mirror of
https://github.com/netbirdio/gvisor.git
synced 2026-05-22 17:12:49 -07:00
Initial commit
This commit is contained in:
Ian Lewis
@@ -0,0 +1 @@
|
||||
/bin/
|
||||
@@ -0,0 +1,28 @@
|
||||
# How to Contribute
|
||||
|
||||
We'd love to accept your patches and contributions to this project. There are
|
||||
just a few small guidelines you need to follow.
|
||||
|
||||
## Contributor License Agreement
|
||||
|
||||
Contributions to this project must be accompanied by a Contributor License
|
||||
Agreement. You (or your employer) retain the copyright to your contribution;
|
||||
this simply gives us permission to use and redistribute your contributions as
|
||||
part of the project. Head over to <https://cla.developers.google.com/> to see
|
||||
your current agreements on file or to sign a new one.
|
||||
|
||||
You generally only need to submit a CLA once, so if you've already submitted one
|
||||
(even if it was for a different project), you probably don't need to do it
|
||||
again.
|
||||
|
||||
## Code reviews
|
||||
|
||||
All submissions, including submissions by project members, require review. We
|
||||
use GitHub pull requests for this purpose. Consult
|
||||
[GitHub Help](https://help.github.com/articles/about-pull-requests/) for more
|
||||
information on using pull requests.
|
||||
|
||||
## Community Guidelines
|
||||
|
||||
This project follows [Google's Open Source Community
|
||||
Guidelines](https://opensource.google.com/conduct/).
|
||||
@@ -0,0 +1,202 @@
|
||||
|
||||
Apache License
|
||||
Version 2.0, January 2004
|
||||
http://www.apache.org/licenses/
|
||||
|
||||
TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
|
||||
|
||||
1. Definitions.
|
||||
|
||||
"License" shall mean the terms and conditions for use, reproduction,
|
||||
and distribution as defined by Sections 1 through 9 of this document.
|
||||
|
||||
"Licensor" shall mean the copyright owner or entity authorized by
|
||||
the copyright owner that is granting the License.
|
||||
|
||||
"Legal Entity" shall mean the union of the acting entity and all
|
||||
other entities that control, are controlled by, or are under common
|
||||
control with that entity. For the purposes of this definition,
|
||||
"control" means (i) the power, direct or indirect, to cause the
|
||||
direction or management of such entity, whether by contract or
|
||||
otherwise, or (ii) ownership of fifty percent (50%) or more of the
|
||||
outstanding shares, or (iii) beneficial ownership of such entity.
|
||||
|
||||
"You" (or "Your") shall mean an individual or Legal Entity
|
||||
exercising permissions granted by this License.
|
||||
|
||||
"Source" form shall mean the preferred form for making modifications,
|
||||
including but not limited to software source code, documentation
|
||||
source, and configuration files.
|
||||
|
||||
"Object" form shall mean any form resulting from mechanical
|
||||
transformation or translation of a Source form, including but
|
||||
not limited to compiled object code, generated documentation,
|
||||
and conversions to other media types.
|
||||
|
||||
"Work" shall mean the work of authorship, whether in Source or
|
||||
Object form, made available under the License, as indicated by a
|
||||
copyright notice that is included in or attached to the work
|
||||
(an example is provided in the Appendix below).
|
||||
|
||||
"Derivative Works" shall mean any work, whether in Source or Object
|
||||
form, that is based on (or derived from) the Work and for which the
|
||||
editorial revisions, annotations, elaborations, or other modifications
|
||||
represent, as a whole, an original work of authorship. For the purposes
|
||||
of this License, Derivative Works shall not include works that remain
|
||||
separable from, or merely link (or bind by name) to the interfaces of,
|
||||
the Work and Derivative Works thereof.
|
||||
|
||||
"Contribution" shall mean any work of authorship, including
|
||||
the original version of the Work and any modifications or additions
|
||||
to that Work or Derivative Works thereof, that is intentionally
|
||||
submitted to Licensor for inclusion in the Work by the copyright owner
|
||||
or by an individual or Legal Entity authorized to submit on behalf of
|
||||
the copyright owner. For the purposes of this definition, "submitted"
|
||||
means any form of electronic, verbal, or written communication sent
|
||||
to the Licensor or its representatives, including but not limited to
|
||||
communication on electronic mailing lists, source code control systems,
|
||||
and issue tracking systems that are managed by, or on behalf of, the
|
||||
Licensor for the purpose of discussing and improving the Work, but
|
||||
excluding communication that is conspicuously marked or otherwise
|
||||
designated in writing by the copyright owner as "Not a Contribution."
|
||||
|
||||
"Contributor" shall mean Licensor and any individual or Legal Entity
|
||||
on behalf of whom a Contribution has been received by Licensor and
|
||||
subsequently incorporated within the Work.
|
||||
|
||||
2. Grant of Copyright License. Subject to the terms and conditions of
|
||||
this License, each Contributor hereby grants to You a perpetual,
|
||||
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
|
||||
copyright license to reproduce, prepare Derivative Works of,
|
||||
publicly display, publicly perform, sublicense, and distribute the
|
||||
Work and such Derivative Works in Source or Object form.
|
||||
|
||||
3. Grant of Patent License. Subject to the terms and conditions of
|
||||
this License, each Contributor hereby grants to You a perpetual,
|
||||
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
|
||||
(except as stated in this section) patent license to make, have made,
|
||||
use, offer to sell, sell, import, and otherwise transfer the Work,
|
||||
where such license applies only to those patent claims licensable
|
||||
by such Contributor that are necessarily infringed by their
|
||||
Contribution(s) alone or by combination of their Contribution(s)
|
||||
with the Work to which such Contribution(s) was submitted. If You
|
||||
institute patent litigation against any entity (including a
|
||||
cross-claim or counterclaim in a lawsuit) alleging that the Work
|
||||
or a Contribution incorporated within the Work constitutes direct
|
||||
or contributory patent infringement, then any patent licenses
|
||||
granted to You under this License for that Work shall terminate
|
||||
as of the date such litigation is filed.
|
||||
|
||||
4. Redistribution. You may reproduce and distribute copies of the
|
||||
Work or Derivative Works thereof in any medium, with or without
|
||||
modifications, and in Source or Object form, provided that You
|
||||
meet the following conditions:
|
||||
|
||||
(a) You must give any other recipients of the Work or
|
||||
Derivative Works a copy of this License; and
|
||||
|
||||
(b) You must cause any modified files to carry prominent notices
|
||||
stating that You changed the files; and
|
||||
|
||||
(c) You must retain, in the Source form of any Derivative Works
|
||||
that You distribute, all copyright, patent, trademark, and
|
||||
attribution notices from the Source form of the Work,
|
||||
excluding those notices that do not pertain to any part of
|
||||
the Derivative Works; and
|
||||
|
||||
(d) If the Work includes a "NOTICE" text file as part of its
|
||||
distribution, then any Derivative Works that You distribute must
|
||||
include a readable copy of the attribution notices contained
|
||||
within such NOTICE file, excluding those notices that do not
|
||||
pertain to any part of the Derivative Works, in at least one
|
||||
of the following places: within a NOTICE text file distributed
|
||||
as part of the Derivative Works; within the Source form or
|
||||
documentation, if provided along with the Derivative Works; or,
|
||||
within a display generated by the Derivative Works, if and
|
||||
wherever such third-party notices normally appear. The contents
|
||||
of the NOTICE file are for informational purposes only and
|
||||
do not modify the License. You may add Your own attribution
|
||||
notices within Derivative Works that You distribute, alongside
|
||||
or as an addendum to the NOTICE text from the Work, provided
|
||||
that such additional attribution notices cannot be construed
|
||||
as modifying the License.
|
||||
|
||||
You may add Your own copyright statement to Your modifications and
|
||||
may provide additional or different license terms and conditions
|
||||
for use, reproduction, or distribution of Your modifications, or
|
||||
for any such Derivative Works as a whole, provided Your use,
|
||||
reproduction, and distribution of the Work otherwise complies with
|
||||
the conditions stated in this License.
|
||||
|
||||
5. Submission of Contributions. Unless You explicitly state otherwise,
|
||||
any Contribution intentionally submitted for inclusion in the Work
|
||||
by You to the Licensor shall be under the terms and conditions of
|
||||
this License, without any additional terms or conditions.
|
||||
Notwithstanding the above, nothing herein shall supersede or modify
|
||||
the terms of any separate license agreement you may have executed
|
||||
with Licensor regarding such Contributions.
|
||||
|
||||
6. Trademarks. This License does not grant permission to use the trade
|
||||
names, trademarks, service marks, or product names of the Licensor,
|
||||
except as required for reasonable and customary use in describing the
|
||||
origin of the Work and reproducing the content of the NOTICE file.
|
||||
|
||||
7. Disclaimer of Warranty. Unless required by applicable law or
|
||||
agreed to in writing, Licensor provides the Work (and each
|
||||
Contributor provides its Contributions) on an "AS IS" BASIS,
|
||||
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
|
||||
implied, including, without limitation, any warranties or conditions
|
||||
of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
|
||||
PARTICULAR PURPOSE. You are solely responsible for determining the
|
||||
appropriateness of using or redistributing the Work and assume any
|
||||
risks associated with Your exercise of permissions under this License.
|
||||
|
||||
8. Limitation of Liability. In no event and under no legal theory,
|
||||
whether in tort (including negligence), contract, or otherwise,
|
||||
unless required by applicable law (such as deliberate and grossly
|
||||
negligent acts) or agreed to in writing, shall any Contributor be
|
||||
liable to You for damages, including any direct, indirect, special,
|
||||
incidental, or consequential damages of any character arising as a
|
||||
result of this License or out of the use or inability to use the
|
||||
Work (including but not limited to damages for loss of goodwill,
|
||||
work stoppage, computer failure or malfunction, or any and all
|
||||
other commercial damages or losses), even if such Contributor
|
||||
has been advised of the possibility of such damages.
|
||||
|
||||
9. Accepting Warranty or Additional Liability. While redistributing
|
||||
the Work or Derivative Works thereof, You may choose to offer,
|
||||
and charge a fee for, acceptance of support, warranty, indemnity,
|
||||
or other liability obligations and/or rights consistent with this
|
||||
License. However, in accepting such obligations, You may act only
|
||||
on Your own behalf and on Your sole responsibility, not on behalf
|
||||
of any other Contributor, and only if You agree to indemnify,
|
||||
defend, and hold each Contributor harmless for any liability
|
||||
incurred by, or claims asserted against, such Contributor by reason
|
||||
of your accepting any such warranty or additional liability.
|
||||
|
||||
END OF TERMS AND CONDITIONS
|
||||
|
||||
APPENDIX: How to apply the Apache License to your work.
|
||||
|
||||
To apply the Apache License to your work, attach the following
|
||||
boilerplate notice, with the fields enclosed by brackets "[]"
|
||||
replaced with your own identifying information. (Don't include
|
||||
the brackets!) The text should be enclosed in the appropriate
|
||||
comment syntax for the file format. We also recommend that a
|
||||
file or class name and description of purpose be included on the
|
||||
same "printed page" as the copyright notice for easier
|
||||
identification within third-party archives.
|
||||
|
||||
Copyright [yyyy] [name of copyright owner]
|
||||
|
||||
Licensed under the Apache License, Version 2.0 (the "License");
|
||||
you may not use this file except in compliance with the License.
|
||||
You may obtain a copy of the License at
|
||||
|
||||
http://www.apache.org/licenses/LICENSE-2.0
|
||||
|
||||
Unless required by applicable law or agreed to in writing, software
|
||||
distributed under the License is distributed on an "AS IS" BASIS,
|
||||
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
See the License for the specific language governing permissions and
|
||||
limitations under the License.
|
||||
@@ -0,0 +1,26 @@
|
||||
# Base path used to install.
|
||||
DESTDIR=/usr/local
|
||||
GO_BUILD_FLAGS=
|
||||
GO_TAGS=
|
||||
GO_LDFLAGS=-ldflags '-s -w -extldflags "-static"'
|
||||
SOURCES=$(shell find cmd/ pkg/ vendor/ -name '*.go')
|
||||
DEPLOY_PATH=cri-containerd-staging/gvisor-containerd-shim
|
||||
VERSION=$(shell git rev-parse HEAD)
|
||||
|
||||
bin/gvisor-containerd-shim: $(SOURCES)
|
||||
CGO_ENABLED=0 go build ${GO_BUILD_FLAGS} -o bin/gvisor-containerd-shim ${SHIM_GO_LDFLAGS} ${GO_TAGS} ./cmd/gvisor-containerd-shim
|
||||
|
||||
|
||||
install: bin/gvisor-containerd-shim
|
||||
mkdir -p $(DESTDIR)/bin
|
||||
install bin/gvisor-containerd-shim $(DESTDIR)/bin
|
||||
|
||||
uninstall:
|
||||
rm -f $(DESTDIR)/bin/gvisor-containerd-shim
|
||||
|
||||
clean:
|
||||
rm -rf bin/*
|
||||
|
||||
push: bin/gvisor-containerd-shim
|
||||
gsutil cp ./bin/gvisor-containerd-shim gs://$(DEPLOY_PATH)/gvisor-containerd-shim-$(VERSION)
|
||||
echo "gvisor-containerd-shim-$(VERSION)" | gsutil cp - "gs://$(DEPLOY_PATH)/latest"
|
||||
@@ -0,0 +1,125 @@
|
||||
# gvisor-containerd-shim
|
||||
|
||||
gvisor-containerd-shim is a containerd shim for [gVisor](https://github.com/google/gvisor/). It implements the containerd v1 shim API. It can be used as a drop-in replacement for [containerd-shim](https://github.com/containerd/containerd/tree/master/cmd/containerd-shim) (though containerd-shim must still be installed). It allows the use of both gVisor (runsc) and normal containers in the same containerd installation by deferring to the runc shim if the desired runtime engine is not runsc.
|
||||
|
||||
## Requirements
|
||||
|
||||
- gvisor (runsc) >= a2ad8fe
|
||||
- containerd, containerd-shim >= 1.1
|
||||
|
||||
## Installation
|
||||
|
||||
1. Download the latest release of the gvisor-containerd-shim and unpack the binary to the desired directory:
|
||||
|
||||
```
|
||||
$ tar xf gvisor-containerd-shim.tar.gz
|
||||
$ mv gvisor-containerd-shim /usr/local/bin
|
||||
```
|
||||
|
||||
2. Create the configuration for the gvisor shim in `/etc/containerd/gvisor-containerd-shim.yaml`:
|
||||
|
||||
```
|
||||
# This is the path to the default runc containerd-shim.
|
||||
runc_shim = "/path/to/containerd-shim"
|
||||
```
|
||||
|
||||
3. Update `/etc/containerd/config.toml`. Be sure to update the path to `gvisor-containerd-shim` and `runsc` if necessary:
|
||||
|
||||
```
|
||||
disabled_plugins = ["restart"]
|
||||
[plugins.linux]
|
||||
shim = "/usr/local/bin/gvisor-containerd-shim"
|
||||
shim_debug = true
|
||||
# Uncomment the following 2 lines if you want runsc to be the default runtime.
|
||||
# runtime = "/usr/local/bin/runsc"
|
||||
# runtime_root = "/run/containerd/runsc"
|
||||
# To support the untrusted-workload annotation.
|
||||
[plugins.cri.containerd.untrusted_workload_runtime]
|
||||
runtime_type = "io.containerd.runtime.v1.linux"
|
||||
runtime_engine = "/usr/local/bin/runsc"
|
||||
runtime_root = "/run/containerd/runsc"
|
||||
[plugins.cri.containerd.runtimes.runsc]
|
||||
runtime_type = "io.containerd.runtime.v1.linux"
|
||||
runtime_engine = "/usr/local/bin/runsc"
|
||||
runtime_root = "/run/containerd/runsc"
|
||||
```
|
||||
|
||||
4. Restart `containerd`
|
||||
|
||||
## Usage
|
||||
|
||||
### CRI
|
||||
|
||||
You can run containers in gVisor via containerd's CRI.
|
||||
|
||||
1. Build and install crictl from HEAD:
|
||||
|
||||
```
|
||||
$ go get github.com/kubernetes-sigs/cri-tools/cmd/crictl
|
||||
$ sudo sh -c 'echo "runtime-endpoint: unix:///run/containerd/containerd.sock" > /etc/crictl.yaml'
|
||||
```
|
||||
|
||||
2. Pull the busybox image
|
||||
|
||||
```
|
||||
$ sudo crictl pull busybox
|
||||
```
|
||||
|
||||
### Containerd 1.1
|
||||
|
||||
If running containerd 1.1 you will need to invoke `runsc` via the `io.kubernetes.cri.untrusted-workload` annotation.
|
||||
|
||||
1. Create a pod config:
|
||||
|
||||
```
|
||||
$ cat > sandbox.json << EOL
|
||||
{
|
||||
"metadata": {
|
||||
"name": "nginx-sandbox",
|
||||
"namespace": "default",
|
||||
"attempt": 1,
|
||||
"uid": "hdishd83djaidwnduwk28bcsb",
|
||||
"annotations": {
|
||||
"io.kubernetes.cri.untrusted-workload": "true"
|
||||
}
|
||||
},
|
||||
"linux": {
|
||||
},
|
||||
"log_directory": "/tmp"
|
||||
}
|
||||
EOL
|
||||
```
|
||||
|
||||
2. Run a sandbox with the `runsc` runtime.
|
||||
|
||||
```
|
||||
$ sudo crictl runp sandbox.json
|
||||
```
|
||||
|
||||
### Containerd 1.2
|
||||
|
||||
If running containerd 1.2 you can specify runsc as the runtime using the new runtime handler.
|
||||
|
||||
1. Create a pod config:
|
||||
|
||||
```
|
||||
$ cat > sandbox.json << EOL
|
||||
{
|
||||
"metadata": {
|
||||
"name": "nginx-sandbox",
|
||||
"namespace": "default",
|
||||
"attempt": 1,
|
||||
"uid": "hdishd83djaidwnduwk28bcsb"
|
||||
},
|
||||
"linux": {
|
||||
},
|
||||
"log_directory": "/tmp"
|
||||
}
|
||||
EOL
|
||||
```
|
||||
|
||||
2. Run a sandbox with the `runsc` runtime.
|
||||
|
||||
```
|
||||
$ sudo crictl runp --runtime=runsc sandbox.json
|
||||
```
|
||||
@@ -0,0 +1,42 @@
|
||||
/*
|
||||
Copyright 2018 Google LLC
|
||||
|
||||
Licensed under the Apache License, Version 2.0 (the "License");
|
||||
you may not use this file except in compliance with the License.
|
||||
You may obtain a copy of the License at
|
||||
|
||||
https://www.apache.org/licenses/LICENSE-2.0
|
||||
|
||||
Unless required by applicable law or agreed to in writing, software
|
||||
distributed under the License is distributed on an "AS IS" BASIS,
|
||||
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
See the License for the specific language governing permissions and
|
||||
limitations under the License.
|
||||
*/
|
||||
|
||||
package main
|
||||
|
||||
import "github.com/BurntSushi/toml"
|
||||
|
||||
// config is the configuration for gvisor containerd shim.
|
||||
type config struct {
|
||||
// RuncShim is the shim binary path for standard containerd-shim for runc.
|
||||
// When the runtime is `runc`, gvisor containerd shim will exec current
|
||||
// process to standard containerd-shim. This is a work around for containerd
|
||||
// 1.1. In containerd 1.2, containerd will choose different containerd-shims
|
||||
// based on runtime.
|
||||
RuncShim string `toml:"runc_shim"`
|
||||
// RunscConfig is configuration for runsc. The key value will be converted
|
||||
// to runsc flags --key=value directly.
|
||||
RunscConfig map[string]string `toml:"runsc_config"`
|
||||
}
|
||||
|
||||
// loadConfig load gvisor containerd shim config from config file.
|
||||
func loadConfig(path string) (*config, error) {
|
||||
var c config
|
||||
_, err := toml.DecodeFile(path, &c)
|
||||
if err != nil {
|
||||
return &c, err
|
||||
}
|
||||
return &c, nil
|
||||
}
|
||||
@@ -0,0 +1,320 @@
|
||||
/*
|
||||
Copyright The containerd Authors.
|
||||
Copyright 2018 Google LLC
|
||||
|
||||
Licensed under the Apache License, Version 2.0 (the "License");
|
||||
you may not use this file except in compliance with the License.
|
||||
You may obtain a copy of the License at
|
||||
|
||||
https://www.apache.org/licenses/LICENSE-2.0
|
||||
|
||||
Unless required by applicable law or agreed to in writing, software
|
||||
distributed under the License is distributed on an "AS IS" BASIS,
|
||||
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
See the License for the specific language governing permissions and
|
||||
limitations under the License.
|
||||
*/
|
||||
|
||||
package main
|
||||
|
||||
import (
|
||||
"bytes"
|
||||
"context"
|
||||
"flag"
|
||||
"fmt"
|
||||
"net"
|
||||
"os"
|
||||
"os/exec"
|
||||
"os/signal"
|
||||
"path/filepath"
|
||||
"runtime"
|
||||
"runtime/debug"
|
||||
"strings"
|
||||
"sync"
|
||||
"syscall"
|
||||
"time"
|
||||
|
||||
"github.com/containerd/containerd/events"
|
||||
"github.com/containerd/containerd/namespaces"
|
||||
"github.com/containerd/containerd/runtime/v1/linux/proc"
|
||||
containerdshim "github.com/containerd/containerd/runtime/v1/shim"
|
||||
shimapi "github.com/containerd/containerd/runtime/v1/shim/v1"
|
||||
"github.com/containerd/ttrpc"
|
||||
"github.com/containerd/typeurl"
|
||||
ptypes "github.com/gogo/protobuf/types"
|
||||
"github.com/opencontainers/runc/libcontainer/system"
|
||||
"github.com/pkg/errors"
|
||||
"github.com/sirupsen/logrus"
|
||||
"golang.org/x/sys/unix"
|
||||
|
||||
runsc "github.com/google/gvisor-containerd-shim/pkg/go-runsc"
|
||||
"github.com/google/gvisor-containerd-shim/pkg/shim"
|
||||
)
|
||||
|
||||
var (
|
||||
debugFlag bool
|
||||
namespaceFlag string
|
||||
socketFlag string
|
||||
addressFlag string
|
||||
workdirFlag string
|
||||
runtimeRootFlag string
|
||||
containerdBinaryFlag string
|
||||
shimConfigFlag string
|
||||
)
|
||||
|
||||
// ShimConfigPath is the default shim config file path.
|
||||
const ShimConfigPath = "/etc/containerd/gvisor-containerd-shim.toml"
|
||||
|
||||
func init() {
|
||||
flag.BoolVar(&debugFlag, "debug", false, "enable debug output in logs")
|
||||
flag.StringVar(&namespaceFlag, "namespace", "", "namespace that owns the shim")
|
||||
flag.StringVar(&socketFlag, "socket", "", "abstract socket path to serve")
|
||||
flag.StringVar(&addressFlag, "address", "", "grpc address back to main containerd")
|
||||
flag.StringVar(&workdirFlag, "workdir", "", "path used to storge large temporary data")
|
||||
// Containerd default to runc, unless another runtime is explicitly specified.
|
||||
// We keep the same default to make the default behavior consistent.
|
||||
flag.StringVar(&runtimeRootFlag, "runtime-root", proc.RuncRoot, "root directory for the runtime")
|
||||
// currently, the `containerd publish` utility is embedded in the daemon binary.
|
||||
// The daemon invokes `containerd-shim -containerd-binary ...` with its own os.Executable() path.
|
||||
flag.StringVar(&containerdBinaryFlag, "containerd-binary", "containerd", "path to containerd binary (used for `containerd publish`)")
|
||||
flag.StringVar(&shimConfigFlag, "config", ShimConfigPath, "path to the shim configuration file")
|
||||
flag.Parse()
|
||||
}
|
||||
|
||||
func main() {
|
||||
// This is a hack. Exec current process to run standard containerd-shim
|
||||
// if runtime root is not `runsc`. We don't need this for shim v2 api.
|
||||
if filepath.Base(runtimeRootFlag) != "runsc" {
|
||||
if err := executeRuncShim(); err != nil {
|
||||
fmt.Fprintf(os.Stderr, "gvisor-containerd-shim: %s\n", err)
|
||||
os.Exit(1)
|
||||
}
|
||||
}
|
||||
|
||||
debug.SetGCPercent(40)
|
||||
go func() {
|
||||
for range time.Tick(30 * time.Second) {
|
||||
debug.FreeOSMemory()
|
||||
}
|
||||
}()
|
||||
|
||||
if debugFlag {
|
||||
logrus.SetLevel(logrus.DebugLevel)
|
||||
}
|
||||
|
||||
if os.Getenv("GOMAXPROCS") == "" {
|
||||
// If GOMAXPROCS hasn't been set, we default to a value of 2 to reduce
|
||||
// the number of Go stacks present in the shim.
|
||||
runtime.GOMAXPROCS(2)
|
||||
}
|
||||
|
||||
// Run regular shim if needed.
|
||||
if err := executeShim(); err != nil {
|
||||
fmt.Fprintf(os.Stderr, "gvisor-containerd-shim: %s\n", err)
|
||||
os.Exit(1)
|
||||
}
|
||||
}
|
||||
|
||||
// executeRuncShim execs current process to a containerd-shim process and
|
||||
// retains all flags and envs.
|
||||
func executeRuncShim() error {
|
||||
c, err := loadConfig(shimConfigFlag)
|
||||
if err != nil && !os.IsNotExist(err) {
|
||||
return errors.Wrap(err, "failed to load shim config")
|
||||
}
|
||||
shimPath := c.RuncShim
|
||||
if shimPath == "" {
|
||||
shimPath, err = exec.LookPath("containerd-shim")
|
||||
if err != nil {
|
||||
return errors.Wrapf(err, "lookup containerd-shim")
|
||||
}
|
||||
}
|
||||
|
||||
args := append([]string{shimPath}, os.Args[1:]...)
|
||||
if err := syscall.Exec(shimPath, args, os.Environ()); err != nil {
|
||||
return errors.Wrapf(err, "exec containerd-shim %q", shimPath)
|
||||
}
|
||||
return nil
|
||||
}
|
||||
|
||||
func executeShim() error {
|
||||
// start handling signals as soon as possible so that things are properly reaped
|
||||
// or if runtime exits before we hit the handler
|
||||
signals, err := setupSignals()
|
||||
if err != nil {
|
||||
return err
|
||||
}
|
||||
dump := make(chan os.Signal, 32)
|
||||
signal.Notify(dump, syscall.SIGUSR1)
|
||||
|
||||
path, err := os.Getwd()
|
||||
if err != nil {
|
||||
return err
|
||||
}
|
||||
server, err := ttrpc.NewServer(ttrpc.WithServerHandshaker(ttrpc.UnixSocketRequireSameUser()))
|
||||
if err != nil {
|
||||
return errors.Wrap(err, "failed creating server")
|
||||
}
|
||||
c, err := loadConfig(shimConfigFlag)
|
||||
if err != nil && !os.IsNotExist(err) {
|
||||
return errors.Wrap(err, "failed to load shim config")
|
||||
}
|
||||
sv, err := shim.NewService(
|
||||
shim.Config{
|
||||
Path: path,
|
||||
Namespace: namespaceFlag,
|
||||
WorkDir: workdirFlag,
|
||||
RuntimeRoot: runtimeRootFlag,
|
||||
RunscConfig: c.RunscConfig,
|
||||
},
|
||||
&remoteEventsPublisher{address: addressFlag},
|
||||
)
|
||||
if err != nil {
|
||||
return err
|
||||
}
|
||||
logrus.Debug("registering ttrpc server")
|
||||
shimapi.RegisterShimService(server, sv)
|
||||
|
||||
socket := socketFlag
|
||||
if err := serve(server, socket); err != nil {
|
||||
return err
|
||||
}
|
||||
logger := logrus.WithFields(logrus.Fields{
|
||||
"pid": os.Getpid(),
|
||||
"path": path,
|
||||
"namespace": namespaceFlag,
|
||||
})
|
||||
go func() {
|
||||
for range dump {
|
||||
dumpStacks(logger)
|
||||
}
|
||||
}()
|
||||
return handleSignals(logger, signals, server, sv)
|
||||
}
|
||||
|
||||
// serve serves the ttrpc API over a unix socket at the provided path
|
||||
// this function does not block
|
||||
func serve(server *ttrpc.Server, path string) error {
|
||||
var (
|
||||
l net.Listener
|
||||
err error
|
||||
)
|
||||
if path == "" {
|
||||
l, err = net.FileListener(os.NewFile(3, "socket"))
|
||||
path = "[inherited from parent]"
|
||||
} else {
|
||||
if len(path) > 106 {
|
||||
return errors.Errorf("%q: unix socket path too long (> 106)", path)
|
||||
}
|
||||
l, err = net.Listen("unix", "\x00"+path)
|
||||
}
|
||||
if err != nil {
|
||||
return err
|
||||
}
|
||||
logrus.WithField("socket", path).Debug("serving api on unix socket")
|
||||
go func() {
|
||||
defer l.Close()
|
||||
if err := server.Serve(context.Background(), l); err != nil &&
|
||||
!strings.Contains(err.Error(), "use of closed network connection") {
|
||||
logrus.WithError(err).Fatal("gvisor-containerd-shim: ttrpc server failure")
|
||||
}
|
||||
}()
|
||||
return nil
|
||||
}
|
||||
|
||||
// setupSignals creates a new signal handler for all signals and sets the shim as a
|
||||
// sub-reaper so that the container processes are reparented
|
||||
func setupSignals() (chan os.Signal, error) {
|
||||
signals := make(chan os.Signal, 32)
|
||||
signal.Notify(signals, unix.SIGTERM, unix.SIGINT, unix.SIGCHLD, unix.SIGPIPE)
|
||||
// make sure runc is setup to use the monitor
|
||||
// for waiting on processes
|
||||
// TODO(random-liu): Move shim/reaper.go to a separate package.
|
||||
runsc.Monitor = containerdshim.Default
|
||||
// set the shim as the subreaper for all orphaned processes created by the container
|
||||
if err := system.SetSubreaper(1); err != nil {
|
||||
return nil, err
|
||||
}
|
||||
return signals, nil
|
||||
}
|
||||
|
||||
func handleSignals(logger *logrus.Entry, signals chan os.Signal, server *ttrpc.Server, sv *shim.Service) error {
|
||||
var (
|
||||
termOnce sync.Once
|
||||
done = make(chan struct{})
|
||||
)
|
||||
|
||||
for {
|
||||
select {
|
||||
case <-done:
|
||||
return nil
|
||||
case s := <-signals:
|
||||
switch s {
|
||||
case unix.SIGCHLD:
|
||||
if err := containerdshim.Reap(); err != nil {
|
||||
logger.WithError(err).Error("reap exit status")
|
||||
}
|
||||
case unix.SIGTERM, unix.SIGINT:
|
||||
go termOnce.Do(func() {
|
||||
ctx := context.TODO()
|
||||
if err := server.Shutdown(ctx); err != nil {
|
||||
logger.WithError(err).Error("failed to shutdown server")
|
||||
}
|
||||
// Ensure our child is dead if any
|
||||
sv.Kill(ctx, &shimapi.KillRequest{
|
||||
Signal: uint32(syscall.SIGKILL),
|
||||
All: true,
|
||||
})
|
||||
sv.Delete(context.Background(), &ptypes.Empty{})
|
||||
close(done)
|
||||
})
|
||||
case unix.SIGPIPE:
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
func dumpStacks(logger *logrus.Entry) {
|
||||
var (
|
||||
buf []byte
|
||||
stackSize int
|
||||
)
|
||||
bufferLen := 16384
|
||||
for stackSize == len(buf) {
|
||||
buf = make([]byte, bufferLen)
|
||||
stackSize = runtime.Stack(buf, true)
|
||||
bufferLen *= 2
|
||||
}
|
||||
buf = buf[:stackSize]
|
||||
logger.Infof("=== BEGIN goroutine stack dump ===\n%s\n=== END goroutine stack dump ===", buf)
|
||||
}
|
||||
|
||||
type remoteEventsPublisher struct {
|
||||
address string
|
||||
}
|
||||
|
||||
func (l *remoteEventsPublisher) Publish(ctx context.Context, topic string, event events.Event) error {
|
||||
ns, _ := namespaces.Namespace(ctx)
|
||||
encoded, err := typeurl.MarshalAny(event)
|
||||
if err != nil {
|
||||
return err
|
||||
}
|
||||
data, err := encoded.Marshal()
|
||||
if err != nil {
|
||||
return err
|
||||
}
|
||||
cmd := exec.CommandContext(ctx, containerdBinaryFlag, "--address", l.address, "publish", "--topic", topic, "--namespace", ns)
|
||||
cmd.Stdin = bytes.NewReader(data)
|
||||
c, err := containerdshim.Default.Start(cmd)
|
||||
if err != nil {
|
||||
return err
|
||||
}
|
||||
status, err := containerdshim.Default.Wait(cmd, c)
|
||||
if err != nil {
|
||||
return err
|
||||
}
|
||||
if status != 0 {
|
||||
return errors.New("failed to publish event")
|
||||
}
|
||||
return nil
|
||||
}
|
||||
File diff suppressed because it is too large
Load Diff
@@ -0,0 +1,56 @@
|
||||
/*
|
||||
Copyright The containerd Authors.
|
||||
Copyright 2018 Google LLC
|
||||
|
||||
Licensed under the Apache License, Version 2.0 (the "License");
|
||||
you may not use this file except in compliance with the License.
|
||||
You may obtain a copy of the License at
|
||||
|
||||
https://www.apache.org/licenses/LICENSE-2.0
|
||||
|
||||
Unless required by applicable law or agreed to in writing, software
|
||||
distributed under the License is distributed on an "AS IS" BASIS,
|
||||
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
See the License for the specific language governing permissions and
|
||||
limitations under the License.
|
||||
*/
|
||||
|
||||
package runsc
|
||||
|
||||
import (
|
||||
"bytes"
|
||||
"strings"
|
||||
"sync"
|
||||
)
|
||||
|
||||
var bytesBufferPool = sync.Pool{
|
||||
New: func() interface{} {
|
||||
return bytes.NewBuffer(nil)
|
||||
},
|
||||
}
|
||||
|
||||
func getBuf() *bytes.Buffer {
|
||||
return bytesBufferPool.Get().(*bytes.Buffer)
|
||||
}
|
||||
|
||||
func putBuf(b *bytes.Buffer) {
|
||||
b.Reset()
|
||||
bytesBufferPool.Put(b)
|
||||
}
|
||||
|
||||
// FormatLogPath parses runsc config, and fill in %ID% in the log path.
|
||||
// * For debug-log, it fills in the id in place;
|
||||
// * For user-log, it fills in the id, returns the user log path, and deletes
|
||||
// the `user-log` entry from the config, because it will only be added to runsc
|
||||
// calls that create a new sandbox.
|
||||
func FormatLogPath(id string, config map[string]string) string {
|
||||
if path, ok := config["debug-log"]; ok {
|
||||
config["debug-log"] = strings.Replace(path, "%ID%", id, -1)
|
||||
}
|
||||
var userLog string
|
||||
if path, ok := config["user-log"]; ok {
|
||||
userLog = strings.Replace(path, "%ID%", id, -1)
|
||||
delete(config, "user-log")
|
||||
}
|
||||
return userLog
|
||||
}
|
||||
@@ -0,0 +1,53 @@
|
||||
/*
|
||||
Copyright The containerd Authors.
|
||||
Copyright 2018 Google LLC
|
||||
|
||||
Licensed under the Apache License, Version 2.0 (the "License");
|
||||
you may not use this file except in compliance with the License.
|
||||
You may obtain a copy of the License at
|
||||
|
||||
https://www.apache.org/licenses/LICENSE-2.0
|
||||
|
||||
Unless required by applicable law or agreed to in writing, software
|
||||
distributed under the License is distributed on an "AS IS" BASIS,
|
||||
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
See the License for the specific language governing permissions and
|
||||
limitations under the License.
|
||||
*/
|
||||
|
||||
package proc
|
||||
|
||||
import (
|
||||
"context"
|
||||
|
||||
"github.com/containerd/console"
|
||||
"github.com/containerd/containerd/runtime/proc"
|
||||
"github.com/pkg/errors"
|
||||
)
|
||||
|
||||
type deletedState struct {
|
||||
}
|
||||
|
||||
func (s *deletedState) Resize(ws console.WinSize) error {
|
||||
return errors.Errorf("cannot resize a deleted process")
|
||||
}
|
||||
|
||||
func (s *deletedState) Start(ctx context.Context) error {
|
||||
return errors.Errorf("cannot start a deleted process")
|
||||
}
|
||||
|
||||
func (s *deletedState) Delete(ctx context.Context) error {
|
||||
return errors.Errorf("cannot delete a deleted process")
|
||||
}
|
||||
|
||||
func (s *deletedState) Kill(ctx context.Context, sig uint32, all bool) error {
|
||||
return errors.Errorf("cannot kill a deleted process")
|
||||
}
|
||||
|
||||
func (s *deletedState) SetExited(status int) {
|
||||
// no op
|
||||
}
|
||||
|
||||
func (s *deletedState) Exec(ctx context.Context, path string, r *ExecConfig) (proc.Process, error) {
|
||||
return nil, errors.Errorf("cannot exec in a deleted state")
|
||||
}
|
||||
@@ -0,0 +1,248 @@
|
||||
/*
|
||||
Copyright The containerd Authors.
|
||||
Copyright 2018 Google LLC
|
||||
|
||||
Licensed under the Apache License, Version 2.0 (the "License");
|
||||
you may not use this file except in compliance with the License.
|
||||
You may obtain a copy of the License at
|
||||
|
||||
https://www.apache.org/licenses/LICENSE-2.0
|
||||
|
||||
Unless required by applicable law or agreed to in writing, software
|
||||
distributed under the License is distributed on an "AS IS" BASIS,
|
||||
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
See the License for the specific language governing permissions and
|
||||
limitations under the License.
|
||||
*/
|
||||
|
||||
package proc
|
||||
|
||||
import (
|
||||
"context"
|
||||
"fmt"
|
||||
"io"
|
||||
"os"
|
||||
"path/filepath"
|
||||
"sync"
|
||||
"syscall"
|
||||
"time"
|
||||
|
||||
"github.com/containerd/console"
|
||||
"github.com/containerd/containerd/errdefs"
|
||||
"github.com/containerd/containerd/runtime/proc"
|
||||
"github.com/containerd/containerd/runtime/v1/shim"
|
||||
"github.com/containerd/fifo"
|
||||
runc "github.com/containerd/go-runc"
|
||||
specs "github.com/opencontainers/runtime-spec/specs-go"
|
||||
"github.com/pkg/errors"
|
||||
"golang.org/x/sys/unix"
|
||||
|
||||
runsc "github.com/google/gvisor-containerd-shim/pkg/go-runsc"
|
||||
)
|
||||
|
||||
type execProcess struct {
|
||||
wg sync.WaitGroup
|
||||
|
||||
proc.State
|
||||
|
||||
mu sync.Mutex
|
||||
id string
|
||||
console console.Console
|
||||
io runc.IO
|
||||
status int
|
||||
exited time.Time
|
||||
pid int
|
||||
internalPid int
|
||||
closers []io.Closer
|
||||
stdin io.Closer
|
||||
stdio proc.Stdio
|
||||
path string
|
||||
spec specs.Process
|
||||
|
||||
parent *Init
|
||||
waitBlock chan struct{}
|
||||
}
|
||||
|
||||
func (e *execProcess) Wait() {
|
||||
<-e.waitBlock
|
||||
}
|
||||
|
||||
func (e *execProcess) ID() string {
|
||||
return e.id
|
||||
}
|
||||
|
||||
func (e *execProcess) Pid() int {
|
||||
e.mu.Lock()
|
||||
defer e.mu.Unlock()
|
||||
return e.pid
|
||||
}
|
||||
|
||||
func (e *execProcess) ExitStatus() int {
|
||||
e.mu.Lock()
|
||||
defer e.mu.Unlock()
|
||||
return e.status
|
||||
}
|
||||
|
||||
func (e *execProcess) ExitedAt() time.Time {
|
||||
e.mu.Lock()
|
||||
defer e.mu.Unlock()
|
||||
return e.exited
|
||||
}
|
||||
|
||||
func (e *execProcess) setExited(status int) {
|
||||
e.status = status
|
||||
e.exited = time.Now()
|
||||
e.parent.Platform.ShutdownConsole(context.Background(), e.console)
|
||||
close(e.waitBlock)
|
||||
}
|
||||
|
||||
func (e *execProcess) delete(ctx context.Context) error {
|
||||
e.wg.Wait()
|
||||
if e.io != nil {
|
||||
for _, c := range e.closers {
|
||||
c.Close()
|
||||
}
|
||||
e.io.Close()
|
||||
}
|
||||
pidfile := filepath.Join(e.path, fmt.Sprintf("%s.pid", e.id))
|
||||
// silently ignore error
|
||||
os.Remove(pidfile)
|
||||
internalPidfile := filepath.Join(e.path, fmt.Sprintf("%s-internal.pid", e.id))
|
||||
// silently ignore error
|
||||
os.Remove(internalPidfile)
|
||||
return nil
|
||||
}
|
||||
|
||||
func (e *execProcess) resize(ws console.WinSize) error {
|
||||
if e.console == nil {
|
||||
return nil
|
||||
}
|
||||
return e.console.Resize(ws)
|
||||
}
|
||||
|
||||
func (e *execProcess) kill(ctx context.Context, sig uint32, _ bool) error {
|
||||
internalPid := e.internalPid
|
||||
if internalPid != 0 {
|
||||
if err := e.parent.runtime.Kill(ctx, e.parent.id, int(sig), &runsc.KillOpts{
|
||||
Pid: internalPid,
|
||||
}); err != nil {
|
||||
// If this returns error, consider the process has already stopped.
|
||||
// TODO: Fix after signal handling is fixed.
|
||||
return errors.Wrapf(errdefs.ErrNotFound, err.Error())
|
||||
}
|
||||
}
|
||||
return nil
|
||||
}
|
||||
|
||||
func (e *execProcess) Stdin() io.Closer {
|
||||
return e.stdin
|
||||
}
|
||||
|
||||
func (e *execProcess) Stdio() proc.Stdio {
|
||||
return e.stdio
|
||||
}
|
||||
|
||||
func (e *execProcess) start(ctx context.Context) (err error) {
|
||||
var (
|
||||
socket *runc.Socket
|
||||
pidfile = filepath.Join(e.path, fmt.Sprintf("%s.pid", e.id))
|
||||
internalPidfile = filepath.Join(e.path, fmt.Sprintf("%s-internal.pid", e.id))
|
||||
)
|
||||
if e.stdio.Terminal {
|
||||
if socket, err = runc.NewTempConsoleSocket(); err != nil {
|
||||
return errors.Wrap(err, "failed to create runc console socket")
|
||||
}
|
||||
defer socket.Close()
|
||||
} else if e.stdio.IsNull() {
|
||||
if e.io, err = runc.NewNullIO(); err != nil {
|
||||
return errors.Wrap(err, "creating new NULL IO")
|
||||
}
|
||||
} else {
|
||||
if e.io, err = runc.NewPipeIO(e.parent.IoUID, e.parent.IoGID, withConditionalIO(e.stdio)); err != nil {
|
||||
return errors.Wrap(err, "failed to create runc io pipes")
|
||||
}
|
||||
}
|
||||
opts := &runsc.ExecOpts{
|
||||
PidFile: pidfile,
|
||||
InternalPidFile: internalPidfile,
|
||||
IO: e.io,
|
||||
Detach: true,
|
||||
}
|
||||
if socket != nil {
|
||||
opts.ConsoleSocket = socket
|
||||
}
|
||||
eventCh := shim.Default.Subscribe()
|
||||
defer func() {
|
||||
// Unsubscribe if an error is returned.
|
||||
if err != nil {
|
||||
shim.Default.Unsubscribe(eventCh)
|
||||
}
|
||||
}()
|
||||
if err := e.parent.runtime.Exec(ctx, e.parent.id, e.spec, opts); err != nil {
|
||||
close(e.waitBlock)
|
||||
return e.parent.runtimeError(err, "OCI runtime exec failed")
|
||||
}
|
||||
if e.stdio.Stdin != "" {
|
||||
sc, err := fifo.OpenFifo(ctx, e.stdio.Stdin, syscall.O_WRONLY|syscall.O_NONBLOCK, 0)
|
||||
if err != nil {
|
||||
return errors.Wrapf(err, "failed to open stdin fifo %s", e.stdio.Stdin)
|
||||
}
|
||||
e.closers = append(e.closers, sc)
|
||||
e.stdin = sc
|
||||
}
|
||||
var copyWaitGroup sync.WaitGroup
|
||||
if socket != nil {
|
||||
console, err := socket.ReceiveMaster()
|
||||
if err != nil {
|
||||
return errors.Wrap(err, "failed to retrieve console master")
|
||||
}
|
||||
if e.console, err = e.parent.Platform.CopyConsole(ctx, console, e.stdio.Stdin, e.stdio.Stdout, e.stdio.Stderr, &e.wg, ©WaitGroup); err != nil {
|
||||
return errors.Wrap(err, "failed to start console copy")
|
||||
}
|
||||
} else if !e.stdio.IsNull() {
|
||||
if err := copyPipes(ctx, e.io, e.stdio.Stdin, e.stdio.Stdout, e.stdio.Stderr, &e.wg, ©WaitGroup); err != nil {
|
||||
return errors.Wrap(err, "failed to start io pipe copy")
|
||||
}
|
||||
}
|
||||
copyWaitGroup.Wait()
|
||||
pid, err := runc.ReadPidFile(opts.PidFile)
|
||||
if err != nil {
|
||||
return errors.Wrap(err, "failed to retrieve OCI runtime exec pid")
|
||||
}
|
||||
e.pid = pid
|
||||
internalPid, err := runc.ReadPidFile(opts.InternalPidFile)
|
||||
if err != nil {
|
||||
return errors.Wrap(err, "failed to retrieve OCI runtime exec internal pid")
|
||||
}
|
||||
e.internalPid = internalPid
|
||||
go func() {
|
||||
defer shim.Default.Unsubscribe(eventCh)
|
||||
for event := range eventCh {
|
||||
if event.Pid == e.pid {
|
||||
ExitCh <- Exit{
|
||||
Timestamp: event.Timestamp,
|
||||
ID: e.id,
|
||||
Status: event.Status,
|
||||
}
|
||||
break
|
||||
}
|
||||
}
|
||||
}()
|
||||
return nil
|
||||
}
|
||||
|
||||
func (e *execProcess) Status(ctx context.Context) (string, error) {
|
||||
e.mu.Lock()
|
||||
defer e.mu.Unlock()
|
||||
// if we don't have a pid then the exec process has just been created
|
||||
if e.pid == 0 {
|
||||
return "created", nil
|
||||
}
|
||||
// if we have a pid and it can be signaled, the process is running
|
||||
// TODO(random-liu): Use `runsc kill --pid`.
|
||||
if err := unix.Kill(e.pid, 0); err == nil {
|
||||
return "running", nil
|
||||
}
|
||||
// else if we have a pid but it can nolonger be signaled, it has stopped
|
||||
return "stopped", nil
|
||||
}
|
||||
@@ -0,0 +1,187 @@
|
||||
/*
|
||||
Copyright The containerd Authors.
|
||||
Copyright 2018 Google LLC
|
||||
|
||||
Licensed under the Apache License, Version 2.0 (the "License");
|
||||
you may not use this file except in compliance with the License.
|
||||
You may obtain a copy of the License at
|
||||
|
||||
https://www.apache.org/licenses/LICENSE-2.0
|
||||
|
||||
Unless required by applicable law or agreed to in writing, software
|
||||
distributed under the License is distributed on an "AS IS" BASIS,
|
||||
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
See the License for the specific language governing permissions and
|
||||
limitations under the License.
|
||||
*/
|
||||
|
||||
package proc
|
||||
|
||||
import (
|
||||
"context"
|
||||
|
||||
"github.com/containerd/console"
|
||||
"github.com/pkg/errors"
|
||||
)
|
||||
|
||||
type execCreatedState struct {
|
||||
p *execProcess
|
||||
}
|
||||
|
||||
func (s *execCreatedState) transition(name string) error {
|
||||
switch name {
|
||||
case "running":
|
||||
s.p.State = &execRunningState{p: s.p}
|
||||
case "stopped":
|
||||
s.p.State = &execStoppedState{p: s.p}
|
||||
case "deleted":
|
||||
s.p.State = &deletedState{}
|
||||
default:
|
||||
return errors.Errorf("invalid state transition %q to %q", stateName(s), name)
|
||||
}
|
||||
return nil
|
||||
}
|
||||
|
||||
func (s *execCreatedState) Resize(ws console.WinSize) error {
|
||||
s.p.mu.Lock()
|
||||
defer s.p.mu.Unlock()
|
||||
|
||||
return s.p.resize(ws)
|
||||
}
|
||||
|
||||
func (s *execCreatedState) Start(ctx context.Context) error {
|
||||
s.p.mu.Lock()
|
||||
defer s.p.mu.Unlock()
|
||||
if err := s.p.start(ctx); err != nil {
|
||||
return err
|
||||
}
|
||||
return s.transition("running")
|
||||
}
|
||||
|
||||
func (s *execCreatedState) Delete(ctx context.Context) error {
|
||||
if err := s.p.delete(ctx); err != nil {
|
||||
return err
|
||||
}
|
||||
s.p.mu.Lock()
|
||||
defer s.p.mu.Unlock()
|
||||
return s.transition("deleted")
|
||||
}
|
||||
|
||||
func (s *execCreatedState) Kill(ctx context.Context, sig uint32, all bool) error {
|
||||
s.p.mu.Lock()
|
||||
defer s.p.mu.Unlock()
|
||||
|
||||
return s.p.kill(ctx, sig, all)
|
||||
}
|
||||
|
||||
func (s *execCreatedState) SetExited(status int) {
|
||||
s.p.mu.Lock()
|
||||
defer s.p.mu.Unlock()
|
||||
|
||||
s.p.setExited(status)
|
||||
|
||||
if err := s.transition("stopped"); err != nil {
|
||||
panic(err)
|
||||
}
|
||||
}
|
||||
|
||||
type execRunningState struct {
|
||||
p *execProcess
|
||||
}
|
||||
|
||||
func (s *execRunningState) transition(name string) error {
|
||||
switch name {
|
||||
case "stopped":
|
||||
s.p.State = &execStoppedState{p: s.p}
|
||||
default:
|
||||
return errors.Errorf("invalid state transition %q to %q", stateName(s), name)
|
||||
}
|
||||
return nil
|
||||
}
|
||||
|
||||
func (s *execRunningState) Resize(ws console.WinSize) error {
|
||||
s.p.mu.Lock()
|
||||
defer s.p.mu.Unlock()
|
||||
|
||||
return s.p.resize(ws)
|
||||
}
|
||||
|
||||
func (s *execRunningState) Start(ctx context.Context) error {
|
||||
s.p.mu.Lock()
|
||||
defer s.p.mu.Unlock()
|
||||
|
||||
return errors.Errorf("cannot start a running process")
|
||||
}
|
||||
|
||||
func (s *execRunningState) Delete(ctx context.Context) error {
|
||||
s.p.mu.Lock()
|
||||
defer s.p.mu.Unlock()
|
||||
|
||||
return errors.Errorf("cannot delete a running process")
|
||||
}
|
||||
|
||||
func (s *execRunningState) Kill(ctx context.Context, sig uint32, all bool) error {
|
||||
s.p.mu.Lock()
|
||||
defer s.p.mu.Unlock()
|
||||
|
||||
return s.p.kill(ctx, sig, all)
|
||||
}
|
||||
|
||||
func (s *execRunningState) SetExited(status int) {
|
||||
s.p.mu.Lock()
|
||||
defer s.p.mu.Unlock()
|
||||
|
||||
s.p.setExited(status)
|
||||
|
||||
if err := s.transition("stopped"); err != nil {
|
||||
panic(err)
|
||||
}
|
||||
}
|
||||
|
||||
type execStoppedState struct {
|
||||
p *execProcess
|
||||
}
|
||||
|
||||
func (s *execStoppedState) transition(name string) error {
|
||||
switch name {
|
||||
case "deleted":
|
||||
s.p.State = &deletedState{}
|
||||
default:
|
||||
return errors.Errorf("invalid state transition %q to %q", stateName(s), name)
|
||||
}
|
||||
return nil
|
||||
}
|
||||
|
||||
func (s *execStoppedState) Resize(ws console.WinSize) error {
|
||||
s.p.mu.Lock()
|
||||
defer s.p.mu.Unlock()
|
||||
|
||||
return errors.Errorf("cannot resize a stopped container")
|
||||
}
|
||||
|
||||
func (s *execStoppedState) Start(ctx context.Context) error {
|
||||
s.p.mu.Lock()
|
||||
defer s.p.mu.Unlock()
|
||||
|
||||
return errors.Errorf("cannot start a stopped process")
|
||||
}
|
||||
|
||||
func (s *execStoppedState) Delete(ctx context.Context) error {
|
||||
if err := s.p.delete(ctx); err != nil {
|
||||
return err
|
||||
}
|
||||
s.p.mu.Lock()
|
||||
defer s.p.mu.Unlock()
|
||||
return s.transition("deleted")
|
||||
}
|
||||
|
||||
func (s *execStoppedState) Kill(ctx context.Context, sig uint32, all bool) error {
|
||||
s.p.mu.Lock()
|
||||
defer s.p.mu.Unlock()
|
||||
|
||||
return s.p.kill(ctx, sig, all)
|
||||
}
|
||||
|
||||
func (s *execStoppedState) SetExited(status int) {
|
||||
// no op
|
||||
}
|
||||
@@ -0,0 +1,413 @@
|
||||
/*
|
||||
Copyright The containerd Authors.
|
||||
Copyright 2018 Google LLC
|
||||
|
||||
Licensed under the Apache License, Version 2.0 (the "License");
|
||||
you may not use this file except in compliance with the License.
|
||||
You may obtain a copy of the License at
|
||||
|
||||
https://www.apache.org/licenses/LICENSE-2.0
|
||||
|
||||
Unless required by applicable law or agreed to in writing, software
|
||||
distributed under the License is distributed on an "AS IS" BASIS,
|
||||
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
See the License for the specific language governing permissions and
|
||||
limitations under the License.
|
||||
*/
|
||||
|
||||
package proc
|
||||
|
||||
import (
|
||||
"context"
|
||||
"encoding/json"
|
||||
"io"
|
||||
"path/filepath"
|
||||
"strings"
|
||||
"sync"
|
||||
"syscall"
|
||||
"time"
|
||||
|
||||
"github.com/containerd/console"
|
||||
"github.com/containerd/containerd/errdefs"
|
||||
"github.com/containerd/containerd/log"
|
||||
"github.com/containerd/containerd/mount"
|
||||
"github.com/containerd/containerd/runtime/proc"
|
||||
"github.com/containerd/fifo"
|
||||
runc "github.com/containerd/go-runc"
|
||||
specs "github.com/opencontainers/runtime-spec/specs-go"
|
||||
"github.com/pkg/errors"
|
||||
|
||||
runsc "github.com/google/gvisor-containerd-shim/pkg/go-runsc"
|
||||
)
|
||||
|
||||
// InitPidFile name of the file that contains the init pid
|
||||
const InitPidFile = "init.pid"
|
||||
|
||||
// Init represents an initial process for a container
|
||||
type Init struct {
|
||||
wg sync.WaitGroup
|
||||
initState
|
||||
|
||||
// mu is used to ensure that `Start()` and `Exited()` calls return in
|
||||
// the right order when invoked in separate go routines.
|
||||
// This is the case within the shim implementation as it makes use of
|
||||
// the reaper interface.
|
||||
mu sync.Mutex
|
||||
|
||||
waitBlock chan struct{}
|
||||
|
||||
WorkDir string
|
||||
|
||||
id string
|
||||
Bundle string
|
||||
console console.Console
|
||||
Platform proc.Platform
|
||||
io runc.IO
|
||||
runtime *runsc.Runsc
|
||||
status int
|
||||
exited time.Time
|
||||
pid int
|
||||
closers []io.Closer
|
||||
stdin io.Closer
|
||||
stdio proc.Stdio
|
||||
Rootfs string
|
||||
IoUID int
|
||||
IoGID int
|
||||
Sandbox bool
|
||||
UserLog string
|
||||
}
|
||||
|
||||
// NewRunsc returns a new runsc instance for a process
|
||||
func NewRunsc(root, path, namespace, runtime string, config map[string]string) *runsc.Runsc {
|
||||
if root == "" {
|
||||
root = RunscRoot
|
||||
}
|
||||
return &runsc.Runsc{
|
||||
Command: runtime,
|
||||
PdeathSignal: syscall.SIGKILL,
|
||||
Log: filepath.Join(path, "log.json"),
|
||||
LogFormat: runc.JSON,
|
||||
Root: filepath.Join(root, namespace),
|
||||
Config: config,
|
||||
}
|
||||
}
|
||||
|
||||
// New returns a new init process
|
||||
func New(id string, runtime *runsc.Runsc, stdio proc.Stdio) *Init {
|
||||
p := &Init{
|
||||
id: id,
|
||||
runtime: runtime,
|
||||
stdio: stdio,
|
||||
status: 0,
|
||||
waitBlock: make(chan struct{}),
|
||||
}
|
||||
p.initState = &createdState{p: p}
|
||||
return p
|
||||
}
|
||||
|
||||
// Create the process with the provided config
|
||||
func (p *Init) Create(ctx context.Context, r *CreateConfig) error {
|
||||
var (
|
||||
err error
|
||||
socket *runc.Socket
|
||||
)
|
||||
if r.Terminal {
|
||||
if socket, err = runc.NewTempConsoleSocket(); err != nil {
|
||||
return errors.Wrap(err, "failed to create OCI runtime console socket")
|
||||
}
|
||||
defer socket.Close()
|
||||
} else if hasNoIO(r) {
|
||||
if p.io, err = runc.NewNullIO(); err != nil {
|
||||
return errors.Wrap(err, "creating new NULL IO")
|
||||
}
|
||||
} else {
|
||||
if p.io, err = runc.NewPipeIO(p.IoUID, p.IoGID, withConditionalIO(p.stdio)); err != nil {
|
||||
return errors.Wrap(err, "failed to create OCI runtime io pipes")
|
||||
}
|
||||
}
|
||||
pidFile := filepath.Join(p.Bundle, InitPidFile)
|
||||
opts := &runsc.CreateOpts{
|
||||
PidFile: pidFile,
|
||||
}
|
||||
if socket != nil {
|
||||
opts.ConsoleSocket = socket
|
||||
}
|
||||
if p.Sandbox {
|
||||
opts.IO = p.io
|
||||
// UserLog is only useful for sandbox.
|
||||
opts.UserLog = p.UserLog
|
||||
}
|
||||
if err := p.runtime.Create(ctx, r.ID, r.Bundle, opts); err != nil {
|
||||
return p.runtimeError(err, "OCI runtime create failed")
|
||||
}
|
||||
if r.Stdin != "" {
|
||||
sc, err := fifo.OpenFifo(ctx, r.Stdin, syscall.O_WRONLY|syscall.O_NONBLOCK, 0)
|
||||
if err != nil {
|
||||
return errors.Wrapf(err, "failed to open stdin fifo %s", r.Stdin)
|
||||
}
|
||||
p.stdin = sc
|
||||
p.closers = append(p.closers, sc)
|
||||
}
|
||||
var copyWaitGroup sync.WaitGroup
|
||||
if socket != nil {
|
||||
console, err := socket.ReceiveMaster()
|
||||
if err != nil {
|
||||
return errors.Wrap(err, "failed to retrieve console master")
|
||||
}
|
||||
console, err = p.Platform.CopyConsole(ctx, console, r.Stdin, r.Stdout, r.Stderr, &p.wg, ©WaitGroup)
|
||||
if err != nil {
|
||||
return errors.Wrap(err, "failed to start console copy")
|
||||
}
|
||||
p.console = console
|
||||
} else if !hasNoIO(r) {
|
||||
if err := copyPipes(ctx, p.io, r.Stdin, r.Stdout, r.Stderr, &p.wg, ©WaitGroup); err != nil {
|
||||
return errors.Wrap(err, "failed to start io pipe copy")
|
||||
}
|
||||
}
|
||||
|
||||
copyWaitGroup.Wait()
|
||||
pid, err := runc.ReadPidFile(pidFile)
|
||||
if err != nil {
|
||||
return errors.Wrap(err, "failed to retrieve OCI runtime container pid")
|
||||
}
|
||||
p.pid = pid
|
||||
return nil
|
||||
}
|
||||
|
||||
// Wait for the process to exit
|
||||
func (p *Init) Wait() {
|
||||
<-p.waitBlock
|
||||
}
|
||||
|
||||
// ID of the process
|
||||
func (p *Init) ID() string {
|
||||
return p.id
|
||||
}
|
||||
|
||||
// Pid of the process
|
||||
func (p *Init) Pid() int {
|
||||
return p.pid
|
||||
}
|
||||
|
||||
// ExitStatus of the process
|
||||
func (p *Init) ExitStatus() int {
|
||||
p.mu.Lock()
|
||||
defer p.mu.Unlock()
|
||||
return p.status
|
||||
}
|
||||
|
||||
// ExitedAt at time when the process exited
|
||||
func (p *Init) ExitedAt() time.Time {
|
||||
p.mu.Lock()
|
||||
defer p.mu.Unlock()
|
||||
return p.exited
|
||||
}
|
||||
|
||||
// Status of the process
|
||||
func (p *Init) Status(ctx context.Context) (string, error) {
|
||||
p.mu.Lock()
|
||||
defer p.mu.Unlock()
|
||||
c, err := p.runtime.State(ctx, p.id)
|
||||
if err != nil {
|
||||
if strings.Contains(err.Error(), "does not exist") {
|
||||
return "stopped", nil
|
||||
}
|
||||
return "", p.runtimeError(err, "OCI runtime state failed")
|
||||
}
|
||||
return p.convertStatus(c.Status), nil
|
||||
}
|
||||
|
||||
func (p *Init) start(context context.Context) error {
|
||||
var cio runc.IO
|
||||
if !p.Sandbox {
|
||||
cio = p.io
|
||||
}
|
||||
if err := p.runtime.Start(context, p.id, cio); err != nil {
|
||||
return p.runtimeError(err, "OCI runtime start failed")
|
||||
}
|
||||
go func() {
|
||||
status, err := p.runtime.Wait(context, p.id)
|
||||
if err != nil {
|
||||
log.G(context).WithError(err).Errorf("Failed to wait for container %q", p.id)
|
||||
// TODO(random-liu): Handle runsc kill error.
|
||||
if err := p.killAll(context); err != nil {
|
||||
log.G(context).WithError(err).Errorf("Failed to kill container %q", p.id)
|
||||
}
|
||||
status = internalErrorCode
|
||||
}
|
||||
ExitCh <- Exit{
|
||||
Timestamp: time.Now(),
|
||||
ID: p.id,
|
||||
Status: status,
|
||||
}
|
||||
}()
|
||||
return nil
|
||||
}
|
||||
|
||||
func (p *Init) setExited(status int) {
|
||||
p.exited = time.Now()
|
||||
p.status = status
|
||||
p.Platform.ShutdownConsole(context.Background(), p.console)
|
||||
close(p.waitBlock)
|
||||
}
|
||||
|
||||
func (p *Init) delete(context context.Context) error {
|
||||
p.killAll(context)
|
||||
p.wg.Wait()
|
||||
err := p.runtime.Delete(context, p.id, nil)
|
||||
// ignore errors if a runtime has already deleted the process
|
||||
// but we still hold metadata and pipes
|
||||
//
|
||||
// this is common during a checkpoint, runc will delete the container state
|
||||
// after a checkpoint and the container will no longer exist within runc
|
||||
if err != nil {
|
||||
if strings.Contains(err.Error(), "does not exist") {
|
||||
err = nil
|
||||
} else {
|
||||
err = p.runtimeError(err, "failed to delete task")
|
||||
}
|
||||
}
|
||||
if p.io != nil {
|
||||
for _, c := range p.closers {
|
||||
c.Close()
|
||||
}
|
||||
p.io.Close()
|
||||
}
|
||||
if err2 := mount.UnmountAll(p.Rootfs, 0); err2 != nil {
|
||||
log.G(context).WithError(err2).Warn("failed to cleanup rootfs mount")
|
||||
if err == nil {
|
||||
err = errors.Wrap(err2, "failed rootfs umount")
|
||||
}
|
||||
}
|
||||
return err
|
||||
}
|
||||
|
||||
func (p *Init) resize(ws console.WinSize) error {
|
||||
if p.console == nil {
|
||||
return nil
|
||||
}
|
||||
return p.console.Resize(ws)
|
||||
}
|
||||
|
||||
func (p *Init) kill(context context.Context, signal uint32, all bool) error {
|
||||
var (
|
||||
killErr error
|
||||
backoff = 100 * time.Millisecond
|
||||
)
|
||||
timeout := 1 * time.Second
|
||||
for start := time.Now(); time.Now().Sub(start) < timeout; {
|
||||
c, err := p.runtime.State(context, p.id)
|
||||
if err != nil {
|
||||
if strings.Contains(err.Error(), "does not exist") {
|
||||
return errors.Wrapf(errdefs.ErrNotFound, "no such process")
|
||||
}
|
||||
return p.runtimeError(err, "OCI runtime state failed")
|
||||
}
|
||||
// For runsc, signal only works when container is running state.
|
||||
// If the container is not in running state, directly return
|
||||
// "no such process"
|
||||
if p.convertStatus(c.Status) == "stopped" {
|
||||
return errors.Wrapf(errdefs.ErrNotFound, "no such process")
|
||||
}
|
||||
killErr = p.runtime.Kill(context, p.id, int(signal), &runsc.KillOpts{
|
||||
All: all,
|
||||
})
|
||||
if killErr == nil {
|
||||
return nil
|
||||
}
|
||||
time.Sleep(backoff)
|
||||
backoff *= 2
|
||||
}
|
||||
return p.runtimeError(killErr, "kill timeout")
|
||||
}
|
||||
|
||||
// KillAll processes belonging to the init process
|
||||
func (p *Init) KillAll(context context.Context) error {
|
||||
p.mu.Lock()
|
||||
defer p.mu.Unlock()
|
||||
return p.killAll(context)
|
||||
}
|
||||
|
||||
func (p *Init) killAll(context context.Context) error {
|
||||
p.runtime.Kill(context, p.id, int(syscall.SIGKILL), &runsc.KillOpts{
|
||||
All: true,
|
||||
})
|
||||
// Ignore error handling for `runsc kill --all` for now.
|
||||
// * If it doesn't return error, it is good;
|
||||
// * If it returns error, consider the container has already stopped.
|
||||
// TODO: Fix `runsc kill --all` error handling.
|
||||
return nil
|
||||
}
|
||||
|
||||
// Stdin of the process
|
||||
func (p *Init) Stdin() io.Closer {
|
||||
return p.stdin
|
||||
}
|
||||
|
||||
// Runtime returns the OCI runtime configured for the init process
|
||||
func (p *Init) Runtime() *runsc.Runsc {
|
||||
return p.runtime
|
||||
}
|
||||
|
||||
// exec returns a new exec'd process
|
||||
func (p *Init) exec(context context.Context, path string, r *ExecConfig) (proc.Process, error) {
|
||||
// process exec request
|
||||
var spec specs.Process
|
||||
if err := json.Unmarshal(r.Spec.Value, &spec); err != nil {
|
||||
return nil, err
|
||||
}
|
||||
spec.Terminal = r.Terminal
|
||||
|
||||
e := &execProcess{
|
||||
id: r.ID,
|
||||
path: path,
|
||||
parent: p,
|
||||
spec: spec,
|
||||
stdio: proc.Stdio{
|
||||
Stdin: r.Stdin,
|
||||
Stdout: r.Stdout,
|
||||
Stderr: r.Stderr,
|
||||
Terminal: r.Terminal,
|
||||
},
|
||||
waitBlock: make(chan struct{}),
|
||||
}
|
||||
e.State = &execCreatedState{p: e}
|
||||
return e, nil
|
||||
}
|
||||
|
||||
// Stdio of the process
|
||||
func (p *Init) Stdio() proc.Stdio {
|
||||
return p.stdio
|
||||
}
|
||||
|
||||
func (p *Init) runtimeError(rErr error, msg string) error {
|
||||
if rErr == nil {
|
||||
return nil
|
||||
}
|
||||
|
||||
rMsg, err := getLastRuntimeError(p.runtime)
|
||||
switch {
|
||||
case err != nil:
|
||||
return errors.Wrapf(rErr, "%s: %s (%s)", msg, "unable to retrieve OCI runtime error", err.Error())
|
||||
case rMsg == "":
|
||||
return errors.Wrap(rErr, msg)
|
||||
default:
|
||||
return errors.Errorf("%s: %s", msg, rMsg)
|
||||
}
|
||||
}
|
||||
|
||||
func (p *Init) convertStatus(status string) string {
|
||||
if status == "created" && !p.Sandbox && p.status == internalErrorCode {
|
||||
// Treat start failure state for non-root container as stopped.
|
||||
return "stopped"
|
||||
}
|
||||
return status
|
||||
}
|
||||
|
||||
func withConditionalIO(c proc.Stdio) runc.IOOpt {
|
||||
return func(o *runc.IOOption) {
|
||||
o.OpenStdin = c.Stdin != ""
|
||||
o.OpenStdout = c.Stdout != ""
|
||||
o.OpenStderr = c.Stderr != ""
|
||||
}
|
||||
}
|
||||
@@ -0,0 +1,224 @@
|
||||
/*
|
||||
Copyright The containerd Authors.
|
||||
Copyright 2018 Google LLC
|
||||
|
||||
Licensed under the Apache License, Version 2.0 (the "License");
|
||||
you may not use this file except in compliance with the License.
|
||||
You may obtain a copy of the License at
|
||||
|
||||
https://www.apache.org/licenses/LICENSE-2.0
|
||||
|
||||
Unless required by applicable law or agreed to in writing, software
|
||||
distributed under the License is distributed on an "AS IS" BASIS,
|
||||
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
See the License for the specific language governing permissions and
|
||||
limitations under the License.
|
||||
*/
|
||||
|
||||
package proc
|
||||
|
||||
import (
|
||||
"context"
|
||||
|
||||
"github.com/containerd/console"
|
||||
"github.com/containerd/containerd/errdefs"
|
||||
"github.com/containerd/containerd/runtime/proc"
|
||||
"github.com/pkg/errors"
|
||||
)
|
||||
|
||||
type initState interface {
|
||||
proc.State
|
||||
|
||||
Exec(context.Context, string, *ExecConfig) (proc.Process, error)
|
||||
}
|
||||
|
||||
type createdState struct {
|
||||
p *Init
|
||||
}
|
||||
|
||||
func (s *createdState) transition(name string) error {
|
||||
switch name {
|
||||
case "running":
|
||||
s.p.initState = &runningState{p: s.p}
|
||||
case "stopped":
|
||||
s.p.initState = &stoppedState{p: s.p}
|
||||
case "deleted":
|
||||
s.p.initState = &deletedState{}
|
||||
default:
|
||||
return errors.Errorf("invalid state transition %q to %q", stateName(s), name)
|
||||
}
|
||||
return nil
|
||||
}
|
||||
|
||||
func (s *createdState) Resize(ws console.WinSize) error {
|
||||
s.p.mu.Lock()
|
||||
defer s.p.mu.Unlock()
|
||||
|
||||
return s.p.resize(ws)
|
||||
}
|
||||
|
||||
func (s *createdState) Start(ctx context.Context) error {
|
||||
s.p.mu.Lock()
|
||||
defer s.p.mu.Unlock()
|
||||
if err := s.p.start(ctx); err != nil {
|
||||
// Containerd doesn't allow deleting container in created state.
|
||||
// However, for gvisor, a non-root container in created state can
|
||||
// only go to running state. If the container can't be started,
|
||||
// it can only stay in created state, and never be deleted.
|
||||
// To work around that, we treat non-root container in start failure
|
||||
// state as stopped.
|
||||
if !s.p.Sandbox {
|
||||
s.p.io.Close()
|
||||
s.p.setExited(internalErrorCode)
|
||||
if err := s.transition("stopped"); err != nil {
|
||||
panic(err)
|
||||
}
|
||||
}
|
||||
return err
|
||||
}
|
||||
return s.transition("running")
|
||||
}
|
||||
|
||||
func (s *createdState) Delete(ctx context.Context) error {
|
||||
s.p.mu.Lock()
|
||||
defer s.p.mu.Unlock()
|
||||
if err := s.p.delete(ctx); err != nil {
|
||||
return err
|
||||
}
|
||||
return s.transition("deleted")
|
||||
}
|
||||
|
||||
func (s *createdState) Kill(ctx context.Context, sig uint32, all bool) error {
|
||||
s.p.mu.Lock()
|
||||
defer s.p.mu.Unlock()
|
||||
|
||||
return s.p.kill(ctx, sig, all)
|
||||
}
|
||||
|
||||
func (s *createdState) SetExited(status int) {
|
||||
s.p.mu.Lock()
|
||||
defer s.p.mu.Unlock()
|
||||
|
||||
s.p.setExited(status)
|
||||
|
||||
if err := s.transition("stopped"); err != nil {
|
||||
panic(err)
|
||||
}
|
||||
}
|
||||
|
||||
func (s *createdState) Exec(ctx context.Context, path string, r *ExecConfig) (proc.Process, error) {
|
||||
s.p.mu.Lock()
|
||||
defer s.p.mu.Unlock()
|
||||
return s.p.exec(ctx, path, r)
|
||||
}
|
||||
|
||||
type runningState struct {
|
||||
p *Init
|
||||
}
|
||||
|
||||
func (s *runningState) transition(name string) error {
|
||||
switch name {
|
||||
case "stopped":
|
||||
s.p.initState = &stoppedState{p: s.p}
|
||||
default:
|
||||
return errors.Errorf("invalid state transition %q to %q", stateName(s), name)
|
||||
}
|
||||
return nil
|
||||
}
|
||||
|
||||
func (s *runningState) Resize(ws console.WinSize) error {
|
||||
s.p.mu.Lock()
|
||||
defer s.p.mu.Unlock()
|
||||
|
||||
return s.p.resize(ws)
|
||||
}
|
||||
|
||||
func (s *runningState) Start(ctx context.Context) error {
|
||||
s.p.mu.Lock()
|
||||
defer s.p.mu.Unlock()
|
||||
|
||||
return errors.Errorf("cannot start a running process")
|
||||
}
|
||||
|
||||
func (s *runningState) Delete(ctx context.Context) error {
|
||||
s.p.mu.Lock()
|
||||
defer s.p.mu.Unlock()
|
||||
|
||||
return errors.Errorf("cannot delete a running process")
|
||||
}
|
||||
|
||||
func (s *runningState) Kill(ctx context.Context, sig uint32, all bool) error {
|
||||
s.p.mu.Lock()
|
||||
defer s.p.mu.Unlock()
|
||||
|
||||
return s.p.kill(ctx, sig, all)
|
||||
}
|
||||
|
||||
func (s *runningState) SetExited(status int) {
|
||||
s.p.mu.Lock()
|
||||
defer s.p.mu.Unlock()
|
||||
|
||||
s.p.setExited(status)
|
||||
|
||||
if err := s.transition("stopped"); err != nil {
|
||||
panic(err)
|
||||
}
|
||||
}
|
||||
|
||||
func (s *runningState) Exec(ctx context.Context, path string, r *ExecConfig) (proc.Process, error) {
|
||||
s.p.mu.Lock()
|
||||
defer s.p.mu.Unlock()
|
||||
return s.p.exec(ctx, path, r)
|
||||
}
|
||||
|
||||
type stoppedState struct {
|
||||
p *Init
|
||||
}
|
||||
|
||||
func (s *stoppedState) transition(name string) error {
|
||||
switch name {
|
||||
case "deleted":
|
||||
s.p.initState = &deletedState{}
|
||||
default:
|
||||
return errors.Errorf("invalid state transition %q to %q", stateName(s), name)
|
||||
}
|
||||
return nil
|
||||
}
|
||||
|
||||
func (s *stoppedState) Resize(ws console.WinSize) error {
|
||||
s.p.mu.Lock()
|
||||
defer s.p.mu.Unlock()
|
||||
|
||||
return errors.Errorf("cannot resize a stopped container")
|
||||
}
|
||||
|
||||
func (s *stoppedState) Start(ctx context.Context) error {
|
||||
s.p.mu.Lock()
|
||||
defer s.p.mu.Unlock()
|
||||
|
||||
return errors.Errorf("cannot start a stopped process")
|
||||
}
|
||||
|
||||
func (s *stoppedState) Delete(ctx context.Context) error {
|
||||
s.p.mu.Lock()
|
||||
defer s.p.mu.Unlock()
|
||||
if err := s.p.delete(ctx); err != nil {
|
||||
return err
|
||||
}
|
||||
return s.transition("deleted")
|
||||
}
|
||||
|
||||
func (s *stoppedState) Kill(ctx context.Context, sig uint32, all bool) error {
|
||||
return errdefs.ToGRPCf(errdefs.ErrNotFound, "process %s not found", s.p.id)
|
||||
}
|
||||
|
||||
func (s *stoppedState) SetExited(status int) {
|
||||
// no op
|
||||
}
|
||||
|
||||
func (s *stoppedState) Exec(ctx context.Context, path string, r *ExecConfig) (proc.Process, error) {
|
||||
s.p.mu.Lock()
|
||||
defer s.p.mu.Unlock()
|
||||
|
||||
return nil, errors.Errorf("cannot exec in a stopped state")
|
||||
}
|
||||
+146
@@ -0,0 +1,146 @@
|
||||
/*
|
||||
Copyright The containerd Authors.
|
||||
Copyright 2018 Google LLC
|
||||
|
||||
Licensed under the Apache License, Version 2.0 (the "License");
|
||||
you may not use this file except in compliance with the License.
|
||||
You may obtain a copy of the License at
|
||||
|
||||
https://www.apache.org/licenses/LICENSE-2.0
|
||||
|
||||
Unless required by applicable law or agreed to in writing, software
|
||||
distributed under the License is distributed on an "AS IS" BASIS,
|
||||
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
See the License for the specific language governing permissions and
|
||||
limitations under the License.
|
||||
*/
|
||||
|
||||
package proc
|
||||
|
||||
import (
|
||||
"context"
|
||||
"fmt"
|
||||
"io"
|
||||
"os"
|
||||
"sync"
|
||||
"syscall"
|
||||
|
||||
"github.com/containerd/fifo"
|
||||
runc "github.com/containerd/go-runc"
|
||||
)
|
||||
|
||||
// TODO(random-liu): This file can be a util.
|
||||
|
||||
var bufPool = sync.Pool{
|
||||
New: func() interface{} {
|
||||
buffer := make([]byte, 32<<10)
|
||||
return &buffer
|
||||
},
|
||||
}
|
||||
|
||||
func copyPipes(ctx context.Context, rio runc.IO, stdin, stdout, stderr string, wg, cwg *sync.WaitGroup) error {
|
||||
var sameFile io.WriteCloser
|
||||
for _, i := range []struct {
|
||||
name string
|
||||
dest func(wc io.WriteCloser, rc io.Closer)
|
||||
}{
|
||||
{
|
||||
name: stdout,
|
||||
dest: func(wc io.WriteCloser, rc io.Closer) {
|
||||
wg.Add(1)
|
||||
cwg.Add(1)
|
||||
go func() {
|
||||
cwg.Done()
|
||||
p := bufPool.Get().(*[]byte)
|
||||
defer bufPool.Put(p)
|
||||
io.CopyBuffer(wc, rio.Stdout(), *p)
|
||||
wg.Done()
|
||||
wc.Close()
|
||||
if rc != nil {
|
||||
rc.Close()
|
||||
}
|
||||
}()
|
||||
},
|
||||
}, {
|
||||
name: stderr,
|
||||
dest: func(wc io.WriteCloser, rc io.Closer) {
|
||||
wg.Add(1)
|
||||
cwg.Add(1)
|
||||
go func() {
|
||||
cwg.Done()
|
||||
p := bufPool.Get().(*[]byte)
|
||||
defer bufPool.Put(p)
|
||||
io.CopyBuffer(wc, rio.Stderr(), *p)
|
||||
wg.Done()
|
||||
wc.Close()
|
||||
if rc != nil {
|
||||
rc.Close()
|
||||
}
|
||||
}()
|
||||
},
|
||||
},
|
||||
} {
|
||||
ok, err := isFifo(i.name)
|
||||
if err != nil {
|
||||
return err
|
||||
}
|
||||
var (
|
||||
fw io.WriteCloser
|
||||
fr io.Closer
|
||||
)
|
||||
if ok {
|
||||
if fw, err = fifo.OpenFifo(ctx, i.name, syscall.O_WRONLY, 0); err != nil {
|
||||
return fmt.Errorf("gvisor-containerd-shim: opening %s failed: %s", i.name, err)
|
||||
}
|
||||
if fr, err = fifo.OpenFifo(ctx, i.name, syscall.O_RDONLY, 0); err != nil {
|
||||
return fmt.Errorf("gvisor-containerd-shim: opening %s failed: %s", i.name, err)
|
||||
}
|
||||
} else {
|
||||
if sameFile != nil {
|
||||
i.dest(sameFile, nil)
|
||||
continue
|
||||
}
|
||||
if fw, err = os.OpenFile(i.name, syscall.O_WRONLY|syscall.O_APPEND, 0); err != nil {
|
||||
return fmt.Errorf("gvisor-containerd-shim: opening %s failed: %s", i.name, err)
|
||||
}
|
||||
if stdout == stderr {
|
||||
sameFile = fw
|
||||
}
|
||||
}
|
||||
i.dest(fw, fr)
|
||||
}
|
||||
if stdin == "" {
|
||||
return nil
|
||||
}
|
||||
f, err := fifo.OpenFifo(ctx, stdin, syscall.O_RDONLY|syscall.O_NONBLOCK, 0)
|
||||
if err != nil {
|
||||
return fmt.Errorf("gvisor-containerd-shim: opening %s failed: %s", stdin, err)
|
||||
}
|
||||
cwg.Add(1)
|
||||
go func() {
|
||||
cwg.Done()
|
||||
p := bufPool.Get().(*[]byte)
|
||||
defer bufPool.Put(p)
|
||||
|
||||
io.CopyBuffer(rio.Stdin(), f, *p)
|
||||
rio.Stdin().Close()
|
||||
f.Close()
|
||||
}()
|
||||
return nil
|
||||
}
|
||||
|
||||
// isFifo checks if a file is a fifo
|
||||
// if the file does not exist then it returns false
|
||||
func isFifo(path string) (bool, error) {
|
||||
stat, err := os.Stat(path)
|
||||
if err != nil {
|
||||
if os.IsNotExist(err) {
|
||||
return false, nil
|
||||
}
|
||||
return false, err
|
||||
}
|
||||
if stat.Mode()&os.ModeNamedPipe == os.ModeNamedPipe {
|
||||
return true, nil
|
||||
}
|
||||
return false, nil
|
||||
}
|
||||
@@ -0,0 +1,39 @@
|
||||
/*
|
||||
Copyright The containerd Authors.
|
||||
Copyright 2018 Google LLC
|
||||
|
||||
Licensed under the Apache License, Version 2.0 (the "License");
|
||||
you may not use this file except in compliance with the License.
|
||||
You may obtain a copy of the License at
|
||||
|
||||
https://www.apache.org/licenses/LICENSE-2.0
|
||||
|
||||
Unless required by applicable law or agreed to in writing, software
|
||||
distributed under the License is distributed on an "AS IS" BASIS,
|
||||
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
See the License for the specific language governing permissions and
|
||||
limitations under the License.
|
||||
*/
|
||||
|
||||
package proc
|
||||
|
||||
import (
|
||||
"github.com/pkg/errors"
|
||||
)
|
||||
|
||||
// RunscRoot is the path to the root runsc state directory
|
||||
const RunscRoot = "/run/containerd/runsc"
|
||||
|
||||
func stateName(v interface{}) string {
|
||||
switch v.(type) {
|
||||
case *runningState, *execRunningState:
|
||||
return "running"
|
||||
case *createdState, *execCreatedState:
|
||||
return "created"
|
||||
case *deletedState:
|
||||
return "deleted"
|
||||
case *stoppedState:
|
||||
return "stopped"
|
||||
}
|
||||
panic(errors.Errorf("invalid state %v", v))
|
||||
}
|
||||
@@ -0,0 +1,62 @@
|
||||
/*
|
||||
Copyright The containerd Authors.
|
||||
Copyright 2018 Google LLC
|
||||
|
||||
Licensed under the Apache License, Version 2.0 (the "License");
|
||||
you may not use this file except in compliance with the License.
|
||||
You may obtain a copy of the License at
|
||||
|
||||
https://www.apache.org/licenses/LICENSE-2.0
|
||||
|
||||
Unless required by applicable law or agreed to in writing, software
|
||||
distributed under the License is distributed on an "AS IS" BASIS,
|
||||
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
See the License for the specific language governing permissions and
|
||||
limitations under the License.
|
||||
*/
|
||||
|
||||
package proc
|
||||
|
||||
import (
|
||||
"time"
|
||||
|
||||
google_protobuf "github.com/gogo/protobuf/types"
|
||||
)
|
||||
|
||||
// Mount holds filesystem mount configuration
|
||||
type Mount struct {
|
||||
Type string
|
||||
Source string
|
||||
Target string
|
||||
Options []string
|
||||
}
|
||||
|
||||
// CreateConfig hold task creation configuration
|
||||
type CreateConfig struct {
|
||||
ID string
|
||||
Bundle string
|
||||
Runtime string
|
||||
Rootfs []Mount
|
||||
Terminal bool
|
||||
Stdin string
|
||||
Stdout string
|
||||
Stderr string
|
||||
Options *google_protobuf.Any
|
||||
}
|
||||
|
||||
// ExecConfig holds exec creation configuration
|
||||
type ExecConfig struct {
|
||||
ID string
|
||||
Terminal bool
|
||||
Stdin string
|
||||
Stdout string
|
||||
Stderr string
|
||||
Spec *google_protobuf.Any
|
||||
}
|
||||
|
||||
// Exit is the type of exit events
|
||||
type Exit struct {
|
||||
Timestamp time.Time
|
||||
ID string
|
||||
Status int
|
||||
}
|
||||
@@ -0,0 +1,94 @@
|
||||
/*
|
||||
Copyright The containerd Authors.
|
||||
Copyright 2018 Google LLC
|
||||
|
||||
Licensed under the Apache License, Version 2.0 (the "License");
|
||||
you may not use this file except in compliance with the License.
|
||||
You may obtain a copy of the License at
|
||||
|
||||
https://www.apache.org/licenses/LICENSE-2.0
|
||||
|
||||
Unless required by applicable law or agreed to in writing, software
|
||||
distributed under the License is distributed on an "AS IS" BASIS,
|
||||
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
See the License for the specific language governing permissions and
|
||||
limitations under the License.
|
||||
*/
|
||||
|
||||
package proc
|
||||
|
||||
import (
|
||||
"encoding/json"
|
||||
"io"
|
||||
"os"
|
||||
"strings"
|
||||
"time"
|
||||
|
||||
runsc "github.com/google/gvisor-containerd-shim/pkg/go-runsc"
|
||||
)
|
||||
|
||||
const (
|
||||
internalErrorCode = 128
|
||||
bufferSize = 32
|
||||
)
|
||||
|
||||
// ExitCh is the exit events channel for containers and exec processes
|
||||
// inside the sandbox.
|
||||
var ExitCh = make(chan Exit, bufferSize)
|
||||
|
||||
// TODO(random-liu): This can be a utility.
|
||||
|
||||
// TODO(mlaventure): move to runc package?
|
||||
func getLastRuntimeError(r *runsc.Runsc) (string, error) {
|
||||
if r.Log == "" {
|
||||
return "", nil
|
||||
}
|
||||
|
||||
f, err := os.OpenFile(r.Log, os.O_RDONLY, 0400)
|
||||
if err != nil {
|
||||
return "", err
|
||||
}
|
||||
|
||||
var (
|
||||
errMsg string
|
||||
log struct {
|
||||
Level string
|
||||
Msg string
|
||||
Time time.Time
|
||||
}
|
||||
)
|
||||
|
||||
dec := json.NewDecoder(f)
|
||||
for err = nil; err == nil; {
|
||||
if err = dec.Decode(&log); err != nil && err != io.EOF {
|
||||
return "", err
|
||||
}
|
||||
if log.Level == "error" {
|
||||
errMsg = strings.TrimSpace(log.Msg)
|
||||
}
|
||||
}
|
||||
|
||||
return errMsg, nil
|
||||
}
|
||||
|
||||
func copyFile(to, from string) error {
|
||||
ff, err := os.Open(from)
|
||||
if err != nil {
|
||||
return err
|
||||
}
|
||||
defer ff.Close()
|
||||
tt, err := os.Create(to)
|
||||
if err != nil {
|
||||
return err
|
||||
}
|
||||
defer tt.Close()
|
||||
|
||||
p := bufPool.Get().(*[]byte)
|
||||
defer bufPool.Put(p)
|
||||
_, err = io.CopyBuffer(tt, ff, *p)
|
||||
return err
|
||||
}
|
||||
|
||||
func hasNoIO(r *CreateConfig) bool {
|
||||
return r.Stdin == "" && r.Stdout == "" && r.Stderr == ""
|
||||
}
|
||||
@@ -0,0 +1,112 @@
|
||||
/*
|
||||
Copyright The containerd Authors.
|
||||
Copyright 2018 Google LLC
|
||||
|
||||
Licensed under the Apache License, Version 2.0 (the "License");
|
||||
you may not use this file except in compliance with the License.
|
||||
You may obtain a copy of the License at
|
||||
|
||||
https://www.apache.org/licenses/LICENSE-2.0
|
||||
|
||||
Unless required by applicable law or agreed to in writing, software
|
||||
distributed under the License is distributed on an "AS IS" BASIS,
|
||||
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
See the License for the specific language governing permissions and
|
||||
limitations under the License.
|
||||
*/
|
||||
|
||||
package shim
|
||||
|
||||
import (
|
||||
"context"
|
||||
"io"
|
||||
"sync"
|
||||
"syscall"
|
||||
|
||||
"github.com/containerd/console"
|
||||
"github.com/containerd/fifo"
|
||||
"github.com/pkg/errors"
|
||||
)
|
||||
|
||||
type linuxPlatform struct {
|
||||
epoller *console.Epoller
|
||||
}
|
||||
|
||||
func (p *linuxPlatform) CopyConsole(ctx context.Context, console console.Console, stdin, stdout, stderr string, wg, cwg *sync.WaitGroup) (console.Console, error) {
|
||||
if p.epoller == nil {
|
||||
return nil, errors.New("uninitialized epoller")
|
||||
}
|
||||
|
||||
epollConsole, err := p.epoller.Add(console)
|
||||
if err != nil {
|
||||
return nil, err
|
||||
}
|
||||
|
||||
if stdin != "" {
|
||||
in, err := fifo.OpenFifo(ctx, stdin, syscall.O_RDONLY, 0)
|
||||
if err != nil {
|
||||
return nil, err
|
||||
}
|
||||
cwg.Add(1)
|
||||
go func() {
|
||||
cwg.Done()
|
||||
p := bufPool.Get().(*[]byte)
|
||||
defer bufPool.Put(p)
|
||||
io.CopyBuffer(epollConsole, in, *p)
|
||||
}()
|
||||
}
|
||||
|
||||
outw, err := fifo.OpenFifo(ctx, stdout, syscall.O_WRONLY, 0)
|
||||
if err != nil {
|
||||
return nil, err
|
||||
}
|
||||
outr, err := fifo.OpenFifo(ctx, stdout, syscall.O_RDONLY, 0)
|
||||
if err != nil {
|
||||
return nil, err
|
||||
}
|
||||
wg.Add(1)
|
||||
cwg.Add(1)
|
||||
go func() {
|
||||
cwg.Done()
|
||||
p := bufPool.Get().(*[]byte)
|
||||
defer bufPool.Put(p)
|
||||
io.CopyBuffer(outw, epollConsole, *p)
|
||||
epollConsole.Close()
|
||||
outr.Close()
|
||||
outw.Close()
|
||||
wg.Done()
|
||||
}()
|
||||
return epollConsole, nil
|
||||
}
|
||||
|
||||
func (p *linuxPlatform) ShutdownConsole(ctx context.Context, cons console.Console) error {
|
||||
if p.epoller == nil {
|
||||
return errors.New("uninitialized epoller")
|
||||
}
|
||||
epollConsole, ok := cons.(*console.EpollConsole)
|
||||
if !ok {
|
||||
return errors.Errorf("expected EpollConsole, got %#v", cons)
|
||||
}
|
||||
return epollConsole.Shutdown(p.epoller.CloseConsole)
|
||||
}
|
||||
|
||||
func (p *linuxPlatform) Close() error {
|
||||
return p.epoller.Close()
|
||||
}
|
||||
|
||||
// initialize a single epoll fd to manage our consoles. `initPlatform` should
|
||||
// only be called once.
|
||||
func (s *Service) initPlatform() error {
|
||||
if s.platform != nil {
|
||||
return nil
|
||||
}
|
||||
epoller, err := console.NewEpoller()
|
||||
if err != nil {
|
||||
return errors.Wrap(err, "failed to initialize epoller")
|
||||
}
|
||||
s.platform = &linuxPlatform{
|
||||
epoller: epoller,
|
||||
}
|
||||
go epoller.Wait()
|
||||
return nil
|
||||
}
|
||||
File diff suppressed because it is too large
Load Diff
Some files were not shown because too many files have changed in this diff Show More
Reference in New Issue
Block a user