Skip to main content

Class: WorkflowClient

client.WorkflowClient

Client for starting Workflow executions and creating Workflow handles.

Typically this client should not be instantiated directly, instead create the high level Client and use Client.workflow to interact with Workflows.

Hierarchy

Constructors

constructor

new WorkflowClient(options?): WorkflowClient

Parameters

NameType
options?WorkflowClientOptions

Returns

WorkflowClient

Overrides

BaseClient.constructor

Properties

connection

Readonly connection: ConnectionLike

The underlying connection used by this client.

Clients are cheap to create, but connections are expensive. Where that make sense, a single connection may and should be reused by multiple Client.

Inherited from

BaseClient.connection


options

Readonly options: LoadedWorkflowClientOptions

Accessors

workflowService

get workflowService(): WorkflowService

Raw gRPC access to the Temporal service.

NOTE: The namespace provided in options is not automatically set on requests made via this service object.

Returns

WorkflowService

Methods

count

count(query?): Promise<CountWorkflowExecution>

Return the number of Workflow Executions matching the given query. If no query is provided, then return the total number of Workflow Executions for this namespace.

Note that the number of Workflow Executions returned is approximate and eventually consistent.

More info on the concept of "visibility" and the query syntax on the Temporal documentation site: https://docs.temporal.io/visibility

Parameters

NameType
query?string

Returns

Promise<CountWorkflowExecution>


execute

execute<T>(workflowTypeOrFunc, options): Promise<WorkflowResultType<T>>

Start a new Workflow execution, then await for its completion and return that Workflow's result.

Type parameters

NameType
Textends Workflow

Parameters

NameType
workflowTypeOrFuncstring | T
optionsWorkflowStartOptions<T>

Returns

Promise<WorkflowResultType<T>>

the result of the Workflow execution


executeUpdateWithStart

executeUpdateWithStart<T, Ret, Args>(updateDef, updateOptions): Promise<Ret>

Start a new Workflow Execution and immediately send an Update to that Workflow, then await and return the Update's result.

The updateOptions object must contain a WithStartWorkflowOperation, which defines the options for the Workflow execution to start (e.g. the Workflow's type, task queue, input arguments, etc.)

The behavior of Update-with-Start in the case where there is already a running Workflow with the given Workflow ID depends on the specified WorkflowIDConflictPolicy. That is, if the policy is USE_EXISTING, then the Update is issued against the already existing Workflow Execution; however, if the policy is FAIL, then an error is thrown. Caller MUST specify the desired WorkflowIDConflictPolicy.

This call will block until the Update has completed. The Workflow handle can be retrieved by awaiting on WithStartWorkflowOperation.workflowHandle, whether or not the Update succeeds.

Type parameters

NameType
Textends Workflow
RetRet
Argsextends any[]

Parameters

NameType
updateDefstring | UpdateDefinition<Ret, Args, string>
updateOptionsWorkflowUpdateOptions & { args?: Args ; startWorkflowOperation: WithStartWorkflowOperation<T> }

Returns

Promise<Ret>

the Update result

Update-with-Start is an experimental feature and may be subject to change.


getHandle

getHandle<T>(workflowId, runId?, options?): WorkflowHandle<T>

Create a handle to an existing Workflow.

  • If only workflowId is passed, and there are multiple Workflow Executions with that ID, the handle will refer to the most recent one.
  • If workflowId and runId are passed, the handle will refer to the specific Workflow Execution with that Run ID.
  • If workflowId and GetWorkflowHandleOptions.firstExecutionRunId are passed, the handle will refer to the most recent Workflow Execution in the Chain that started with firstExecutionRunId.

A Chain is a series of Workflow Executions that share the same Workflow ID and are connected by:

This method does not validate workflowId. If there is no Workflow Execution with the given workflowId, handle methods like handle.describe() will throw a WorkflowNotFoundError error.

Type parameters

NameType
Textends Workflow

Parameters

NameType
workflowIdstring
runId?string
options?GetWorkflowHandleOptions

Returns

WorkflowHandle<T>


list

list(options?): AsyncWorkflowListIterable

Return a list of Workflow Executions matching the given query.

Note that the list of Workflow Executions returned is approximate and eventually consistent.

More info on the concept of "visibility" and the query syntax on the Temporal documentation site: https://docs.temporal.io/visibility

Parameters

NameType
options?ListOptions

Returns

AsyncWorkflowListIterable


result

result<T>(workflowId, runId?, opts?): Promise<WorkflowResultType<T>>

Get the result of a Workflow execution.

Follow the chain of execution in case Workflow continues as new, or has a cron schedule or retry policy.

Type parameters

NameType
Textends Workflow

Parameters

NameType
workflowIdstring
runId?string
opts?WorkflowResultOptions

Returns

Promise<WorkflowResultType<T>>


signalWithStart

signalWithStart<WorkflowFn, SignalArgs>(workflowTypeOrFunc, options): Promise<WorkflowHandleWithSignaledRunId<WorkflowFn>>

Start a new Workflow Execution and immediately send a Signal to that Workflow.

The behavior of Signal-with-Start in the case where there is already a running Workflow with the given Workflow ID depends on the WorkflowIDConflictPolicy. That is, if the policy is USE_EXISTING, then the Signal is issued against the already existing Workflow Execution; however, if the policy is FAIL, then an error is thrown. If no policy is specified, Signal-with-Start defaults to USE_EXISTING.

Type parameters

NameType
WorkflowFnextends Workflow
SignalArgsextends any[] = []

Parameters

NameType
workflowTypeOrFuncstring | WorkflowFn
optionsWithWorkflowArgs<WorkflowFn, WorkflowSignalWithStartOptions<SignalArgs>>

Returns

Promise<WorkflowHandleWithSignaledRunId<WorkflowFn>>

a WorkflowHandle to the started Workflow


start

start<T>(workflowTypeOrFunc, options): Promise<WorkflowHandleWithFirstExecutionRunId<T>>

Start a new Workflow execution.

Type parameters

NameType
Textends Workflow

Parameters

NameType
workflowTypeOrFuncstring | T
optionsWorkflowStartOptions<T>

Returns

Promise<WorkflowHandleWithFirstExecutionRunId<T>>

a WorkflowHandle to the started Workflow


startUpdateWithStart

startUpdateWithStart<T, Ret, Args>(updateDef, updateOptions): Promise<WorkflowUpdateHandle<Ret>>

Start a new Workflow Execution and immediately send an Update to that Workflow, then return a WorkflowUpdateHandle for that Update.

The updateOptions object must contain a WithStartWorkflowOperation, which defines the options for the Workflow execution to start (e.g. the Workflow's type, task queue, input arguments, etc.)

The behavior of Update-with-Start in the case where there is already a running Workflow with the given Workflow ID depends on the specified WorkflowIDConflictPolicy. That is, if the policy is USE_EXISTING, then the Update is issued against the already existing Workflow Execution; however, if the policy is FAIL, then an error is thrown. Caller MUST specify the desired WorkflowIDConflictPolicy.

This call will block until the Update has reached the specified WorkflowUpdateStage. Note that this means that the call will not return successfully until the Update has been delivered to a Worker. The Workflow handle can be retrieved by awaiting on WithStartWorkflowOperation.workflowHandle, whether or not the Update succeeds.

Type parameters

NameType
Textends Workflow
RetRet
Argsextends any[]

Parameters

NameType
updateDefstring | UpdateDefinition<Ret, Args, string>
updateOptionsWorkflowUpdateOptions & { args?: Args ; startWorkflowOperation: WithStartWorkflowOperation<T> ; waitForStage: "ACCEPTED" }

Returns

Promise<WorkflowUpdateHandle<Ret>>

a WorkflowUpdateHandle to the started Update

Update-with-Start is an experimental feature and may be subject to change.


withAbortSignal

withAbortSignal<R>(abortSignal, fn): Promise<R>

Set an AbortSignal that, when aborted, cancels any ongoing service requests executed in fn's scope. This will locally result in the request call throwing a _grpc.ServiceError|ServiceError with code _grpc.status.CANCELLED|CANCELLED; see isGrpcCancelledError.

This method is only a convenience wrapper around Connection.withAbortSignal.

Type parameters

Name
R

Parameters

NameType
abortSignalAbortSignal
fn() => Promise<R>

Returns

Promise<R>

value returned from fn

See

https://developer.mozilla.org/en-US/docs/Web/API/AbortSignal

Inherited from

BaseClient.withAbortSignal


withDeadline

withDeadline<R>(deadline, fn): Promise<R>

Set a deadline for any service requests executed in fn's scope.

The deadline is a point in time after which any pending gRPC request will be considered as failed; this will locally result in the request call throwing a _grpc.ServiceError|ServiceError with code _grpc.status.DEADLINE_EXCEEDED|DEADLINE_EXCEEDED; see isGrpcDeadlineError.

It is stronly recommended to explicitly set deadlines. If no deadline is set, then it is possible for the client to end up waiting forever for a response.

This method is only a convenience wrapper around Connection.withDeadline.

Type parameters

Name
R

Parameters

NameTypeDescription
deadlinenumber | Datea point in time after which the request will be considered as failed; either a Date object, or a number of milliseconds since the Unix epoch (UTC).
fn() => Promise<R>-

Returns

Promise<R>

the value returned from fn

See

https://grpc.io/docs/guides/deadlines/

Inherited from

BaseClient.withDeadline


withMetadata

withMetadata<R>(metadata, fn): Promise<R>

Set metadata for any service requests executed in fn's scope.

This method is only a convenience wrapper around Connection.withMetadata.

Type parameters

Name
R

Parameters

NameType
metadataMetadata
fn() => Promise<R>

Returns

Promise<R>

returned value of fn

Inherited from

BaseClient.withMetadata