Add opt-in Timeout to PurgeInstancesFilter for partial purge

[YunchuWang]

Summary

Add opt-in TimeSpan? Timeout property to PurgeInstancesFilter and send it as google.protobuf.Duration in the gRPC proto request.

Changes

• PurgeInstancesFilter: Add TimeSpan? Timeout { get; init; } property (default null)
• GrpcDurableTaskClient.PurgeAllInstancesAsync: Send timeout as google.protobuf.Duration in proto when set
• Proto: Add google.protobuf.Duration timeout = 4 to PurgeInstanceFilter message

Usage

// Opt-in — existing callers unaffected
var filter = new PurgeInstancesFilter(from, to) { Timeout = TimeSpan.FromSeconds(25) };
var result = await client.PurgeAllInstancesAsync(filter);
// result.IsComplete == false → call again

Breaking Changes

None. Timeout defaults to null — existing behavior unchanged.

Replaces #679 (closed).

Dependencies

• Proto: Add opt-in timeout field to PurgeInstanceFilter for partial purge durabletask-protobuf#67
• DTFx: Enhance purge with parallel batch deletes and partial purge timeout Azure/durabletask#1321
• Extension: Add partial purge timeout support in gRPC proto mapping Azure/azure-functions-durable-extension#3398

[Copilot]

Pull request overview

This PR introduces an opt-in, per-call timeout for purge operations by extending the client-side PurgeInstancesFilter API and propagating the value through the gRPC contract to enable time-bounded (“partial”) purge behavior when supported by the backend.

Changes:

• Add TimeSpan? Timeout (init-only) to PurgeInstancesFilter.
• Send Timeout as google.protobuf.Duration in GrpcDurableTaskClient.PurgeAllInstancesAsync when provided.
• Extend the gRPC proto PurgeInstanceFilter message with a new timeout field.

Reviewed changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 3 comments.

File	Description
src/Client/Core/PurgeInstancesFilter.cs	Adds the new optional Timeout property and XML documentation.
src/Client/Grpc/GrpcDurableTaskClient.cs	Maps PurgeInstancesFilter.Timeout into the gRPC request message.
src/Grpc/orchestrator_service.proto	Extends PurgeInstanceFilter with google.protobuf.Duration timeout = 4.

---

You can also share your feedback on Copilot code review. Take the survey.

[Copilot]

Pull request overview

Adds an optional timeout to purge operations so callers can request time-bounded (“partial”) purges and loop until completion, improving reliability for large purge workloads over gRPC.

Changes:

• Added TimeSpan? Timeout to PurgeInstancesFilter (opt-in, default null).
• Forwarded Timeout through GrpcDurableTaskClient.PurgeAllInstancesAsync as google.protobuf.Duration, with validation for non-positive values.
• Extended the gRPC proto PurgeInstanceFilter message to include a timeout field, plus added unit tests around timeout validation.

Reviewed changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated 5 comments.

File	Description
test/Client/Grpc.Tests/GrpcDurableTaskClientTests.cs	Adds unit tests for Timeout validation behavior
src/Grpc/orchestrator_service.proto	Adds google.protobuf.Duration timeout = 4 to purge filter message
src/Client/Grpc/GrpcDurableTaskClient.cs	Validates and maps PurgeInstancesFilter.Timeout into the protobuf request
src/Client/Core/PurgeInstancesFilter.cs	Adds the new Timeout property with XML documentation

[Copilot]

Pull request overview

Copilot reviewed 4 out of 4 changed files in this pull request and generated 1 comment.