Files
podman/pkg/api/handlers/compat/containers.go
Matthew Heon 997de2f8e9 Initial implementation of renaming containers
Basic theory: We remove the container, but *only from the DB*.
We leave it in c/storage, we leave the lock allocated, we leave
it running (if it is). Then we create an identical container with
an altered name, and add that back to the database. Theoretically
we now have a renamed container.

The advantage of this approach is that it doesn't just apply to
rename - we can use this to make *any* configuration change to a
container that does not alter its container ID.

Potential problems are numerous. This process is *THOROUGHLY*
non-atomic at present - if you `kill -9` Podman mid-rename things
will be in a bad place, for example. Also, we can't rename
containers that can't be removed normally - IE, containers with
dependencies (pod infra containers, for example).

The largest potential improvement will be to move the majority of
the work into the DB, with a `RecreateContainer()` method - that
will add atomicity, and let us remove the container without
worrying about depencies and similar issues.

Potential problems: long-running processes that edit the DB and
may have an older version of the configuration around. Most
notable example is `podman run --rm` - the removal command needed
to be manually edited to avoid this one. This begins to get at
the heart of me not wanting to do this in the first place...

This provides CLI and API implementations for frontend, but no
tunnel implementation. It will be added in a future release (just
held back for time now - we need this in 3.0 and are running low
on time).

This is honestly kind of horrifying, but I think it will work.

Signed-off-by: Matthew Heon <mheon@redhat.com>
2021-01-14 18:29:28 -05:00

499 lines
14 KiB
Go

package compat
import (
"encoding/json"
"fmt"
"net/http"
"sort"
"strconv"
"strings"
"syscall"
"time"
"github.com/containers/podman/v2/libpod"
"github.com/containers/podman/v2/libpod/define"
"github.com/containers/podman/v2/pkg/api/handlers"
"github.com/containers/podman/v2/pkg/api/handlers/utils"
"github.com/containers/podman/v2/pkg/domain/filters"
"github.com/containers/podman/v2/pkg/ps"
"github.com/containers/podman/v2/pkg/signal"
"github.com/docker/docker/api/types"
"github.com/docker/docker/api/types/container"
"github.com/docker/go-connections/nat"
"github.com/gorilla/mux"
"github.com/gorilla/schema"
"github.com/pkg/errors"
"github.com/sirupsen/logrus"
)
func RemoveContainer(w http.ResponseWriter, r *http.Request) {
decoder := r.Context().Value("decoder").(*schema.Decoder)
query := struct {
Force bool `schema:"force"`
Vols bool `schema:"v"`
Link bool `schema:"link"`
}{
// override any golang type defaults
}
if err := decoder.Decode(&query, r.URL.Query()); err != nil {
utils.Error(w, http.StatusText(http.StatusBadRequest), http.StatusBadRequest,
errors.Wrapf(err, "failed to parse parameters for %s", r.URL.String()))
return
}
if query.Link && !utils.IsLibpodRequest(r) {
utils.Error(w, http.StatusText(http.StatusBadRequest), http.StatusBadRequest,
utils.ErrLinkNotSupport)
return
}
runtime := r.Context().Value("runtime").(*libpod.Runtime)
name := utils.GetName(r)
con, err := runtime.LookupContainer(name)
if err != nil && errors.Cause(err) == define.ErrNoSuchCtr {
// Failed to get container. If force is specified, get the container's ID
// and evict it
if !query.Force {
utils.ContainerNotFound(w, name, err)
return
}
if _, err := runtime.EvictContainer(r.Context(), name, query.Vols); err != nil {
if errors.Cause(err) == define.ErrNoSuchCtr {
logrus.Debugf("Ignoring error (--allow-missing): %q", err)
w.WriteHeader(http.StatusNoContent)
return
}
logrus.Warn(errors.Wrapf(err, "failed to evict container: %q", name))
utils.InternalServerError(w, err)
return
}
w.WriteHeader(http.StatusNoContent)
return
}
if err := runtime.RemoveContainer(r.Context(), con, query.Force, query.Vols); err != nil {
utils.InternalServerError(w, err)
return
}
utils.WriteResponse(w, http.StatusNoContent, nil)
}
func ListContainers(w http.ResponseWriter, r *http.Request) {
runtime := r.Context().Value("runtime").(*libpod.Runtime)
decoder := r.Context().Value("decoder").(*schema.Decoder)
query := struct {
All bool `schema:"all"`
Limit int `schema:"limit"`
Size bool `schema:"size"`
Filters map[string][]string `schema:"filters"`
}{
// override any golang type defaults
}
if err := decoder.Decode(&query, r.URL.Query()); err != nil {
utils.Error(w, "Something went wrong.", http.StatusBadRequest, errors.Wrapf(err, "failed to parse parameters for %s", r.URL.String()))
return
}
filterFuncs := make([]libpod.ContainerFilter, 0, len(query.Filters))
all := query.All || query.Limit > 0
if len(query.Filters) > 0 {
for k, v := range query.Filters {
generatedFunc, err := filters.GenerateContainerFilterFuncs(k, v, runtime)
if err != nil {
utils.InternalServerError(w, err)
return
}
filterFuncs = append(filterFuncs, generatedFunc)
}
}
// Docker thinks that if status is given as an input, then we should override
// the all setting and always deal with all containers.
if len(query.Filters["status"]) > 0 {
all = true
}
if !all {
runningOnly, err := filters.GenerateContainerFilterFuncs("status", []string{define.ContainerStateRunning.String()}, runtime)
if err != nil {
utils.InternalServerError(w, err)
return
}
filterFuncs = append(filterFuncs, runningOnly)
}
containers, err := runtime.GetContainers(filterFuncs...)
if err != nil {
utils.InternalServerError(w, err)
return
}
if _, found := r.URL.Query()["limit"]; found && query.Limit > 0 {
// Sort the libpod containers
sort.Sort(ps.SortCreateTime{SortContainers: containers})
// we should perform the lopping before we start getting
// the expensive information on containers
if len(containers) > query.Limit {
containers = containers[:query.Limit]
}
}
var list = make([]*handlers.Container, len(containers))
for i, ctnr := range containers {
api, err := LibpodToContainer(ctnr, query.Size)
if err != nil {
utils.InternalServerError(w, err)
return
}
list[i] = api
}
utils.WriteResponse(w, http.StatusOK, list)
}
func GetContainer(w http.ResponseWriter, r *http.Request) {
runtime := r.Context().Value("runtime").(*libpod.Runtime)
decoder := r.Context().Value("decoder").(*schema.Decoder)
query := struct {
Size bool `schema:"size"`
}{
// override any golang type defaults
}
if err := decoder.Decode(&query, r.URL.Query()); err != nil {
utils.Error(w, "Something went wrong.", http.StatusBadRequest, errors.Wrapf(err, "failed to parse parameters for %s", r.URL.String()))
return
}
name := utils.GetName(r)
ctnr, err := runtime.LookupContainer(name)
if err != nil {
utils.ContainerNotFound(w, name, err)
return
}
api, err := LibpodToContainerJSON(ctnr, query.Size)
if err != nil {
utils.InternalServerError(w, err)
return
}
utils.WriteResponse(w, http.StatusOK, api)
}
func KillContainer(w http.ResponseWriter, r *http.Request) {
// /{version}/containers/(name)/kill
runtime := r.Context().Value("runtime").(*libpod.Runtime)
decoder := r.Context().Value("decoder").(*schema.Decoder)
query := struct {
Signal string `schema:"signal"`
}{
Signal: "KILL",
}
if err := decoder.Decode(&query, r.URL.Query()); err != nil {
utils.Error(w, "Something went wrong.", http.StatusBadRequest, errors.Wrapf(err, "failed to parse parameters for %s", r.URL.String()))
return
}
sig, err := signal.ParseSignalNameOrNumber(query.Signal)
if err != nil {
utils.InternalServerError(w, err)
return
}
name := utils.GetName(r)
con, err := runtime.LookupContainer(name)
if err != nil {
utils.ContainerNotFound(w, name, err)
return
}
state, err := con.State()
if err != nil {
utils.InternalServerError(w, err)
return
}
// If the Container is stopped already, send a 409
if state == define.ContainerStateStopped || state == define.ContainerStateExited {
utils.Error(w, fmt.Sprintf("Container %s is not running", name), http.StatusConflict, errors.New(fmt.Sprintf("Cannot kill Container %s, it is not running", name)))
return
}
signal := uint(sig)
err = con.Kill(signal)
if err != nil {
utils.Error(w, "Something went wrong.", http.StatusInternalServerError, errors.Wrapf(err, "unable to kill Container %s", name))
return
}
// Docker waits for the container to stop if the signal is 0 or
// SIGKILL.
if !utils.IsLibpodRequest(r) && (signal == 0 || syscall.Signal(signal) == syscall.SIGKILL) {
if _, err = con.Wait(); err != nil {
utils.Error(w, "Something went wrong.", http.StatusInternalServerError, errors.Wrapf(err, "failed to wait for Container %s", con.ID()))
return
}
}
// Success
utils.WriteResponse(w, http.StatusNoContent, nil)
}
func WaitContainer(w http.ResponseWriter, r *http.Request) {
var msg string
// /{version}/containers/(name)/wait
exitCode, err := utils.WaitContainer(w, r)
if err != nil {
logrus.Warnf("failed to wait on container %q: %v", mux.Vars(r)["name"], err)
return
}
utils.WriteResponse(w, http.StatusOK, handlers.ContainerWaitOKBody{
StatusCode: int(exitCode),
Error: struct {
Message string
}{
Message: msg,
},
})
}
func LibpodToContainer(l *libpod.Container, sz bool) (*handlers.Container, error) {
imageID, imageName := l.Image()
var (
err error
sizeRootFs int64
sizeRW int64
state define.ContainerStatus
)
if state, err = l.State(); err != nil {
return nil, err
}
stateStr := state.String()
if stateStr == "configured" {
stateStr = "created"
}
if sz {
if sizeRW, err = l.RWSize(); err != nil {
return nil, err
}
if sizeRootFs, err = l.RootFsSize(); err != nil {
return nil, err
}
}
return &handlers.Container{Container: types.Container{
ID: l.ID(),
Names: []string{fmt.Sprintf("/%s", l.Name())},
Image: imageName,
ImageID: imageID,
Command: strings.Join(l.Command(), " "),
Created: l.CreatedTime().Unix(),
Ports: nil,
SizeRw: sizeRW,
SizeRootFs: sizeRootFs,
Labels: l.Labels(),
State: stateStr,
Status: "",
HostConfig: struct {
NetworkMode string `json:",omitempty"`
}{
"host"},
NetworkSettings: nil,
Mounts: nil,
},
ContainerCreateConfig: types.ContainerCreateConfig{},
}, nil
}
func LibpodToContainerJSON(l *libpod.Container, sz bool) (*types.ContainerJSON, error) {
_, imageName := l.Image()
inspect, err := l.Inspect(sz)
if err != nil {
return nil, err
}
i, err := json.Marshal(inspect.State)
if err != nil {
return nil, err
}
state := types.ContainerState{}
if err := json.Unmarshal(i, &state); err != nil {
return nil, err
}
// docker considers paused to be running
if state.Paused {
state.Running = true
}
formatCapabilities(inspect.HostConfig.CapDrop)
formatCapabilities(inspect.HostConfig.CapAdd)
h, err := json.Marshal(inspect.HostConfig)
if err != nil {
return nil, err
}
hc := container.HostConfig{}
if err := json.Unmarshal(h, &hc); err != nil {
return nil, err
}
g, err := json.Marshal(inspect.GraphDriver)
if err != nil {
return nil, err
}
graphDriver := types.GraphDriverData{}
if err := json.Unmarshal(g, &graphDriver); err != nil {
return nil, err
}
cb := types.ContainerJSONBase{
ID: l.ID(),
Created: l.CreatedTime().Format(time.RFC3339Nano),
Path: inspect.Path,
Args: inspect.Args,
State: &state,
Image: imageName,
ResolvConfPath: inspect.ResolvConfPath,
HostnamePath: inspect.HostnamePath,
HostsPath: inspect.HostsPath,
LogPath: l.LogPath(),
Node: nil,
Name: fmt.Sprintf("/%s", l.Name()),
RestartCount: int(inspect.RestartCount),
Driver: inspect.Driver,
Platform: "linux",
MountLabel: inspect.MountLabel,
ProcessLabel: inspect.ProcessLabel,
AppArmorProfile: inspect.AppArmorProfile,
ExecIDs: inspect.ExecIDs,
HostConfig: &hc,
GraphDriver: graphDriver,
SizeRw: inspect.SizeRw,
SizeRootFs: &inspect.SizeRootFs,
}
// set Path and Args
processArgs := l.Config().Spec.Process.Args
if len(processArgs) > 0 {
cb.Path = processArgs[0]
}
if len(processArgs) > 1 {
cb.Args = processArgs[1:]
}
stopTimeout := int(l.StopTimeout())
exposedPorts := make(nat.PortSet)
for ep := range inspect.HostConfig.PortBindings {
splitp := strings.SplitN(ep, "/", 2)
if len(splitp) != 2 {
return nil, errors.Errorf("PORT/PROTOCOL Format required for %q", ep)
}
exposedPort, err := nat.NewPort(splitp[1], splitp[0])
if err != nil {
return nil, err
}
exposedPorts[exposedPort] = struct{}{}
}
config := container.Config{
Hostname: l.Hostname(),
Domainname: inspect.Config.DomainName,
User: l.User(),
AttachStdin: inspect.Config.AttachStdin,
AttachStdout: inspect.Config.AttachStdout,
AttachStderr: inspect.Config.AttachStderr,
ExposedPorts: exposedPorts,
Tty: inspect.Config.Tty,
OpenStdin: inspect.Config.OpenStdin,
StdinOnce: inspect.Config.StdinOnce,
Env: inspect.Config.Env,
Cmd: l.Command(),
Healthcheck: nil,
ArgsEscaped: false,
Image: imageName,
Volumes: nil,
WorkingDir: l.WorkingDir(),
Entrypoint: l.Entrypoint(),
NetworkDisabled: false,
MacAddress: "",
OnBuild: nil,
Labels: l.Labels(),
StopSignal: strconv.Itoa(int(l.StopSignal())),
StopTimeout: &stopTimeout,
Shell: nil,
}
m, err := json.Marshal(inspect.Mounts)
if err != nil {
return nil, err
}
mounts := []types.MountPoint{}
if err := json.Unmarshal(m, &mounts); err != nil {
return nil, err
}
p, err := json.Marshal(inspect.NetworkSettings.Ports)
if err != nil {
return nil, err
}
ports := nat.PortMap{}
if err := json.Unmarshal(p, &ports); err != nil {
return nil, err
}
n, err := json.Marshal(inspect.NetworkSettings)
if err != nil {
return nil, err
}
networkSettings := types.NetworkSettings{}
if err := json.Unmarshal(n, &networkSettings); err != nil {
return nil, err
}
c := types.ContainerJSON{
ContainerJSONBase: &cb,
Mounts: mounts,
Config: &config,
NetworkSettings: &networkSettings,
}
return &c, nil
}
func formatCapabilities(slice []string) {
for i := range slice {
slice[i] = strings.TrimPrefix(slice[i], "CAP_")
}
}
func RenameContainer(w http.ResponseWriter, r *http.Request) {
runtime := r.Context().Value("runtime").(*libpod.Runtime)
decoder := r.Context().Value("decoder").(*schema.Decoder)
name := utils.GetName(r)
query := struct {
Name string `schema:"name"`
}{}
if err := decoder.Decode(&query, r.URL.Query()); err != nil {
utils.Error(w, "Something went wrong.", http.StatusBadRequest, errors.Wrapf(err, "failed to parse parameters for %s", r.URL.String()))
return
}
ctr, err := runtime.LookupContainer(name)
if err != nil {
utils.ContainerNotFound(w, name, err)
return
}
if _, err := runtime.RenameContainer(r.Context(), ctr, query.Name); err != nil {
if errors.Cause(err) == define.ErrPodExists || errors.Cause(err) == define.ErrCtrExists {
utils.Error(w, "Something went wrong.", http.StatusConflict, err)
return
}
utils.InternalServerError(w, err)
return
}
utils.WriteResponse(w, http.StatusNoContent, nil)
}