This provides a way to pass in hours to be ignored via the command line.

openshift_metrics/invoice.py

@@ -1,7 +1,7 @@
 import math
 from dataclasses import dataclass, field
 from collections import namedtuple
-from typing import List, Tuple
+from typing import List, Tuple, Optional
 from decimal import Decimal, ROUND_HALF_UP
 import datetime
 
@@ -138,7 +138,7 @@ def get_runtime(
     def end_time(self) -> int:
         return self.start_time + self.duration
 
-    def generate_pod_row(self):
+    def generate_pod_row(self, ignore_times):
         """
         This returns a row to represent pod data.
         It converts the epoch_time stamps to datetime timestamps so it's more readable.
@@ -154,7 +154,7 @@ def generate_pod_row(self):
         memory_request = self.memory_request.quantize(
             Decimal(".0001"), rounding=ROUND_HALF_UP
         )
-        runtime = self.get_runtime().quantize(Decimal(".0001"), rounding=ROUND_HALF_UP)
+        runtime = self.get_runtime(ignore_times).quantize(Decimal(".0001"), rounding=ROUND_HALF_UP)
         return [
             self.namespace,
             start_time,
@@ -195,6 +195,7 @@ class ProjectInvoce:
     intitution: str
     institution_specific_code: str
     rates: Rates
+    ignore_hours: Optional[List[Tuple[datetime.datetime, datetime.datetime]]] = None
     su_hours: dict = field(
         default_factory=lambda: {
             SU_CPU: 0,
@@ -210,7 +211,7 @@ class ProjectInvoce:
     def add_pod(self, pod: Pod) -> None:
         """Aggregate a pods data"""
         su_type, su_count, _ = pod.get_service_unit()
-        duration_in_hours = pod.get_runtime()
+        duration_in_hours = pod.get_runtime(self.ignore_hours)
         self.su_hours[su_type] += su_count * duration_in_hours
 
     def get_rate(self, su_type) -> Decimal:

openshift_metrics/merge.py

@@ -5,6 +5,7 @@
 import argparse
 from datetime import datetime
 import json
+from typing import Tuple
 
 import utils
 
@@ -17,6 +18,19 @@ def compare_dates(date_str1, date_str2):
     return date1 < date2
 
 
+def parse_timestamp_range(timestamp_range: str) -> Tuple[datetime, datetime]:
+    try:
+        start_str, end_str = timestamp_range.split(",")
+        start_dt = datetime.fromisoformat(start_str)
+        end_dt = datetime.fromisoformat(end_str)
+        if start_dt < end_dt:
+            raise argparse.ArgumentTypeError("Ignore start time is after ignore end time")
+        return start_dt, end_dt
+    except ValueError:
+        raise argparse.ArgumentTypeError(
+            "Timestamp range must be in the format 'YYYY-MM-DDTHH:MM:SS,YYYY-MM-DDTHH:MM:SS'"
+        )
+
 def main():
     """Reads the metrics from files and generates the reports"""
     parser = argparse.ArgumentParser()
@@ -26,13 +40,21 @@ def main():
         "--upload-to-s3",
         action="store_true"
     )
+    parser.add_argument(
+        "--ignore-hours",
+        type=parse_timestamp_range,
+        nargs="*",
+        help="List of timestamp ranges to ignore in the format 'YYYY-MM-DDTHH:MM:SS,YYYY-MM-DDTHH:MM:SS'"
+    )
+
     args = parser.parse_args()
     files = args.files
 
     if args.output_file:
         output_file = args.output_file
     else:
         output_file = f"{datetime.today().strftime('%Y-%m-%d')}.csv"
+    ignore_hours = args.ignore_hours
 
     report_start_date = None
     report_end_date = None
@@ -77,9 +99,10 @@ def main():
     utils.write_metrics_by_namespace(
         condensed_metrics_dict,
         output_file,
-        report_month
+        report_month,
+        ignore_hours,
     )
-    utils.write_metrics_by_pod(condensed_metrics_dict, "pod-" + output_file)
+    utils.write_metrics_by_pod(condensed_metrics_dict, "pod-" + output_file, ignore_hours)
 
     if args.upload_to_s3:
         primary_location = (

openshift_metrics/utils.py

@@ -110,7 +110,7 @@ def csv_writer(rows, file_name):
         csvwriter.writerows(rows)
 
 
-def write_metrics_by_namespace(condensed_metrics_dict, file_name, report_month):
+def write_metrics_by_namespace(condensed_metrics_dict, file_name, report_month, ignore_hours=None):
     """
     Process metrics dictionary to aggregate usage by namespace and then write that to a file
     """
@@ -157,7 +157,8 @@ def write_metrics_by_namespace(condensed_metrics_dict, file_name, report_month):
                 invoice_address="",
                 intitution="",
                 institution_specific_code=cf_institution_code,
-                rates=rates
+                rates=rates,
+                ignore_hours=ignore_hours,
             )
             invoices[namespace] = project_invoice
 
@@ -186,7 +187,7 @@ def write_metrics_by_namespace(condensed_metrics_dict, file_name, report_month):
     csv_writer(rows, file_name)
 
 
-def write_metrics_by_pod(condensed_metrics_dict, file_name):
+def write_metrics_by_pod(condensed_metrics_dict, file_name, ignore_hours=None):
     """
     Generates metrics report by pod.
     """
@@ -227,6 +228,6 @@ def write_metrics_by_pod(condensed_metrics_dict, file_name):
                     node_hostname=pod_metric_dict.get("node", "Unknown Node"),
                     node_model=pod_metric_dict.get("node_model", "Unknown Model"),
                 )
-                rows.append(pod_obj.generate_pod_row())
+                rows.append(pod_obj.generate_pod_row(ignore_hours))
 
     csv_writer(rows, file_name)