{% include 'notebox_end' %}
+h3. CrunchRunCommand: Using host networking for containers
+
+Older Linux kernels (prior to 3.18) have bugs in network namespace handling which can lead to compute node lockups. This by is indicated by blocked kernel tasks in "Workqueue: netns cleanup_net". If you are experiencing this problem, as a workaround you can disable use of network namespaces by Docker across the cluster. Be aware this reduces container isolation, which may be a security risk.
+
+<notextile>
+<pre><code class="userinput">Client:
+ APIHost: <b>zzzzz.arvadosapi.com</b>
+ AuthToken: <b>zzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz</b>
+CrunchRunCommand:
+- <b>crunch-run</b>
+- <b>"-container-enable-networking=always"</b>
+- <b>"-container-network-mode=host"</b>
+</code></pre>
+</notextile>
+
h3. MinRetryPeriod: Rate-limit repeated attempts to start containers
If SLURM is unable to run a container, the dispatcher will submit it again after the next PollPeriod. If PollPeriod is very short, this can be excessive. If MinRetryPeriod is set, the dispatcher will avoid submitting the same container to SLURM more than once in the given time span.
cStateLock sync.Mutex
cStarted bool // StartContainer() succeeded
cCancelled bool // StopContainer() invoked
+
+ enableNetwork string // one of "default" or "always"
+ networkMode string // passed through to HostConfig.NetworkMode
}
// SetupSignals sets up signal handling to gracefully terminate the underlying
for k, v := range runner.Container.Environment {
runner.ContainerConfig.Env = append(runner.ContainerConfig.Env, k+"="+v)
}
+
+ runner.HostConfig = dockerclient.HostConfig{
+ Binds: runner.Binds,
+ CgroupParent: runner.setCgroupParent,
+ LogConfig: dockerclient.LogConfig{
+ Type: "none",
+ },
+ }
+
if wantAPI := runner.Container.RuntimeConstraints.API; wantAPI != nil && *wantAPI {
tok, err := runner.ContainerToken()
if err != nil {
"ARVADOS_API_HOST="+os.Getenv("ARVADOS_API_HOST"),
"ARVADOS_API_HOST_INSECURE="+os.Getenv("ARVADOS_API_HOST_INSECURE"),
)
- runner.ContainerConfig.NetworkDisabled = false
+ runner.HostConfig.NetworkMode = runner.networkMode
} else {
- runner.ContainerConfig.NetworkDisabled = true
+ if runner.enableNetwork == "always" {
+ runner.HostConfig.NetworkMode = runner.networkMode
+ } else {
+ runner.HostConfig.NetworkMode = "none"
+ }
}
var err error
return fmt.Errorf("While creating container: %v", err)
}
- runner.HostConfig = dockerclient.HostConfig{
- Binds: runner.Binds,
- CgroupParent: runner.setCgroupParent,
- LogConfig: dockerclient.LogConfig{
- Type: "none",
- },
- }
-
return runner.AttachStreams()
}
cgroupParent := flag.String("cgroup-parent", "docker", "name of container's parent cgroup (ignored if -cgroup-parent-subsystem is used)")
cgroupParentSubsystem := flag.String("cgroup-parent-subsystem", "", "use current cgroup for given subsystem as parent cgroup for container")
caCertsPath := flag.String("ca-certs", "", "Path to TLS root certificates")
+ enableNetwork := flag.String("container-enable-networking", "default",
+ `Specify if networking should be enabled for container. One of 'default', 'always':
+ default: only enable networking if container requests it.
+ always: containers always have networking enabled
+ `)
+ networkMode := flag.String("container-network-mode", "default",
+ `Set networking mode for container. Corresponds to Docker network mode (--net).
+ `)
flag.Parse()
containerId := flag.Arg(0)
cr.statInterval = *statInterval
cr.cgroupRoot = *cgroupRoot
cr.expectCgroupParent = *cgroupParent
+ cr.enableNetwork = *enableNetwork
+ cr.networkMode = *networkMode
if *cgroupParentSubsystem != "" {
p := findCgroup(*cgroupParentSubsystem)
cr.setCgroupParent = p