Performance Regression or Improvement: gbk_python_batch_load_test_2gb_of_10B_records:runtime

[github-actions]

Performance change found in the
test: gbk_python_batch_load_test_2gb_of_10B_records for the metric: runtime.

For more information on how to triage the alerts, please look at
Triage performance alert issues section of the README.

Test description: GroupByKey Python Load test - 2GB of 10B records
python | GBK | Small records (10B)
Test Dashboard - http://metrics.beam.apache.org/d/UYZ-oJ3Zk/gbk-load-tests?orgId=1&var-processingType=batch&var-sdk=python&from=now-90d&to=now&viewPanel=2
Test link - https://github.com/apache/beam/blob/5e38decf9e723a385057131b01bbd33d8c60bda3/.test-infra/jenkins/job_LoadTests_GBK_Python.groovy#L36C25-L36C72

timestamp: Tue Mar 18 16:22:11 2025, metric_value: 1018.00
timestamp: Mon Mar 17 16:24:00 2025, metric_value: 1014.00
timestamp: Sun Mar 16 16:24:04 2025, metric_value: 1010.00
timestamp: Sat Mar 15 16:24:35 2025, metric_value: 1018.00
timestamp: Fri Mar 14 16:23:50 2025, metric_value: 1020.00
timestamp: Thu Mar 13 16:26:20 2025, metric_value: 941.00
timestamp: Wed Mar 12 16:23:03 2025, metric_value: 937.00
timestamp: Tue Mar 11 16:25:01 2025, metric_value: 972.00 <---- Anomaly
timestamp: Mon Mar 10 16:19:21 2025, metric_value: 660.00
timestamp: Sun Mar  9 16:18:25 2025, metric_value: 647.00
timestamp: Sat Mar  8 16:24:53 2025, metric_value: 958.00
timestamp: Fri Mar  7 16:24:18 2025, metric_value: 937.00
timestamp: Thu Mar  6 16:23:00 2025, metric_value: 948.00
timestamp: Wed Mar  5 16:21:20 2025, metric_value: 924.00
timestamp: Tue Mar  4 16:22:05 2025, metric_value: 953.00
timestamp: Mon Mar  3 16:23:02 2025, metric_value: 913.00
timestamp: Sun Mar  2 16:18:52 2025, metric_value: 662.00
timestamp: Sat Mar  1 16:17:05 2025, metric_value: 659.00

[kennknowles]

There was an unrealistic dip in timing on http://35.193.202.176/d/UYZ-oJ3Zk/gbk-load-tests?orgId=1&var-processingType=batch&var-sdk=python&from=now-90d&to=now&viewPanel=2