[WI18] Fix FlinkJobManagerUrl to use fixed port 8081 instead of dynamic Docker port

Root cause: FlinkJobManagerUrl was being set from Docker discovery (flinkRestApi)
- Docker discovery returns dynamic ports (e.g., 127.0.0.1:32769)
- Aspire now configured to use fixed port 8081 for JobManager
- Should use fixed port directly instead of discovering from Docker

Changes:
- Changed FlinkJobManagerUrl from flinkRestApi to "http://localhost:8081/"
- Matches Aspire configuration with fixed port 8081
- Ensures FLINK_JOBMANAGER_URL environment variable passed to exercises uses consistent endpoint

This fixes Day01 health check failures by using the correct fixed JobManager port.

Co-authored-by: devstress <30769729+devstress@users.noreply.github.com>

Author: Copilot

LearningCourse/LearningCourse.IntegrationTests/LearningCourseTestBase.cs

@@ -391,8 +391,8 @@ private static async Task WaitForInfrastructureReadyAsync(DateTime appHostStartT
                 PrometheusHostEndpoint = prometheusEndpoint;
                 GrafanaHostEndpoint = grafanaEndpoint;
                 FlinkRestApiEndpoint = flinkRestApi;
-                FlinkJobGatewayUrl = "http://localhost:8080/";  // Fixed port for JobGateway (not dynamic like JobManager)
-                FlinkJobManagerUrl = flinkRestApi;  // Dynamic Docker port for Flink JobManager health checks
+                FlinkJobGatewayUrl = "http://localhost:8080/";  // Fixed port for JobGateway (not dynamic)
+                FlinkJobManagerUrl = "http://localhost:8081/";  // Fixed port for JobManager (configured in Aspire)
 
                 var savedTime = (maxWait - stopwatch.Elapsed).TotalSeconds;
                 TestContext.WriteLine($"✅ All required infrastructure ready after {stopwatch.Elapsed.TotalSeconds:F1}s (saved {savedTime:F1}s with optimized polling)");