Errorf("cannot share the host's network namespace when user namespaces are enabled")
}
if hostConfig.PidMode.IsHost() && !hostConfig.UsernsMode.IsHost() {
return warnings, fmt.Errorf("cannot share the host PID namespace when user namespaces are enabled")
}
}
if hostConfig.CgroupParent != "" && UsingSystemd(daemon.configStore) {
// CgroupParent for systemd cgroup should be named as "xxx.slice"
if len(hostConfig.CgroupParent) <= 6 || !strings.HasSuffix(hostConfig.CgroupParent, ".slice") {
return warnings, fmt.Errorf("cgroup-parent for systemd cgroup should be a valid slice named as \"xxx.slice\"")
}
}
if hostConfig.Runtime == "" {
hostConfig.Runtime = daemon.configStore.GetDefaultRuntimeName()
}
if rt := daemon.configStore.GetRuntime(hostConfig.Runtime); rt == nil {
return warnings, fmt.Errorf("Unknown runtime specified %s", hostConfig.Runtime)
}
parser := volumemounts.NewParser()
for dest := range hostConfig.Tmpfs {
if err := parser.ValidateTmpfsMountDestination(dest); err != nil {
return warnings, err
}
}
if !hostConfig.CgroupnsMode.Valid() {
return warnings, fmt.Errorf("invalid cgroup namespace mode: %v", hostConfig.CgroupnsMode)
}
if hostConfig.CgroupnsMode.IsPrivate() {
if !sysInfo.CgroupNamespaces {
warnings = append(warnings, "Your kernel does not support cgroup namespaces. Cgroup namespace setting discarded.")
}
}
if hostConfig.Runtime == config.LinuxV1RuntimeName || (hostConfig.Runtime == "" && daemon.configStore.DefaultRuntime == config.LinuxV1RuntimeName) {
warnings = append(warnings, fmt.Sprintf("Configured runtime %q is deprecated and will be removed in the next release.", config.LinuxV1RuntimeName))
}
return warnings, nil
}
// verifyDaemonSettings performs validation of daemon config struct
func verifyDaemonSettings(conf *config.Config) error {
if conf.ContainerdNamespace == conf.ContainerdPluginNamespace {
return errors.New("containers namespace and plugins namespace cannot be the same")
}
// Check for mutually incompatible config options
if conf.BridgeConfig.Iface != "" && conf.BridgeConfig.IP != "" {
return fmt.Errorf("You specified -b & --bip, mutually exclusive options. Please specify only one")
}
if !conf.BridgeConfig.EnableIPTables && !conf.BridgeConfig.InterContainerCommunication {
return fmt.Errorf("You specified --iptables=false with --icc=false. ICC=false uses iptables to function. Please set --icc or --iptables to true")
}
if conf.BridgeConfig.EnableIP6Tables && !conf.Experimental {
return fmt.Errorf("ip6tables rules are only available if experimental features are enabled")
}
if !conf.BridgeConfig.EnableIPTables && conf.BridgeConfig.EnableIPMasq {
conf.BridgeConfig.EnableIPMasq = false
}
if err := verifyCgroupDriver(conf); err != nil {
return err
}
if conf.CgroupParent != "" && UsingSystemd(conf) {
if len(conf.CgroupParent) <= 6 || !strings.HasSuffix(conf.CgroupParent, ".slice") {
return fmt.Errorf("cgroup-parent for systemd cgroup should be a valid slice named as \"xxx.slice\"")
}
}
if conf.Rootless && UsingSystemd(conf) && cgroups.Mode() != cgroups.Unified {
return fmt.Errorf("exec-opt native.cgroupdriver=systemd requires cgroup v2 for rootless mode")
}
configureRuntimes(conf)
if rtName := conf.GetDefaultRuntimeName(); rtName != "" {
if conf.GetRuntime(rtName) == nil {
return fmt.Errorf("specified default runtime '%s' does not exist", rtName)
}
if rtName == config.LinuxV1RuntimeName {
logrus.Warnf("Configured default runtime %q is deprecated and will be removed in the next release.", config.LinuxV1RuntimeName)
}
}
return nil
}
// checkSystem validates platform-specific requirements
func checkSystem() error {
return nil
}
// configureMaxThreads sets the Go runtime max threads threshold
// which is 90% of the kernel setting from /proc/sys/kernel/threads-max
func configureMaxThreads(config *config.Config) error {
mt, err := os.ReadFile("/proc/sys/kernel/threads-max")
if err != nil {
return err
}
mtint, err := strconv.Atoi(strings.TrimSpace(string(mt)))
if err != nil {
return err
}
maxThreads := (mtint / 100) * 90
debug.SetMaxThreads(maxThreads)
logrus.Debugf("Golang's threads limit set to %d", maxThreads)
return nil
}
func overlaySupportsSelinux() (bool, error) {
f, err := os.Open("/proc/kallsyms")
if err != nil {
if os.IsNotExist(err) {
return false, nil
}
return false, err
}
defer f.Close()
s := bufio.NewScanner(f)
for s.Scan() {
if strings.HasSuffix(s.Text(), " security_inode_copy_up") {
return true, nil
}
}
return false, s.Err()
}
// configureKernelSecuritySupport configures and validates security support for the kernel
func configureKernelSecuritySupport(config *config.Config, driverName string) error {
if config.EnableSelinuxSupport {
if !selinux.GetEnabled() {
logrus.Warn("Docker could not enable SELinux on the host system")
return nil
}
if driverName == "overlay" || driverName == "overlay2" {
// If driver is overlay or overlay2, make sure kernel
// supports selinux with overlay.
supported, err := overlaySupportsSelinux()
if err != nil {
return err
}
if !supported {
logrus.Warnf("SELinux is not supported with the %v graph driver on this kernel", driverName)
}
}
} else {
selinux.SetDisabled()
}
return nil
}
func (daemon *Daemon) initNetworkController(config *config.Config, activeSandboxes map[string]interface{}) (libnetwork.NetworkController, error) {
netOptions, err := daemon.networkOptions(config, daemon.PluginStore, activeSandboxes)
if err != nil {
return nil, err
}
controller, err := libnetwork.New(netOptions...)
if err != nil {
return nil, fmt.Errorf("error obtaining controller instance: %v", err)
}
if len(activeSandboxes) > 0 {
logrus.Info("There are old running containers, the network config will not take affect")
setHostGatewayIP(daemon.configStore, controller)
return controller, nil
}
// Initialize default network on "null"
if n, _ := controller.NetworkByName("none"); n == nil {
if _, err := controller.NewNetwork("null", "none", "", libnetwork.NetworkOptionPersist(true)); err != nil {
return nil, fmt.Errorf("Error creating default \"null\" network: %v", err)
}
}
// Initialize default network on "host"
if n, _ := controller.NetworkByName("host"); n == nil {
if _, err := controller.NewNetwork("host", "host", "", libnetwork.NetworkOptionPersist(true)); err != nil {
return nil, fmt.Errorf("Error creating default \"host\" network: %v", err)
}
}
// Clear stale bridge network
if n, err := controller.NetworkByName("bridge"); err == nil {
if err = n.Delete(); err != nil {
return nil, fmt.Errorf("could not delete the default bridge network: %v", err)
}
if len(config.NetworkConfig.DefaultAddressPools.Value()) > 0 && !daemon.configStore.LiveRestoreEnabled {
removeDefaultBridgeInterface()
}
}
if !config.DisableBridge {
// Initialize default driver "bridge"
if err := initBridgeDriver(controller, config); err != nil {
return nil, err
}
} else {
removeDefaultBridgeInterface()
}
// Set HostGatewayIP to the default bridge's IP if it is empty
setHostGatewayIP(daemon.configStore, controller)
return controller, nil
}
// setHostGatewayIP sets cfg.HostGatewayIP to the default bridge's IP if it is empty.
func setHostGatewayIP(config *config.Config, controller libnetwork.NetworkController) {
if config.HostGatewayIP != nil {
return
}
if n, err := controller.NetworkByName("bridge"); err == nil {
v4Info, v6Info := n.Info().IpamInfo()
var gateway net.IP
if len(v4Info) > 0 {
gateway = v4Info[0].Gateway.IP
} else if len(v6Info) > 0 {
gateway = v6Info[0].Gateway.IP
}
config.HostGatewayIP = gateway
}
}
func driverOptions(config *config.Config) nwconfig.Option {
return nwconfig.OptionDriverConfig("bridge", options.Generic{
netlabel.GenericData: options.Generic{
"EnableIPForwarding": config.BridgeConfig.EnableIPForward,
"EnableIPTables": config.BridgeConfig.EnableIPTables,
"EnableIP6Tables": config.BridgeConfig.EnableIP6Tables,
"EnableUserlandProxy": config.BridgeConfig.EnableUserlandProxy,
"UserlandProxyPath": config.BridgeConfig.UserlandProxyPath,
},
})
}
func initBridgeDriver(controller libnetwork.NetworkController, config *config.Config) error {
bridgeName := bridge.DefaultBridgeName
if config.BridgeConfig.Iface != "" {
bridgeName = config.BridgeConfig.Iface
}
netOption := map[string]string{
bridge.BridgeName: bridgeName,
bridge.DefaultBridge: strconv.FormatBool(true),
netlabel.DriverMTU: strconv.Itoa(config.Mtu),
bridge.EnableIPMasquerade: strconv.FormatBool(config.BridgeConfig.EnableIPMasq),
bridge.EnableICC: strconv.FormatBool(config.BridgeConfig.InterContainerCommunication),
}
// --ip processing
if config.BridgeConfig.DefaultIP != nil {
netOption[bridge.DefaultBindingIP] = config.BridgeConfig.DefaultIP.String()
}
ipamV4Conf := &libnetwork.IpamConf{AuxAddresses: make(map[string]string)}
nwList, nw6List, err := netutils.ElectInterfaceAddresses(bridgeName)
if err != nil {
return errors.Wrap(err, "list bridge addresses failed")
}
nw := nwList[0]
if len(nwList) > 1 && config.BridgeConfig.FixedCIDR != "" {
_, fCIDR, err := net.ParseCIDR(config.BridgeConfig.FixedCIDR)
if err != nil {
return errors.Wrap(err, "parse CIDR failed")
}
// Iterate through in case there are multiple addresses for the bridge
for _, entry := range nwList {
if fCIDR.Contains(entry.IP) {
nw = entry
break
}
}
}
ipamV4Conf.PreferredPool = lntypes.GetIPNetCanonical(nw).String()
hip, _ := lntypes.GetHostPartIP(nw.IP, nw.Mask)
if hip.IsGlobalUnicast() {
ipamV4Conf.Gateway = nw.IP.String()
}
if config.BridgeConfig.IP != "" {
ip, ipNet, err := net.ParseCIDR(config.BridgeConfig.IP)
if err != nil {
return err
}
ipamV4Conf.PreferredPool = ipNet.String()
ipamV4Conf.Gateway = ip.String()
} else if bridgeName == bridge.DefaultBridgeName && ipamV4Conf.PreferredPool != "" {
logrus.Infof("Default bridge (%s) is assigned with an IP address %s. Daemon option --bip can be used to set a preferred IP address", bridgeName, ipamV4Conf.PreferredPool)
}
if config.BridgeConfig.FixedCIDR != "" {
_, fCIDR, err := net.ParseCIDR(config.BridgeConfig.FixedCIDR)
if err != nil {
return err
}
ipamV4Conf.SubPool = fCIDR.String()
}
if config.BridgeConfig.DefaultGatewayIPv4 != nil {
ipamV4Conf.AuxAddresses["DefaultGatewayIPv4"] = config.BridgeConfig.DefaultGatewayIPv4.String()
}
var (
deferIPv6Alloc bool
ipamV6Conf *libnetwork.IpamConf
)
if config.BridgeConfig.EnableIPv6 && config.BridgeConfig.FixedCIDRv6 == "" {
return errdefs.InvalidParameter(errors.New("IPv6 is enabled for the default bridge, but no subnet is configured. Specify an IPv6 subnet using --fixed-cidr-v6"))