Performance Regression or Improvement: gbk_python_batch_load_test_fanout_4_times_with_2GB_10byte_records_total:runtime

[github-actions]

Performance change found in the
test: gbk_python_batch_load_test_fanout_4_times_with_2GB_10byte_records_total 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 - fanout 4 times with 2GB 10-byte records total python | GBK | Fanout 4 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=5 Test link -

beam/.test-infra/jenkins/job_LoadTests_GBK_Python.groovy

Line 106 in 5e38dec

test : 'apache_beam.testing.load_tests.group_by_key_test',

timestamp: Sat Oct  7 21:26:14 2023, metric_value: 54.00
timestamp: Sat Oct  7 09:27:08 2023, metric_value: 53.00
timestamp: Fri Oct  6 21:29:10 2023, metric_value: 58.00
timestamp: Fri Oct  6 09:30:13 2023, metric_value: 61.00
timestamp: Thu Oct  5 21:30:32 2023, metric_value: 58.00
timestamp: Thu Oct  5 17:42:52 2023, metric_value: 56.00 <---- Anomaly
timestamp: Thu Oct  5 10:37:08 2023, metric_value: 373.00
timestamp: Wed Oct  4 14:11:19 2023, metric_value: 54.00
timestamp: Wed Oct  4 10:37:15 2023, metric_value: 365.00
timestamp: Wed Oct  4 04:52:21 2023, metric_value: 354.00
timestamp: Tue Oct  3 14:17:44 2023, metric_value: 355.00
timestamp: Tue Oct  3 14:10:12 2023, metric_value: 58.00
timestamp: Mon Oct  2 14:11:19 2023, metric_value: 92.00
timestamp: Sun Oct  1 14:09:29 2023, metric_value: 54.00
timestamp: Sat Sep 30 14:09:27 2023, metric_value: 55.00
timestamp: Fri Sep 29 14:11:35 2023, metric_value: 47.00

[AnandInguva]

This test itself seems noisy. I think we should disable the alert for this test.

[tvalentyn]

This reads as though the test had a temporary blip on October 4 that is now gone.

[tvalentyn]

this noise would also be removed with #28238