client

package
Version: v0.3.2 Latest Latest
Warning

This package is not in the latest version of its module.

Go to latest
Published: Apr 22, 2016 License: MPL-2.0 Imports: 26 Imported by: 0

Documentation

Index

Constants

View Source
const (
	ReasonNoRestartsAllowed   = "Policy allows no restarts"
	ReasonUnrecoverableErrror = "Error was unrecoverable"
	ReasonWithinPolicy        = "Restart within policy"
	ReasonDelay               = "Exceeded allowed attempts, applying a delay"
)

Variables

This section is empty.

Functions

func DefaultConfig

func DefaultConfig() *config.Config

DefaultConfig returns the default configuration

Types

type AllocRunner

type AllocRunner struct {
	// contains filtered or unexported fields
}

AllocRunner is used to wrap an allocation and provide the execution context.

func NewAllocRunner

func NewAllocRunner(logger *log.Logger, config *config.Config, updater AllocStateUpdater,
	alloc *structs.Allocation) *AllocRunner

NewAllocRunner is used to create a new allocation context

func (*AllocRunner) Alloc

func (r *AllocRunner) Alloc() *structs.Allocation

Alloc returns the associated allocation

func (*AllocRunner) Destroy

func (r *AllocRunner) Destroy()

Destroy is used to indicate that the allocation context should be destroyed

func (*AllocRunner) DestroyContext

func (r *AllocRunner) DestroyContext() error

DestroyContext is used to destroy the context

func (*AllocRunner) DestroyState

func (r *AllocRunner) DestroyState() error

DestroyState is used to cleanup after ourselves

func (*AllocRunner) RestoreState

func (r *AllocRunner) RestoreState() error

RestoreState is used to restore the state of the alloc runner

func (*AllocRunner) Run

func (r *AllocRunner) Run()

Run is a long running goroutine used to manage an allocation

func (*AllocRunner) SaveState

func (r *AllocRunner) SaveState() error

SaveState is used to snapshot the state of the alloc runner if the fullSync is marked as false only the state of the Alloc Runner is snapshotted. If fullSync is marked as true, we snapshot all the Task Runners associated with the Alloc

func (*AllocRunner) Update

func (r *AllocRunner) Update(update *structs.Allocation)

Update is used to update the allocation of the context

func (*AllocRunner) WaitCh added in v0.1.2

func (r *AllocRunner) WaitCh() <-chan struct{}

WaitCh returns a channel to wait for termination

type AllocStateUpdater

type AllocStateUpdater func(alloc *structs.Allocation)

AllocStateUpdater is used to update the status of an allocation

type Client

type Client struct {
	// contains filtered or unexported fields
}

Client is used to implement the client interaction with Nomad. Clients are expected to register as a schedulable node to the servers, and to run allocations as determined by the servers.

func NewClient

func NewClient(cfg *config.Config) (*Client, error)

NewClient is used to create a new client from the given configuration

func (*Client) GetAllocFS added in v0.3.0

func (c *Client) GetAllocFS(allocID string) (allocdir.AllocDirFS, error)

GetAllocFS returns the AllocFS interface for the alloc dir of an allocation

func (*Client) Leave

func (c *Client) Leave() error

Leave is used to prepare the client to leave the cluster

func (*Client) Node

func (c *Client) Node() *structs.Node

Node returns the locally registered node

func (*Client) RPC

func (c *Client) RPC(method string, args interface{}, reply interface{}) error

RPC is used to forward an RPC call to a nomad server, or fail if no servers

func (*Client) Servers

func (c *Client) Servers() []string

Servers is used to return the current known servers list. When an agent is first started, this list comes directly from configuration files.

func (*Client) SetServers

func (c *Client) SetServers(servers []string)

SetServers is used to modify the known servers list. This avoids forcing a config rollout + rolling restart and enables auto-join features. The full set of servers is passed to support adding and/or removing servers.

func (*Client) Shutdown

func (c *Client) Shutdown() error

Shutdown is used to tear down the client

func (*Client) Stats

func (c *Client) Stats() map[string]map[string]string

Stats is used to return statistics for debugging and insight for various sub-systems

type RestartTracker added in v0.3.0

type RestartTracker struct {
	// contains filtered or unexported fields
}

func (*RestartTracker) GetReason added in v0.3.2

func (r *RestartTracker) GetReason() string

GetReason returns a human-readable description for the last state returned by GetState.

func (*RestartTracker) GetState added in v0.3.1

func (r *RestartTracker) GetState() (string, time.Duration)

GetState returns the tasks next state given the set exit code and start error. One of the following states are returned: * TaskRestarting - Task should be restarted * TaskNotRestarting - Task should not be restarted and has exceeded its

restart policy.

* TaskTerminated - Task has terminated successfully and does not need a

restart.

If TaskRestarting is returned, the duration is how long to wait until starting the task again.

func (*RestartTracker) SetPolicy added in v0.3.0

func (r *RestartTracker) SetPolicy(policy *structs.RestartPolicy)

SetPolicy updates the policy used to determine restarts.

func (*RestartTracker) SetStartError added in v0.3.1

func (r *RestartTracker) SetStartError(err error) *RestartTracker

SetStartError is used to mark the most recent start error. If starting was successful the error should be nil.

func (*RestartTracker) SetWaitResult added in v0.3.1

func (r *RestartTracker) SetWaitResult(res *cstructs.WaitResult) *RestartTracker

SetWaitResult is used to mark the most recent wait result.

type TaskRunner

type TaskRunner struct {
	// contains filtered or unexported fields
}

TaskRunner is used to wrap a task within an allocation and provide the execution context.

func NewTaskRunner

func NewTaskRunner(logger *log.Logger, config *config.Config,
	updater TaskStateUpdater, ctx *driver.ExecContext,
	alloc *structs.Allocation, task *structs.Task) *TaskRunner

NewTaskRunner is used to create a new task context

func (*TaskRunner) Destroy

func (r *TaskRunner) Destroy()

Destroy is used to indicate that the task context should be destroyed

func (*TaskRunner) DestroyState

func (r *TaskRunner) DestroyState() error

DestroyState is used to cleanup after ourselves

func (*TaskRunner) MarkReceived added in v0.3.2

func (r *TaskRunner) MarkReceived()

MarkReceived marks the task as received.

func (*TaskRunner) RestoreState

func (r *TaskRunner) RestoreState() error

RestoreState is used to restore our state

func (*TaskRunner) Run

func (r *TaskRunner) Run()

Run is a long running routine used to manage the task

func (*TaskRunner) SaveState

func (r *TaskRunner) SaveState() error

SaveState is used to snapshot our state

func (*TaskRunner) Update

func (r *TaskRunner) Update(update *structs.Allocation)

Update is used to update the task of the context

func (*TaskRunner) WaitCh

func (r *TaskRunner) WaitCh() <-chan struct{}

WaitCh returns a channel to wait for termination

type TaskStateUpdater

type TaskStateUpdater func(taskName, state string, event *structs.TaskEvent)

TaskStateUpdater is used to signal that tasks state has changed.

Directories

Path Synopsis
Functions shared between linux/darwin.
Functions shared between linux/darwin.
env

Jump to

Keyboard shortcuts

? : This menu
/ : Search site
f or F : Jump to
y or Y : Canonical URL