[WIP] add traffic replay feature to experiment service

[adriangudas]

Note: Not suitable for production. Just testing and experimenting. This may form the basis for a real PR later

Example output:

adriangudas@Mac experiments % ruby example_with_traffic_replay.rb sample_traffic_log.json
adriangudas@Adrians-MacBook-Pro experiments % ruby example_with_traffic_replay.rb sample_traffic_log.json
=== Semian ExperimentalResource - Traffic Replay Example ===

Example: Using traffic replay from Grafana export
------------------------------------------------------------
Traffic replay mode enabled. Timeline duration: 12.0s with 25 requests

Resource created successfully!
Starting to process requests...
Press Ctrl+C to stop

I, [2025-10-15T10:59:03.915914 #22481]  INFO -- : Semian Resource my_service is running in log-mode for configuration validation. This means that Semian will not raise an error if the configuration is invalid. This is not recommended for production environments.

[IMPORTANT] PLEASE UPDATE YOUR CONFIGURATION TO USE `force_config_validation: true` TO ENABLE STRICT CONFIGURATION VALIDATION.
---

I, [2025-10-15T10:59:03.915966 #22481]  INFO -- : Semian sysv semaphores are not supported on arm64-darwin25 - all operations will no-op
[10:59:03] Request #1 - Latency: 2.5ms
[10:59:04] Request #2 - Latency: 2.5ms
[10:59:04] Request #3 - Latency: 2.5ms
[10:59:04] Request #4 - Latency: 3.2ms
[10:59:04] Request #5 - Latency: 3.2ms
[10:59:04] Request #6 - Latency: 3.2ms
[10:59:04] Request #7 - Latency: 3.2ms
[10:59:04] Request #8 - Latency: 2.8ms
[10:59:04] Request #9 - Latency: 2.8ms
[10:59:04] Request #10 - Latency: 2.8ms
[10:59:05] Request #11 - Latency: 2.8ms
[10:59:05] Request #12 - Latency: 2.8ms
[10:59:05] Request #13 - Latency: 5.1ms
[10:59:05] Request #14 - Latency: 5.1ms
[10:59:05] Request #15 - Latency: 5.1ms
[10:59:05] Request #16 - Latency: 5.1ms
[10:59:05] Request #17 - Latency: 5.1ms
[10:59:05] Request #18 - Latency: 8.5ms
[10:59:05] Request #19 - Latency: 8.5ms
[10:59:06] Request #20 - Latency: 8.5ms
[10:59:06] Request #21 - Latency: 8.5ms
[10:59:06] Request #22 - Latency: 15.3ms
[10:59:06] Request #23 - Latency: 15.3ms
[10:59:06] Request #24 - Latency: 15.3ms
[10:59:06] Request #25 - Latency: 15.3ms
[10:59:06] Request #26 - Latency: 25.7ms
[10:59:06] Request #27 - Latency: 25.7ms
[10:59:07] Request #28 - Latency: 25.7ms
[10:59:07] Request #29 - Latency: 25.7ms
[10:59:07] Request #30 - Latency: 45.2ms
[10:59:07] Request #31 - Latency: 45.2ms
[10:59:07] Request #32 - Latency: 45.2ms
[10:59:07] Request #33 - Latency: 78.4ms
[10:59:07] Request #34 - Latency: 78.4ms
[10:59:08] Request #35 - Latency: 78.4ms
[10:59:08] Timeout: Request timed out after 0.1s (would have taken 0.126s)
[10:59:09] Timeout: Request timed out after 0.1s (would have taken 0.246s)
I, [2025-10-15T10:59:10.599327 #22481]  INFO -- : [Semian::CircuitBreaker] State transition from closed to open. success_count=0 error_count=3 success_count_threshold=2 error_count_threshold=3 error_timeout=0.2 error_last_at="583143.281439" name="my_service" last_error_message="Request timed out after 0.1s (would have taken 0.313s)"
[10:59:10] Timeout: Request timed out after 0.1s (would have taken 0.313s)
I, [2025-10-15T10:59:11.604952 #22481]  INFO -- : [Semian::CircuitBreaker] State transition from open to half_open. success_count=0 error_count=3 success_count_threshold=2 error_count_threshold=3 error_timeout=0.2 error_last_at="583143.281439" name="my_service"
I, [2025-10-15T10:59:11.710493 #22481]  INFO -- : [Semian::CircuitBreaker] State transition from half_open to open. success_count=0 error_count=3 success_count_threshold=2 error_count_threshold=3 error_timeout=0.2 error_last_at="583144.392658" name="my_service" last_error_message="Request timed out after 0.1s (would have taken 0.234s)"
[10:59:11] Timeout: Request timed out after 0.1s (would have taken 0.234s)
I, [2025-10-15T10:59:12.715998 #22481]  INFO -- : [Semian::CircuitBreaker] State transition from open to half_open. success_count=0 error_count=3 success_count_threshold=2 error_count_threshold=3 error_timeout=0.2 error_last_at="583144.392658" name="my_service"
[10:59:12] Request #36 - Latency: 82.1ms
[10:59:12] Request #37 - Latency: 82.1ms
I, [2025-10-15T10:59:12.991463 #22481]  INFO -- : [Semian::CircuitBreaker] State transition from half_open to closed. success_count=2 error_count=3 success_count_threshold=2 error_count_threshold=3 error_timeout=0.2 error_last_at="583144.392658" name="my_service"
[10:59:13] Request #38 - Latency: 82.1ms
[10:59:13] Request #39 - Latency: 45.7ms
[10:59:13] Request #40 - Latency: 45.7ms
[10:59:13] Request #41 - Latency: 45.7ms
[10:59:13] Request #42 - Latency: 25.3ms
[10:59:13] Request #43 - Latency: 25.3ms
[10:59:14] Request #44 - Latency: 25.3ms
[10:59:14] Request #45 - Latency: 25.3ms
[10:59:14] Request #46 - Latency: 12.8ms
[10:59:14] Request #47 - Latency: 12.8ms
[10:59:14] Request #48 - Latency: 12.8ms
[10:59:14] Request #49 - Latency: 12.8ms
[10:59:14] Request #50 - Latency: 6.5ms
[10:59:14] Request #51 - Latency: 6.5ms
[10:59:15] Request #52 - Latency: 6.5ms
[10:59:15] Request #53 - Latency: 6.5ms
[10:59:15] Request #54 - Latency: 3.8ms
[10:59:15] Request #55 - Latency: 3.8ms
[10:59:15] Request #56 - Latency: 3.8ms
[10:59:15] Request #57 - Latency: 3.8ms
[10:59:15] Request #58 - Latency: 3.8ms
[10:59:15] Request #59 - Latency: 2.7ms
[10:59:15] Request #60 - Latency: 2.7ms

=== Traffic replay completed ===
Service has been running longer than the traffic log timeline.
No more requests will be processed.

Traffic replay completed!
Total requests processed: 60

=== Replay Complete ===