logs added

deployment/nbeats/main.py

@@ -4,6 +4,7 @@ import stomp
 import json
 from amq_message_python_library import *  # python amq-message-python-library
 import logging
+import time
 
 AMQ_USER = os.environ.get("AMQ_USER", "admin")
 AMQ_PASSWORD = os.environ.get("AMQ_PASSWORD", "admin")
@@ -97,12 +98,11 @@ def main():
     # msg1 = Msg()
     # msg1.body = '[{"metric": "cpu_usage", "level": 3, "publish_rate": 60000}]'
     # msg2 = Msg()
-    # msg2.body = """{
-    #     "metrics": ["cpu_usage"],
-    #     "timestamp": 0,
-    #     "epoch_start": 0,
-    #     "number_of_forward_predictions": 8,
-    #     "prediction_horizon": 60}"""
+    # msg2.body = (
+    #     "{"
+    #     + f'"metrics": ["cpu_usage"],"timestamp": {int(time.time())}, "epoch_start": {int(time.time()) + 30}, "number_of_forward_predictions": 8,"prediction_horizon": 60'
+    #     + "}"
+    # )
 
     # StartListener(start_conn.conn, START_APP_TOPIC).on_message(msg1)
     # StartForecastingListener(start_conn.conn, START_APP_TOPIC).on_message(msg2)

deployment/nbeats/predict.py

@@ -112,9 +112,7 @@ def main():
         logging.debug("prediction")
         dataset_preprocessor.prepare_csv()
         global time_0
-        time_1 = time_0
         print("time 00000000 ", time_0)
-        time_0 = time_0 + prediction_cycle
         for metric in predicted_metrics:
             predictions = None
             for i in range(number_of_forward_predictions[metric]):
@@ -128,7 +126,7 @@ def main():
                 )
                 if i == (number_of_forward_predictions[metric] - 1):
                     print(
-                        f"time_0 difference seconds  {start_time + (i + 1) *prediction_horizon // 1000  - int(time.time())}"
+                        f"time_0 difference seconds  {start_time + (i + 1) * prediction_horizon // 1000  - int(time.time())}"
                     )
                 if predictions is not None:
                     predictions = pd.concat(
@@ -139,11 +137,18 @@ def main():
 
                 if prediction_msgs:
                     dest = f"{PRED_TOPIC_PREF}.{metric}"
+                    print(
+                        f'{int(prediction_msgs[metric]["predictionTime"]) - int(prediction_msgs[metric]["timestamp"])} difference between timestamp and predicted in secnds'
+                    )
+                    print(
+                        f'{int(prediction_msgs[metric]["predictionTime"]) - int(time.time())} difference between current time and predicted in secnds'
+                    )
                     start_conn.send_to_topic(dest, prediction_msgs[metric])
                     influxdb_conn.send_to_influxdb(metric, prediction_msgs)
 
         end_time = int(time.time())
         print(f"sleeping {prediction_cycle - (end_time - start_time)} seconds")
+        time_0 = time_0 + prediction_cycle
         time.sleep(prediction_cycle - (end_time - start_time))
 
 

deployment/tft/predict.py

@@ -113,7 +113,6 @@ def main():
         dataset_preprocessor.prepare_csv()
         global time_0
         print("time 00000000 ", time_0)
-        time_0 = time_0 + prediction_cycle
         for metric in predicted_metrics:
             predictions = None
             for i in range(number_of_forward_predictions[metric]):
@@ -127,7 +126,7 @@ def main():
                 )
                 if i == (number_of_forward_predictions[metric] - 1):
                     print(
-                        f"time_0 difference seconds  {start_time + (i + 1) *prediction_horizon // 1000  - int(time.time())}"
+                        f"time_0 difference seconds  {start_time + (i + 1) * prediction_horizon // 1000  - int(time.time())}"
                     )
                 if predictions is not None:
                     predictions = pd.concat(
@@ -139,13 +138,17 @@ def main():
                 if prediction_msgs:
                     dest = f"{PRED_TOPIC_PREF}.{metric}"
                     print(
-                        f'{int(prediction_msgs[metric]["predictionTime"]) - int(prediction_msgs[metric]["timestamp"])} difference between current and predicted in secnds'
+                        f'{int(prediction_msgs[metric]["predictionTime"]) - int(prediction_msgs[metric]["timestamp"])} difference between timestamp and predicted in secnds'
+                    )
+                    print(
+                        f'{int(prediction_msgs[metric]["predictionTime"]) - int(time.time())} difference between current time and predicted in secnds'
                     )
                     start_conn.send_to_topic(dest, prediction_msgs[metric])
                     influxdb_conn.send_to_influxdb(metric, prediction_msgs)
 
         end_time = int(time.time())
         print(f"sleeping {prediction_cycle - (end_time - start_time)} seconds")
+        time_0 = time_0 + prediction_cycle
         time.sleep(prediction_cycle - (end_time - start_time))