jrmccluskey commented on code in PR #33702: URL: https://github.com/apache/beam/pull/33702#discussion_r1941387833
########## sdks/python/apache_beam/testing/benchmarks/README.md: ########## @@ -0,0 +1,117 @@ +<!-- + Licensed to the Apache Software Foundation (ASF) under one + or more contributor license agreements. See the NOTICE file + distributed with this work for additional information + regarding copyright ownership. The ASF licenses this file + to you under the Apache License, Version 2.0 (the + "License"); you may not use this file except in compliance + with the License. You may obtain a copy of the License at + + http://www.apache.org/licenses/LICENSE-2.0 + + Unless required by applicable law or agreed to in writing, + software distributed under the License is distributed on an + "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY + KIND, either express or implied. See the License for the + specific language governing permissions and limitations + under the License. +--> + +# Writing a Dataflow Cost Benchmark + +Writing a Dataflow Cost Benchmark to estimate the financial cost of executing a pipeline on Google Cloud Platform Dataflow requires 4 components in the repository: + +1. A pipeline to execute (ideally one located in the examples directory) +1. A text file with pipeline options in the `.github/workflows/cost-benchmarks-pipeline-options` [directory](../../../../../.github/workflows/cost-benchmarks-pipeline-options) +1. A test class inheriting from the `DataflowCostBenchmark` [class](../load_tests/dataflow_cost_benchmark.py) +1. An entry to execute the pipeline as part of the cost benchmarks workflow action + +### Choosing a Pipeline +Pipelines that are worth benchmarking in terms of performance and cost have a few straightforward requirements. + +1. The transforms used in the pipeline should be native to Beam *or* be lightweight and readily available in the given pipeline +1. The pipeline itself should run on a consistent data set and have consistent internals (such as model versions for `RunInference` workloads.) Review Comment: This is referring to keeping the same version of a model in a RunInference pipeline rather than doing something like automatically updating to the latest version. A fully specified benchmark should be running on an identical configuration every time, from details like model version and framework all the way up to the GCP region the job runs in. I'll see if I can nail down better wording -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
