Durable Task Client SDK for Java

This repo contains the Java SDK for the Durable Task Framework as well as classes and annotations to support running Azure Durable Functions for Java. With this SDK, you can define, schedule, and manage durable orchestrations using ordinary Java code.

// *** Simple, fault-tolerant, sequential orchestration ***
String result = "";
result += ctx.callActivity("SayHello", "Tokyo", String.class).await() + ", ";
result += ctx.callActivity("SayHello", "London", String.class).await() + ", ";
result += ctx.callActivity("SayHello", "Seattle", String.class).await();
return result;

Orchestrator code re-executes while rebuilding state from history. Wrap an existing java.util.logging.Logger to suppress log output during those replay segments:

Logger logger = ctx.createReplaySafeLogger(
    Logger.getLogger(MyOrchestration.class.getName()));

logger.info(() -> "Starting orchestration " + ctx.getInstanceId());
String result = ctx.callActivity("ProcessItem", input, String.class).await();
logger.info(() -> "Activity returned: " + result);

In Azure Functions, pass ExecutionContext.getLogger() instead of creating a named logger so the output retains its invocation ID and normal host routing:

Logger logger = ctx.createReplaySafeLogger(executionContext.getLogger());

Replay-safe logging suppresses calls made while replaying; it does not guarantee exactly-once log delivery across failed or retried live orchestration turns.

// Get the list of work-items to process
List<?> batch = ctx.callActivity("GetWorkBatch", List.class).await();

// Schedule each task to run in parallel
List<Task<Integer>> parallelTasks = batch.stream()
        .map(item -> ctx.callActivity("ProcessItem", item, Integer.class))
        .collect(Collectors.toList());

// Wait for all tasks to complete, then return the aggregated sum of the results
List<Integer> results = ctx.allOf(parallelTasks).await();
return results.stream().reduce(0, Integer::sum);

ApprovalInfo approvalInfo = ctx.getInput(ApprovalInfo.class);
ctx.callActivity("RequestApproval", approvalInfo).await();

Duration timeout = Duration.ofHours(72);
try {
    // Wait for an approval. A TaskCanceledException will be thrown if the timeout expires.
    boolean approved = ctx.waitForExternalEvent("ApprovalEvent", timeout, boolean.class).await();
    approvalInfo.setApproved(approved);

    ctx.callActivity("ProcessApproval", approvalInfo).await();
} catch (TaskCanceledException timeoutEx) {
    ctx.callActivity("Escalate", approvalInfo).await();
}

JobInfo jobInfo = ctx.getInput(JobInfo.class);
String jobId = jobInfo.getJobId();

String status = ctx.callActivity("GetJobStatus", jobId, String.class).await();
if (status.equals("Completed")) {
    // The job is done - we can exit now
    ctx.callActivity("SendAlert", jobId).await();
} else {
    // wait N minutes before doing the next poll
    Duration pollingDelay = jobInfo.getPollingDelay();
    ctx.createTimer(pollingDelay).await();

    // restart from the beginning
    ctx.continueAsNew(jobInfo);
}

return null;

The following packages are produced from this repo.

Package	Latest version
Durable Task - Client
Durable Task - Azure Functions

The com.microsoft:durabletask-azuremanaged package configures clients and workers through DurableTaskSchedulerClientOptions, DurableTaskSchedulerWorkerOptions, or the corresponding DurableTaskSchedulerClientExtensions and DurableTaskSchedulerWorkerExtensions convenience methods.

Set resourceId using setResourceId(...) on either options class, the optional last argument of the createClientBuilder, createWorkerBuilder, and useDurableTaskScheduler overloads, or ResourceId in a connection string. This is a token audience URI, not an Azure Resource Manager resource path. Existing overloads remain supported.

Default behavior change: applications running in US Government or DoD regions now select the government audience when no explicit audience is provided. Set ResourceId=https://durabletask.io to retain the public audience in those regions. Prefixes, not substrings, are matched: chinaeast2, notusgov, and notusdod still use the public default. No audience is inferred from the endpoint.

Explicit values have surrounding whitespace (including Unicode whitespace such as em spaces and non-breaking spaces) and trailing / characters removed, then one existing /.default suffix removed case-insensitively, followed by any remaining trailing / characters. URI casing is otherwise preserved. For example, https://durabletask.azure.us//.DEFAULT// requests https://durabletask.azure.us/.default, and api://CustomAudience/resource/.DEFAULT/ requests api://CustomAudience/resource/.default. Whitespace-only input, ///, /.default, and /.DEFAULT/// throw IllegalArgumentException; use an omitted or genuinely empty value for the default.

Defaults are resolved per options instance or parsed connection string, rather than at class initialization. Setting a null or empty audience explicitly resolves the default again at that point. The selected audience is retained when creating channels, refreshing tokens, and reconnecting. Connection-string conversion does not normalize the audience again.

The audience, credential authority/cloud, and service endpoint are independent settings. Neither resourceId nor REGION_NAME changes the endpoint or credential authority. For an already-created TokenCredential, configure authority on that credential; token requests do not override it.

import com.azure.core.credential.TokenCredential;
import com.azure.identity.AzureAuthorityHosts;
import com.azure.identity.DefaultAzureCredentialBuilder;
import com.microsoft.durabletask.DurableTaskGrpcClientBuilder;
import com.microsoft.durabletask.DurableTaskGrpcWorkerBuilder;
import com.microsoft.durabletask.azuremanaged.DurableTaskSchedulerClientExtensions;
import com.microsoft.durabletask.azuremanaged.DurableTaskSchedulerWorkerExtensions;

// Set these to your scheduler's actual endpoint and task hub.
String endpoint = System.getenv("DTS_ENDPOINT");
String taskHub = System.getenv("DTS_TASK_HUB");
TokenCredential credential = new DefaultAzureCredentialBuilder()
    .authorityHost(AzureAuthorityHosts.AZURE_GOVERNMENT)
    .build();

DurableTaskGrpcClientBuilder clientBuilder =
    DurableTaskSchedulerClientExtensions.createClientBuilder(
        endpoint, taskHub, credential, "https://durabletask.azure.us");
DurableTaskGrpcWorkerBuilder workerBuilder =
    DurableTaskSchedulerWorkerExtensions.createWorkerBuilder(
        endpoint, taskHub, credential, "https://durabletask.azure.us");

When the SDK constructs the credential from a connection string, use the optional AuthorityHost property:

Endpoint=<your-scheduler-endpoint>;TaskHub=<your-task-hub>;Authentication=DefaultAzure;ResourceId=https://durabletask.azure.us;AuthorityHost=https://login.microsoftonline.us/

AuthorityHost is forwarded to Azure Identity for DefaultAzure, Environment, WorkloadIdentity, and InteractiveBrowser authentication. Omission or an empty value leaves Azure Identity's defaults intact, including AZURE_AUTHORITY_HOST where supported. It is not an authority override on client or worker options.

Managed identity uses the hosting environment's identity endpoint; an Entra authority override does not apply. Developer-tool credentials (AzureCli, AzurePowerShell, VisualStudioCode, and IntelliJ) use those tools' cloud configuration, not the connection string's AuthorityHost. Configure them separately, including when they are used by DefaultAzureCredential (for example, az cloud set --name AzureUSGovernment before signing in with Azure CLI). Authentication=None remains anonymous.

For information about how to get started with Durable Functions for Java, see the Azure Functions README.md content.

This project welcomes contributions and suggestions. Most contributions require you to agree to a Contributor License Agreement (CLA) declaring that you have the right to, and actually do, grant us the rights to use your contribution. For details, visit https://cla.opensource.microsoft.com.

When you submit a pull request, a CLA bot will automatically determine whether you need to provide a CLA and decorate the PR appropriately (e.g., status check, comment). Simply follow the instructions provided by the bot. You will only need to do this once across all repos using our CLA.

This project has adopted the Microsoft Open Source Code of Conduct. For more information see the Code of Conduct FAQ or contact opencode@microsoft.com with any additional questions or comments.

This project may contain trademarks or logos for projects, products, or services. Authorized use of Microsoft trademarks or logos is subject to and must follow Microsoft's Trademark & Brand Guidelines. Use of Microsoft trademarks or logos in modified versions of this project must not cause confusion or imply Microsoft sponsorship. Any use of third-party trademarks or logos are subject to those third-party's policies.