Add purge support for in-memory persistence

[Copilot]

Describe the change
Long-running hosts using the in-memory persistence provider retain terminated workflows indefinitely, leading to accumulating instance history. This adds purge support so terminal workflows can be cleared from memory as expected.

• Memory purge behavior: Added purge support for terminal (completed/terminated) workflows and their subscriptions in the in-memory store.
• Coverage: Added unit tests that verify purge removes terminal workflows while keeping active ones.

Describe your implementation or design

• Purger wiring: Implemented and registered an in-memory IWorkflowPurger aligned with existing provider patterns.
• Tests: Updated memory persistence tests to use async/await and cover purge scenarios.

await purger.PurgeWorkflows(WorkflowStatus.Terminated, DateTime.UtcNow.AddMinutes(-1), 100);

Tests
Not run (requires .NET 6 runtime for net6.0 targets in this environment).

Breaking change
No.

Additional context
None.

Warning

Firewall rules blocked me from connecting to one or more addresses (expand for details)

I tried to connect to the following addresses, but was blocked by firewall rules:

• www.example.com
  • Triggering command: /opt/oracle/product/26ai/dbhomeFree/bin/oracle oracleFREE (DESCRIPTION=(LOCAL=YES)(ADDRESS=(PROTOCOL=beq))) --bundle /run/containerd/io.containerd.runtime.v2.task/moby/6509a570663d48122afa393102abb172.17.0.4 --pid-file /run/containerd/io.containerd.runtime.v2.task/moby/6509a570663d48122afa393102abb1521 6509a570663d48122afa393102abb66444aa9f52d78a6b18172be33803ad2e6b ep/bin/linux-x64/rg (dns block)

If you need me to access, download, or install something from one of these locations, you can either:

• Configure Actions setup steps to set up my environment, which run before the firewall is enabled
• Add the appropriate URLs or hosts to the custom allowlist in this repository's Copilot coding agent settings (admins only)

Original prompt

---

This section details on the original issue you should resolve

<issue_title>host.PersistenceStore memory always exists after workflow terminates</issue_title>
<issue_description>- I can start work flow host in web api main program by call app.StartHost() method:

var builder = WebApplication.CreateBuilder(args);
Log.Information("Starting web host");
var app = builder.Build();
try
{
  app.StartHost();
  app.Run();
}
finally
{
  Log.CloseAndFlush();
}

• StartHost method implemnts:

public static WebApplication StartHost(this WebApplication app)
{
  var services = app.Services;
  var log = app.Services.GetRequiredService<ILoggerFactory>().CreateLogger("WorkflowStartup");
  var registry = services.GetService<IWorkflowRegistry>();
  var host = app.Services.GetRequiredService<IWorkflowHost>();
  if (registry is not null && !registry.IsRegistered(nameof(StartWorkflow), 1))
  {
    host.RegisterWorkflow<StartWorkflow, WorkflowData<string>>();
    host.RegisterWorkflow<PauseWorkflow, PauseWorkflowData>();
    log.LogInformation("Registered Workflow ");
  }
  LifeCycleEventHandler lifeCycleHandler = async evt =>
  {
    try
    {
      if (evt is WorkflowCore.Models.LifeCycleEvents.WorkflowError stepErr)
      {
        log.LogError(
          "Workflow {Id} step {StepId} error: {Message}",
          stepErr.WorkflowInstanceId,
          stepErr.StepId,
          stepErr.Message
        );
        var z = host.PersistenceStore;
        await host.TerminateWorkflow(evt.WorkflowInstanceId);
      }
    }
    catch
    {
      log.LogError("Error handling life cycle event");
    }
  };

  host.OnLifeCycleEvent += lifeCycleHandler;
  log.LogInformation("Starting WorkflowCore host...");
  host.Start();

  app.Lifetime.ApplicationStopping.Register(() =>
  {
    host.OnLifeCycleEvent -= lifeCycleHandler;
    host.Stop();
  });

• i can start the work flow in the service from DI container as below

    var workflowId = await host.StartWorkflow(
      nameof(StartWorkflow),
      new JobWorkflowData<string>() 
    );

Question:
The host and work flow started normally. it is expected action that when error occurs ,then the workflow was terminated, then i can start the work flow again.
But i found that the inner instances of host persistenceStore are added after per work flow terminated , i have a bit worry about the memory will up with time passes, I want to know how to clear terminated history workflow instances?

below is the memory snapshot ( inner instances of host persistenceStore):

</issue_description>

Comments on the Issue (you are @copilot in this section)

• Fixes host.PersistenceStore memory always exists after workflow terminates #1409

---

💬 We'd love your input! Share your thoughts on Copilot coding agent in our 2 minute survey.