Closed kostmo closed 5 years ago
Example test case is "caffe2_py2_clang3_8_ubuntu16_04_build" build here: https://dr.pytorch.org/master-timeline.html?count=14&sha1=a625734f6aa56ff7bd5119bbedcdc06d53ebe942&should_suppress_scheduled_builds=true&should_suppress_fully_successful_columns=true&max_columns_suppress_successful=50
Denser example: https://dr.pytorch.org/master-timeline.html?count=8&sha1=e66e00cd1700df179816a1555f28691cdec029c1&should_suppress_scheduled_builds=true&should_suppress_fully_successful_columns=true&max_columns_suppress_successful=50
I'm now using a more efficient "edge-detection"-based query; when a job has missing builds across one or more commits, the "state" of that job at any commit is the most recently observed build result.
Example test case is "caffe2_py2_clang3_8_ubuntu16_04_build" build here: https://dr.pytorch.org/master-timeline.html?count=14&sha1=a625734f6aa56ff7bd5119bbedcdc06d53ebe942&should_suppress_scheduled_builds=true&should_suppress_fully_successful_columns=true&max_columns_suppress_successful=50
Denser example: https://dr.pytorch.org/master-timeline.html?count=8&sha1=e66e00cd1700df179816a1555f28691cdec029c1&should_suppress_scheduled_builds=true&should_suppress_fully_successful_columns=true&max_columns_suppress_successful=50