fix(sdk): remove 3-second timeout from checkpoint queue completion (#368)

Author: ParidelPooya

packages/aws-durable-execution-sdk-js-examples/src/examples/run-in-child-context/checkpoint-size-limit/run-in-child-context-checkpoint-size-limit.test.ts

@@ -6,9 +6,9 @@ createTests({
   name: "run-in-child-context-checkpoint-size-limit boundary test",
   functionName: "run-in-child-context-checkpoint-size-limit",
   handler,
-  // localRunnerConfig: {
-  //   checkpointDelay: 100,
-  // },
+  localRunnerConfig: {
+    checkpointDelay: 100,
+  },
   tests: (runner, { assertEventSignatures }) => {
     it("should handle 100 iterations near checkpoint size limit", async () => {
       const execution = await runner.run();
@@ -20,15 +20,12 @@ createTests({
 
       // Verify totalIterations matches actual operations created
       expect(result.totalIterations).toBe(
-        // TODO: https://github.com/aws/aws-durable-execution-sdk-js/issues/365
-        // execution.getOperations({
-        //   status: "SUCCEEDED",
-        // }).length,
-        execution.getOperations().length,
+        execution.getOperations({
+          status: "SUCCEEDED",
+        }).length,
       );
 
-      // TODO: https://github.com/aws/aws-durable-execution-sdk-js/issues/365
-      // assertEventSignatures(execution.getHistoryEvents(), historyEvents);
+      assertEventSignatures(execution.getHistoryEvents(), historyEvents);
     }, 120000);
   },
 });

packages/aws-durable-execution-sdk-js/src/utils/checkpoint/checkpoint-manager.ts

@@ -41,7 +41,6 @@ export class CheckpointManager implements Checkpoint {
     reject: (error: Error) => void;
   }> = [];
   private queueCompletionResolver: (() => void) | null = null;
-  private queueCompletionTimeout: NodeJS.Timeout | null = null;
   private readonly MAX_PAYLOAD_SIZE = 750 * 1024; // 750KB in bytes
   private isTerminating = false;
   private static textEncoder = new TextEncoder();
@@ -152,17 +151,8 @@ export class CheckpointManager implements Checkpoint {
       return;
     }
 
-    return new Promise<void>((resolve, reject) => {
+    return new Promise<void>((resolve) => {
       this.queueCompletionResolver = resolve;
-
-      // Set a timeout to prevent infinite waiting
-      this.queueCompletionTimeout = setTimeout(() => {
-        this.queueCompletionResolver = null;
-        this.queueCompletionTimeout = null;
-        // Clear the queue since it's taking too long
-        this.clearQueue();
-        reject(new Error("Timeout waiting for checkpoint queue completion"));
-      }, 3000); // 3 second timeout
     });
   }
 
@@ -393,10 +383,6 @@ export class CheckpointManager implements Checkpoint {
 
   private notifyQueueCompletion(): void {
     if (this.queueCompletionResolver) {
-      if (this.queueCompletionTimeout) {
-        clearTimeout(this.queueCompletionTimeout);
-        this.queueCompletionTimeout = null;
-      }
       this.queueCompletionResolver();
       this.queueCompletionResolver = null;
     }

packages/aws-durable-execution-sdk-js/src/utils/checkpoint/checkpoint-queue-completion.test.ts

@@ -67,58 +67,6 @@ describe("CheckpointManager Queue Completion", () => {
 
       await expect(waits).resolves.toEqual([undefined, undefined, undefined]);
     });
-
-    it("should timeout after 3 seconds if queue doesn't complete", async () => {
-      jest.useFakeTimers();
-
-      const mockCheckpoint = jest.fn().mockImplementation(
-        () => new Promise(() => {}), // Never resolves
-      );
-      (checkpointManager as any).storage = { checkpoint: mockCheckpoint };
-
-      // Add item to queue
-      checkpointManager.checkpoint("test-step", {});
-
-      const waitPromise = checkpointManager.waitForQueueCompletion();
-
-      // Fast-forward time by 3 seconds
-      jest.advanceTimersByTime(3000);
-
-      await expect(waitPromise).rejects.toThrow(
-        "Timeout waiting for checkpoint queue completion",
-      );
-
-      jest.useRealTimers();
-    });
-
-    it("should clear queue on timeout", async () => {
-      jest.useFakeTimers();
-
-      const mockCheckpoint = jest.fn().mockImplementation(
-        () => new Promise(() => {}), // Never resolves
-      );
-      (checkpointManager as any).storage = { checkpoint: mockCheckpoint };
-
-      // Add items to queue
-      checkpointManager.checkpoint("test-step-1", {});
-      checkpointManager.checkpoint("test-step-2", {});
-
-      expect((checkpointManager as any).queue.length).toBe(2);
-
-      const waitPromise = checkpointManager.waitForQueueCompletion();
-
-      // Fast-forward time by 3 seconds
-      jest.advanceTimersByTime(3000);
-
-      await expect(waitPromise).rejects.toThrow(
-        "Timeout waiting for checkpoint queue completion",
-      );
-
-      // Queue should be cleared
-      expect((checkpointManager as any).queue.length).toBe(0);
-
-      jest.useRealTimers();
-    });
   });
 
   describe("clearQueue", () => {

packages/aws-durable-execution-sdk-js/src/with-durable-execution-queue-completion.test.ts

@@ -104,37 +104,6 @@ describe("withDurableExecution Queue Completion", () => {
     waitSpy.mockRestore();
   });
 
-  it("should handle waitForQueueCompletion timeout gracefully", async () => {
-    // Mock waitForQueueCompletion to reject with timeout error after 3 seconds
-    const waitSpy = jest
-      .spyOn(CheckpointManager.prototype, "waitForQueueCompletion")
-      .mockImplementation(
-        () =>
-          new Promise((_, reject) =>
-            setTimeout(
-              () =>
-                reject(
-                  new Error("Timeout waiting for checkpoint queue completion"),
-                ),
-              3000,
-            ),
-          ),
-      );
-
-    const mockHandler = jest.fn().mockResolvedValue("success");
-    const wrappedHandler = withDurableExecution(mockHandler);
-
-    const startTime = Date.now();
-    await wrappedHandler(mockEvent, mockContext);
-    const endTime = Date.now();
-
-    // Should complete within timeout period (3 seconds + buffer for test overhead)
-    expect(endTime - startTime).toBeLessThan(7000);
-    expect(waitSpy).toHaveBeenCalled();
-
-    waitSpy.mockRestore();
-  }, 10000);
-
   it("should handle waitForQueueCompletion errors gracefully", async () => {
     const waitSpy = jest
       .spyOn(CheckpointManager.prototype, "waitForQueueCompletion")