moby/daemon/daemon_unix_test.go
Cory Snider d222bf097c daemon: reload runtimes w/o breaking containers
The existing runtimes reload logic went to great lengths to replace the
directory containing runtime wrapper scripts as atomically as possible
within the limitations of the Linux filesystem ABI. Trouble is,
atomically swapping the wrapper scripts directory solves the wrong
problem! The runtime configuration is "locked in" when a container is
started, including the path to the runC binary. If a container is
started with a runtime which requires a daemon-managed wrapper script
and then the daemon is reloaded with a config which no longer requires
the wrapper script (i.e. some args -> no args, or the runtime is dropped
from the config), that container would become unmanageable. Any attempts
to stop, exec or otherwise perform lifecycle management operations on
the container are likely to fail due to the wrapper script no longer
existing at its original path.

Atomically swapping the wrapper scripts is also incompatible with the
read-copy-update paradigm for reloading configuration. A handler in the
daemon could retain a reference to the pre-reload configuration for an
indeterminate amount of time after the daemon configuration has been
reloaded and updated. It is possible for the daemon to attempt to start
a container using a deleted wrapper script if a request to run a
container races a reload.

Solve the problem of deleting referenced wrapper scripts by ensuring
that all wrapper scripts are *immutable* for the lifetime of the daemon
process. Any given runtime wrapper script must always exist with the
same contents, no matter how many times the daemon config is reloaded,
or what changes are made to the config. This is accomplished by using
everyone's favourite design pattern: content-addressable storage. Each
wrapper script file name is suffixed with the SHA-256 digest of its
contents to (probabilistically) guarantee immutability without needing
any concurrency control. Stale runtime wrapper scripts are only cleaned
up on the next daemon restart.

Split the derived runtimes configuration from the user-supplied
configuration to have a place to store derived state without mutating
the user-supplied configuration or exposing daemon internals in API
struct types. Hold the derived state and the user-supplied configuration
in a single struct value so that they can be updated as an atomic unit.

Signed-off-by: Cory Snider <csnider@mirantis.com>
2023-06-01 14:45:25 -04:00

440 lines
13 KiB
Go

//go:build !windows
package daemon // import "github.com/docker/docker/daemon"
import (
"errors"
"os"
"path/filepath"
"testing"
"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/pkg/sysinfo"
"github.com/opencontainers/selinux/go-selinux"
"golang.org/x/sys/unix"
"gotest.tools/v3/assert"
is "gotest.tools/v3/assert/cmp"
)
type fakeContainerGetter struct {
containers map[string]*container.Container
}
func (f *fakeContainerGetter) GetContainer(cid string) (*container.Container, error) {
ctr, ok := f.containers[cid]
if !ok {
return nil, errors.New("container not found")
}
return ctr, nil
}
// Unix test as uses settings which are not available on Windows
func TestAdjustSharedNamespaceContainerName(t *testing.T) {
fakeID := "abcdef1234567890"
hostConfig := &containertypes.HostConfig{
IpcMode: containertypes.IpcMode("container:base"),
PidMode: containertypes.PidMode("container:base"),
NetworkMode: containertypes.NetworkMode("container:base"),
}
containerStore := &fakeContainerGetter{}
containerStore.containers = make(map[string]*container.Container)
containerStore.containers["base"] = &container.Container{
ID: fakeID,
}
adaptSharedNamespaceContainer(containerStore, hostConfig)
if hostConfig.IpcMode != containertypes.IpcMode("container:"+fakeID) {
t.Errorf("Expected IpcMode to be container:%s", fakeID)
}
if hostConfig.PidMode != containertypes.PidMode("container:"+fakeID) {
t.Errorf("Expected PidMode to be container:%s", fakeID)
}
if hostConfig.NetworkMode != containertypes.NetworkMode("container:"+fakeID) {
t.Errorf("Expected NetworkMode to be container:%s", fakeID)
}
}
// Unix test as uses settings which are not available on Windows
func TestAdjustCPUShares(t *testing.T) {
tmp, err := os.MkdirTemp("", "docker-daemon-unix-test-")
if err != nil {
t.Fatal(err)
}
defer os.RemoveAll(tmp)
daemon := &Daemon{
repository: tmp,
root: tmp,
}
cfg := &config.Config{}
muteLogs()
hostConfig := &containertypes.HostConfig{
Resources: containertypes.Resources{CPUShares: linuxMinCPUShares - 1},
}
daemon.adaptContainerSettings(cfg, hostConfig, true)
if hostConfig.CPUShares != linuxMinCPUShares {
t.Errorf("Expected CPUShares to be %d", linuxMinCPUShares)
}
hostConfig.CPUShares = linuxMaxCPUShares + 1
daemon.adaptContainerSettings(cfg, hostConfig, true)
if hostConfig.CPUShares != linuxMaxCPUShares {
t.Errorf("Expected CPUShares to be %d", linuxMaxCPUShares)
}
hostConfig.CPUShares = 0
daemon.adaptContainerSettings(cfg, hostConfig, true)
if hostConfig.CPUShares != 0 {
t.Error("Expected CPUShares to be unchanged")
}
hostConfig.CPUShares = 1024
daemon.adaptContainerSettings(cfg, hostConfig, true)
if hostConfig.CPUShares != 1024 {
t.Error("Expected CPUShares to be unchanged")
}
}
// Unix test as uses settings which are not available on Windows
func TestAdjustCPUSharesNoAdjustment(t *testing.T) {
tmp, err := os.MkdirTemp("", "docker-daemon-unix-test-")
if err != nil {
t.Fatal(err)
}
defer os.RemoveAll(tmp)
daemon := &Daemon{
repository: tmp,
root: tmp,
}
cfg := &config.Config{}
hostConfig := &containertypes.HostConfig{
Resources: containertypes.Resources{CPUShares: linuxMinCPUShares - 1},
}
daemon.adaptContainerSettings(cfg, hostConfig, false)
if hostConfig.CPUShares != linuxMinCPUShares-1 {
t.Errorf("Expected CPUShares to be %d", linuxMinCPUShares-1)
}
hostConfig.CPUShares = linuxMaxCPUShares + 1
daemon.adaptContainerSettings(cfg, hostConfig, false)
if hostConfig.CPUShares != linuxMaxCPUShares+1 {
t.Errorf("Expected CPUShares to be %d", linuxMaxCPUShares+1)
}
hostConfig.CPUShares = 0
daemon.adaptContainerSettings(cfg, hostConfig, false)
if hostConfig.CPUShares != 0 {
t.Error("Expected CPUShares to be unchanged")
}
hostConfig.CPUShares = 1024
daemon.adaptContainerSettings(cfg, hostConfig, false)
if hostConfig.CPUShares != 1024 {
t.Error("Expected CPUShares to be unchanged")
}
}
// Unix test as uses settings which are not available on Windows
func TestParseSecurityOptWithDeprecatedColon(t *testing.T) {
opts := &container.SecurityOptions{}
cfg := &containertypes.HostConfig{}
// test apparmor
cfg.SecurityOpt = []string{"apparmor=test_profile"}
if err := parseSecurityOpt(opts, cfg); err != nil {
t.Fatalf("Unexpected parseSecurityOpt error: %v", err)
}
if opts.AppArmorProfile != "test_profile" {
t.Fatalf("Unexpected AppArmorProfile, expected: \"test_profile\", got %q", opts.AppArmorProfile)
}
// test seccomp
sp := "/path/to/seccomp_test.json"
cfg.SecurityOpt = []string{"seccomp=" + sp}
if err := parseSecurityOpt(opts, cfg); err != nil {
t.Fatalf("Unexpected parseSecurityOpt error: %v", err)
}
if opts.SeccompProfile != sp {
t.Fatalf("Unexpected AppArmorProfile, expected: %q, got %q", sp, opts.SeccompProfile)
}
// test valid label
cfg.SecurityOpt = []string{"label=user:USER"}
if err := parseSecurityOpt(opts, cfg); err != nil {
t.Fatalf("Unexpected parseSecurityOpt error: %v", err)
}
// test invalid label
cfg.SecurityOpt = []string{"label"}
if err := parseSecurityOpt(opts, cfg); err == nil {
t.Fatal("Expected parseSecurityOpt error, got nil")
}
// test invalid opt
cfg.SecurityOpt = []string{"test"}
if err := parseSecurityOpt(opts, cfg); err == nil {
t.Fatal("Expected parseSecurityOpt error, got nil")
}
}
func TestParseSecurityOpt(t *testing.T) {
t.Run("apparmor", func(t *testing.T) {
secOpts := &container.SecurityOptions{}
err := parseSecurityOpt(secOpts, &containertypes.HostConfig{
SecurityOpt: []string{"apparmor=test_profile"},
})
assert.Check(t, err)
assert.Equal(t, secOpts.AppArmorProfile, "test_profile")
})
t.Run("apparmor using legacy separator", func(t *testing.T) {
secOpts := &container.SecurityOptions{}
err := parseSecurityOpt(secOpts, &containertypes.HostConfig{
SecurityOpt: []string{"apparmor:test_profile"},
})
assert.Check(t, err)
assert.Equal(t, secOpts.AppArmorProfile, "test_profile")
})
t.Run("seccomp", func(t *testing.T) {
secOpts := &container.SecurityOptions{}
err := parseSecurityOpt(secOpts, &containertypes.HostConfig{
SecurityOpt: []string{"seccomp=/path/to/seccomp_test.json"},
})
assert.Check(t, err)
assert.Equal(t, secOpts.SeccompProfile, "/path/to/seccomp_test.json")
})
t.Run("valid label", func(t *testing.T) {
secOpts := &container.SecurityOptions{}
err := parseSecurityOpt(secOpts, &containertypes.HostConfig{
SecurityOpt: []string{"label=user:USER"},
})
assert.Check(t, err)
if selinux.GetEnabled() {
// TODO(thaJeztah): set expected labels here (or "partial" if depends on host)
// assert.Check(t, is.Equal(secOpts.MountLabel, ""))
// assert.Check(t, is.Equal(secOpts.ProcessLabel, ""))
} else {
assert.Check(t, is.Equal(secOpts.MountLabel, ""))
assert.Check(t, is.Equal(secOpts.ProcessLabel, ""))
}
})
t.Run("invalid label", func(t *testing.T) {
secOpts := &container.SecurityOptions{}
err := parseSecurityOpt(secOpts, &containertypes.HostConfig{
SecurityOpt: []string{"label"},
})
assert.Error(t, err, `invalid --security-opt 1: "label"`)
})
t.Run("invalid option (no value)", func(t *testing.T) {
secOpts := &container.SecurityOptions{}
err := parseSecurityOpt(secOpts, &containertypes.HostConfig{
SecurityOpt: []string{"unknown"},
})
assert.Error(t, err, `invalid --security-opt 1: "unknown"`)
})
t.Run("unknown option", func(t *testing.T) {
secOpts := &container.SecurityOptions{}
err := parseSecurityOpt(secOpts, &containertypes.HostConfig{
SecurityOpt: []string{"unknown=something"},
})
assert.Error(t, err, `invalid --security-opt 2: "unknown=something"`)
})
}
func TestParseNNPSecurityOptions(t *testing.T) {
daemonCfg := &configStore{Config: config.Config{NoNewPrivileges: true}}
daemon := &Daemon{}
daemon.configStore.Store(daemonCfg)
opts := &container.SecurityOptions{}
cfg := &containertypes.HostConfig{}
// test NNP when "daemon:true" and "no-new-privileges=false""
cfg.SecurityOpt = []string{"no-new-privileges=false"}
if err := daemon.parseSecurityOpt(&daemonCfg.Config, opts, cfg); err != nil {
t.Fatalf("Unexpected daemon.parseSecurityOpt error: %v", err)
}
if opts.NoNewPrivileges {
t.Fatalf("container.NoNewPrivileges should be FALSE: %v", opts.NoNewPrivileges)
}
// test NNP when "daemon:false" and "no-new-privileges=true""
daemonCfg.NoNewPrivileges = false
cfg.SecurityOpt = []string{"no-new-privileges=true"}
if err := daemon.parseSecurityOpt(&daemonCfg.Config, opts, cfg); err != nil {
t.Fatalf("Unexpected daemon.parseSecurityOpt error: %v", err)
}
if !opts.NoNewPrivileges {
t.Fatalf("container.NoNewPrivileges should be TRUE: %v", opts.NoNewPrivileges)
}
}
func TestVerifyPlatformContainerResources(t *testing.T) {
t.Parallel()
var (
no = false
yes = true
)
withMemoryLimit := func(si *sysinfo.SysInfo) {
si.MemoryLimit = true
}
withSwapLimit := func(si *sysinfo.SysInfo) {
si.SwapLimit = true
}
withOomKillDisable := func(si *sysinfo.SysInfo) {
si.OomKillDisable = true
}
tests := []struct {
name string
resources containertypes.Resources
sysInfo sysinfo.SysInfo
update bool
expectedWarnings []string
}{
{
name: "no-oom-kill-disable",
resources: containertypes.Resources{},
sysInfo: sysInfo(t, withMemoryLimit),
expectedWarnings: []string{},
},
{
name: "oom-kill-disable-disabled",
resources: containertypes.Resources{
OomKillDisable: &no,
},
sysInfo: sysInfo(t, withMemoryLimit),
expectedWarnings: []string{},
},
{
name: "oom-kill-disable-not-supported",
resources: containertypes.Resources{
OomKillDisable: &yes,
},
sysInfo: sysInfo(t, withMemoryLimit),
expectedWarnings: []string{
"Your kernel does not support OomKillDisable. OomKillDisable discarded.",
},
},
{
name: "oom-kill-disable-without-memory-constraints",
resources: containertypes.Resources{
OomKillDisable: &yes,
Memory: 0,
},
sysInfo: sysInfo(t, withMemoryLimit, withOomKillDisable, withSwapLimit),
expectedWarnings: []string{
"OOM killer is disabled for the container, but no memory limit is set, this can result in the system running out of resources.",
},
},
{
name: "oom-kill-disable-with-memory-constraints-but-no-memory-limit-support",
resources: containertypes.Resources{
OomKillDisable: &yes,
Memory: linuxMinMemory,
},
sysInfo: sysInfo(t, withOomKillDisable),
expectedWarnings: []string{
"Your kernel does not support memory limit capabilities or the cgroup is not mounted. Limitation discarded.",
"OOM killer is disabled for the container, but no memory limit is set, this can result in the system running out of resources.",
},
},
{
name: "oom-kill-disable-with-memory-constraints",
resources: containertypes.Resources{
OomKillDisable: &yes,
Memory: linuxMinMemory,
},
sysInfo: sysInfo(t, withMemoryLimit, withOomKillDisable, withSwapLimit),
expectedWarnings: []string{},
},
}
for _, tc := range tests {
tc := tc
t.Run(tc.name, func(t *testing.T) {
t.Parallel()
warnings, err := verifyPlatformContainerResources(&tc.resources, &tc.sysInfo, tc.update)
assert.NilError(t, err)
for _, w := range tc.expectedWarnings {
assert.Assert(t, is.Contains(warnings, w))
}
})
}
}
func sysInfo(t *testing.T, opts ...func(*sysinfo.SysInfo)) sysinfo.SysInfo {
t.Helper()
si := sysinfo.SysInfo{}
for _, opt := range opts {
opt(&si)
}
if si.OomKillDisable {
t.Log(t.Name(), "OOM disable supported")
}
return si
}
const (
// prepare major 0x1FD(509 in decimal) and minor 0x130(304)
DEVNO = 0x11FD30
MAJOR = 509
MINOR = 304
WEIGHT = 1024
)
func deviceTypeMock(t *testing.T, testAndCheck func(string)) {
if os.Getuid() != 0 {
t.Skip("root required") // for mknod
}
t.Parallel()
tempDir, err := os.MkdirTemp("", "tempDevDir"+t.Name())
assert.NilError(t, err, "create temp file")
tempFile := filepath.Join(tempDir, "dev")
defer os.RemoveAll(tempDir)
if err = unix.Mknod(tempFile, unix.S_IFCHR, DEVNO); err != nil {
t.Fatalf("mknod error %s(%x): %v", tempFile, DEVNO, err)
}
testAndCheck(tempFile)
}
func TestGetBlkioWeightDevices(t *testing.T) {
deviceTypeMock(t, func(tempFile string) {
mockResource := containertypes.Resources{
BlkioWeightDevice: []*blkiodev.WeightDevice{{Path: tempFile, Weight: WEIGHT}},
}
weightDevs, err := getBlkioWeightDevices(mockResource)
assert.NilError(t, err, "getBlkioWeightDevices")
assert.Check(t, is.Len(weightDevs, 1), "getBlkioWeightDevices")
assert.Check(t, weightDevs[0].Major == MAJOR, "get major device type")
assert.Check(t, weightDevs[0].Minor == MINOR, "get minor device type")
assert.Check(t, *weightDevs[0].Weight == WEIGHT, "get device weight")
})
}
func TestGetBlkioThrottleDevices(t *testing.T) {
deviceTypeMock(t, func(tempFile string) {
mockDevs := []*blkiodev.ThrottleDevice{{Path: tempFile, Rate: WEIGHT}}
retDevs, err := getBlkioThrottleDevices(mockDevs)
assert.NilError(t, err, "getBlkioThrottleDevices")
assert.Check(t, is.Len(retDevs, 1), "getBlkioThrottleDevices")
assert.Check(t, retDevs[0].Major == MAJOR, "get major device type")
assert.Check(t, retDevs[0].Minor == MINOR, "get minor device type")
assert.Check(t, retDevs[0].Rate == WEIGHT, "get device rate")
})
}