diff --git a/cron/README.md b/cron/README.md new file mode 100644 index 0000000..787dc4c --- /dev/null +++ b/cron/README.md @@ -0,0 +1,50 @@ +# Automating Benchmarking + +This directory contains a bash script and asv configuration files for automating benchmarking for dask and dask-distributed. The script takes advantage of an asv feature that allows it to benchmark all commits since the last benchmark. If no new commits were found for both repositories, the script will just exit. If commits were found for at least one, it will benchmark, generate html files, and publish to the gh-pages branch. + +Note that asv does have a built in feature for doing all the steps to publish to gh-pages. This was not used as we are benchmarking two packages and to achieve a custom url structure. + +Also note that distributed has a dependency on dask. The master branch for distributed has a strong dependency on dask master. This requires pip installing the dask dependency from github. As this is not directly supported by asv, a small hack is required as commented in the script. + +## Setting up new machine (for CentOS, adapt as needed) + +Install requirements: + +``` +sudo yum update +sudo yum upgrade +sudo yum install wget git gcc gcc-c++ bzip2 +git clone git@github.com:dask/dask-benchmarks.git +wget https://repo.continuum.io/miniconda/Miniconda3-latest-Linux-x86_64.sh +bash Miniconda3-latest-Linux-x86_64.sh # install to default location, yes to append path in bashrc +source ~/.bashrc +``` + +Generate key and add to your github ssh keys: + +``` +ssh-keygen -t rsa -b 4096 +cat .ssh/id_rsa.pub +``` + +Set git username and email + +``` +git config --global user.name "My Name" +git config --global user.email "me@email.com" +``` + +## Configuring benchmark script + +The script pulls down the latest dask-benchmarks from the repository. If you want to autodeploy changes to the asv config files and the script, you can run the script directly from the cloned repository. Otherwise, copy the files to another location and configure cron accordingly. + +The script needs to know the location of the config files and the benchmark clone. It defaults to the common directory structure for an AWS EC2 instance but you can override by setting environment variables for `BENCHMARK_REPOSITORY`, `DASK_ASV_CONFIG`, `DISTRIBUTED_ASV_CONFIG`. + +## Configuring cron + +Run `crontab -e` and add the following line: + +``` +0 12 * * * /path/to/run_benchmarks.sh > /path/to/benchmarking.log 2>&1 +``` + diff --git a/cron/asv.dask.conf.json b/cron/asv.dask.conf.json new file mode 100644 index 0000000..1ff5c69 --- /dev/null +++ b/cron/asv.dask.conf.json @@ -0,0 +1,146 @@ +{ + // The version of the config file format. Do not change, unless + // you know what you are doing. + "version": 1, + + // The name of the project being benchmarked + "project": "dask", + + // The project's homepage + "project_url": "http://dask.pydata.org/", + + // The URL or local path of the source code repository for the + // project being benchmarked + "repo": "https://github.com/dask/dask.git", + + // List of branches to benchmark. If not provided, defaults to "master" + // (for git) or "default" (for mercurial). + "branches": ["master"], // for git + // "branches": ["default"], // for mercurial + + // The DVCS being used. If not set, it will be automatically + // determined from "repo" by looking at the protocol in the URL + // (if remote), or by looking for special directories, such as + // ".git" (if local). + "dvcs": "git", + + // The tool to use to create environments. May be "conda", + // "virtualenv" or other value depending on the plugins in use. + // If missing or the empty string, the tool will be automatically + // determined by looking for tools on the PATH environment + // variable. + "environment_type": "conda", + + // timeout in seconds for installing any dependencies in environment + // defaults to 10 min + //"install_timeout": 600, + + // the base URL to show a commit for the project. + // "show_commit_url": "http://github.com/dask/dask/", + + // The Pythons you'd like to test against. If not provided, defaults + // to the current version of Python used to run `asv`. + "pythons": ["2.7", "3.5"], + + // The matrix of dependencies to test. Each key is the name of a + // package (in PyPI) and the values are version numbers. An empty + // list or empty string indicates to just test against the default + // (latest) version. null indicates that the package is to not be + // installed. If the package to be tested is only available from + // PyPi, and the 'environment_type' is conda, then you can preface + // the package name by 'pip+', and the package will be installed via + // pip (with all the conda available packages installed first, + // followed by the pip installed packages). + // + "matrix": { + "numpy": [], + "toolz": [], + "cloudpickle": [], + "partd": [], + "pandas": [], + "distributed": [], + "s3fs": [], + "pytables": [] + }, + + // Combinations of libraries/python versions can be excluded/included + // from the set to test. Each entry is a dictionary containing additional + // key-value pairs to include/exclude. + // + // An exclude entry excludes entries where all values match. The + // values are regexps that should match the whole string. + // + // An include entry adds an environment. Only the packages listed + // are installed. The 'python' key is required. The exclude rules + // do not apply to includes. + // + // In addition to package names, the following keys are available: + // + // - python + // Python version, as in the *pythons* variable above. + // - environment_type + // Environment type, as above. + // - sys_platform + // Platform, as in sys.platform. Possible values for the common + // cases: 'linux2', 'win32', 'cygwin', 'darwin'. + // + // "exclude": [ + // {"python": "3.2", "sys_platform": "win32"}, // skip py3.2 on windows + // {"environment_type": "conda", "six": null}, // don't run without six on conda + // ], + // + // "include": [ + // // additional env for python2.7 + // {"python": "2.7", "numpy": "1.8"}, + // // additional env if run on windows+conda + // {"platform": "win32", "environment_type": "conda", "python": "2.7", "libpython": ""}, + // ], + + // The directory (relative to the current directory) that benchmarks are + // stored in. If not provided, defaults to "benchmarks" + "benchmark_dir": "benchmarks", + + // The directory (relative to the current directory) to cache the Python + // environments in. If not provided, defaults to "env" + "env_dir": ".asv/env", + + // The directory (relative to the current directory) that raw benchmark + // results are stored in. If not provided, defaults to "results". + "results_dir": "/home/ec2-user/results/dask", + + // The directory (relative to the current directory) that the html tree + // should be written to. If not provided, defaults to "html". + "html_dir": "/home/ec2-user/html/dask", + + // The number of characters to retain in the commit hashes. + // "hash_length": 8, + + // `asv` will cache wheels of the recent builds in each + // environment, making them faster to install next time. This is + // number of builds to keep, per environment. + "wheel_cache_size": 1 + + // The commits after which the regression search in `asv publish` + // should start looking for regressions. Dictionary whose keys are + // regexps matching to benchmark names, and values corresponding to + // the commit (exclusive) after which to start looking for + // regressions. The default is to start from the first commit + // with results. If the commit is `null`, regression detection is + // skipped for the matching benchmark. + // + // "regressions_first_commits": { + // "some_benchmark": "352cdf", // Consider regressions only after this commit + // "another_benchmark": null, // Skip regression detection altogether + // } + + // The thresholds for relative change in results, after which `asv + // publish` starts reporting regressions. Dictionary of the same + // form as in ``regressions_first_commits``, with values + // indicating the thresholds. If multiple entries match, the + // maximum is taken. If no entry matches, the default is 5%. + // + // "regressions_thresholds": { + // "some_benchmark": 0.01, // Threshold of 1% + // "another_benchmark": 0.5, // Threshold of 50% + // } +} diff --git a/cron/asv.distributed.conf.json b/cron/asv.distributed.conf.json new file mode 100644 index 0000000..82a06a9 --- /dev/null +++ b/cron/asv.distributed.conf.json @@ -0,0 +1,147 @@ +{ + // The version of the config file format. Do not change, unless + // you know what you are doing. + "version": 1, + + // The name of the project being benchmarked + "project": "distributed", + + // The project's homepage + "project_url": "https://distributed.readthedocs.io/en/latest/", + + // The URL or local path of the source code repository for the + // project being benchmarked + "repo": "https://github.com/dask/distributed", + + // List of branches to benchmark. If not provided, defaults to "master" + // (for git) or "default" (for mercurial). + "branches": ["master"], // for git + // "branches": ["default"], // for mercurial + + // The DVCS being used. If not set, it will be automatically + // determined from "repo" by looking at the protocol in the URL + // (if remote), or by looking for special directories, such as + // ".git" (if local). + "dvcs": "git", + + // The tool to use to create environments. May be "conda", + // "virtualenv" or other value depending on the plugins in use. + // If missing or the empty string, the tool will be automatically + // determined by looking for tools on the PATH environment + // variable. + "environment_type": "conda", + + // timeout in seconds for installing any dependencies in environment + // defaults to 10 min + //"install_timeout": 600, + + // the base URL to show a commit for the project. + // "show_commit_url": "http://github.com/dask/dask/", + + // The Pythons you'd like to test against. If not provided, defaults + // to the current version of Python used to run `asv`. + "pythons": ["2.7", "3.5"], + + // The matrix of dependencies to test. Each key is the name of a + // package (in PyPI) and the values are version numbers. An empty + // list or empty string indicates to just test against the default + // (latest) version. null indicates that the package is to not be + // installed. If the package to be tested is only available from + // PyPi, and the 'environment_type' is conda, then you can preface + // the package name by 'pip+', and the package will be installed via + // pip (with all the conda available packages installed first, + // followed by the pip installed packages). + // + "matrix": { + "numpy": [], + "toolz": [], + "cloudpickle": [], + "partd": [], + "pandas": [], + "s3fs": [], + "pip+blosc": [], + "pip+lz4": [], + "pip+git+https://github.com/dask/dask.git": [], + }, + + // Combinations of libraries/python versions can be excluded/included + // from the set to test. Each entry is a dictionary containing additional + // key-value pairs to include/exclude. + // + // An exclude entry excludes entries where all values match. The + // values are regexps that should match the whole string. + // + // An include entry adds an environment. Only the packages listed + // are installed. The 'python' key is required. The exclude rules + // do not apply to includes. + // + // In addition to package names, the following keys are available: + // + // - python + // Python version, as in the *pythons* variable above. + // - environment_type + // Environment type, as above. + // - sys_platform + // Platform, as in sys.platform. Possible values for the common + // cases: 'linux2', 'win32', 'cygwin', 'darwin'. + // + // "exclude": [ + // {"python": "3.2", "sys_platform": "win32"}, // skip py3.2 on windows + // {"environment_type": "conda", "six": null}, // don't run without six on conda + // ], + // + // "include": [ + // // additional env for python2.7 + // {"python": "2.7", "numpy": "1.8"}, + // // additional env if run on windows+conda + // {"platform": "win32", "environment_type": "conda", "python": "2.7", "libpython": ""}, + // ], + + // The directory (relative to the current directory) that benchmarks are + // stored in. If not provided, defaults to "benchmarks" + "benchmark_dir": "/home/ec2-user/dask-benchmarks/distributed/benchmarks", + + // The directory (relative to the current directory) to cache the Python + // environments in. If not provided, defaults to "env" + "env_dir": ".asv/env", + + // The directory (relative to the current directory) that raw benchmark + // results are stored in. If not provided, defaults to "results". + "results_dir": "/home/ec2-user/results/distributed", + + // The directory (relative to the current directory) that the html tree + // should be written to. If not provided, defaults to "html". + "html_dir": "/home/ec2-user/html/distributed", + + // The number of characters to retain in the commit hashes. + // "hash_length": 8, + + // `asv` will cache wheels of the recent builds in each + // environment, making them faster to install next time. This is + // number of builds to keep, per environment. + "wheel_cache_size": 1 + + // The commits after which the regression search in `asv publish` + // should start looking for regressions. Dictionary whose keys are + // regexps matching to benchmark names, and values corresponding to + // the commit (exclusive) after which to start looking for + // regressions. The default is to start from the first commit + // with results. If the commit is `null`, regression detection is + // skipped for the matching benchmark. + // + // "regressions_first_commits": { + // "some_benchmark": "352cdf", // Consider regressions only after this commit + // "another_benchmark": null, // Skip regression detection altogether + // } + + // The thresholds for relative change in results, after which `asv + // publish` starts reporting regressions. Dictionary of the same + // form as in ``regressions_first_commits``, with values + // indicating the thresholds. If multiple entries match, the + // maximum is taken. If no entry matches, the default is 5%. + // + // "regressions_thresholds": { + // "some_benchmark": 0.01, // Threshold of 1% + // "another_benchmark": 0.5, // Threshold of 50% + // } +} diff --git a/cron/run_benchmarks.sh b/cron/run_benchmarks.sh new file mode 100755 index 0000000..0f849f8 --- /dev/null +++ b/cron/run_benchmarks.sh @@ -0,0 +1,67 @@ +#!/usr/bin/bash +echo "Running benchmark update `date`" +BENCHMARK_REPO=${BENCHMARK_REPOSITORY:-$HOME/dask-benchmarks} +DASK_DIR=$BENCHMARK_REPO/dask +DISTRIBUTED_DIR=$BENCHMARK_REPO/distributed +DASK_CONFIG=${DASK_ASV_CONFIG:-$HOME/asv.dask.conf.json} +DISTRIBUTED_CONFIG=${DISTRIBUTED_ASV_CONFIG:-$HOME/asv.distributed.conf.json} + +echo "Creating conda environment..." +conda create -n dask-asv python=3.5 +pip install asv +source activate dask-asv + +echo "Updating benchmark repo..." +cd $BENCHMARK_REPO +git checkout master +git pull + +echo "Running dask benchmarks..." +cd $DASK_DIR +echo " Benchmarking new commits..." +asv --config $DASK_CONFIG run NEW +DASK_STATUS=$? +echo " Running new benchmarks on existing commits..." +asv --config $DASK_CONFIG run EXISTING --skip-existing-successful +DASK_STATUS=$(($DASK_STATUS + $?)) +if [ "$DASK_STATUS" -lt "2" ]; then + echo "Generating dask html files..." + asv --config $DASK_CONFIG publish +fi + +echo "Running distributed benchmarks..." +cd $DISTRIBUTED_DIR +echo " Benchmarking new commits..." +asv --config $DISTRIBUTED_CONFIG run NEW +DISTRIBUTED_STATUS=$? +echo " Running new benchmarks on existing commits..." +asv --config $DISTRIBUTED_CONFIG run EXISTING --skip-existing-successful +STATUS=$(($DISTRIBUTED_STATUS + $?)) +if [ "$DISTRIBUTED_STATUS" -lt "2" ]; then + echo "Generating distributed html files..." + # Currently install dask dependency for distributed via pip install git+http to + # get current dask master. asv does not directly support this even though you + # can get it to work. However directory structure gets messed up and machine.json + # is not in the correct location to generate the graphs. Thus this hack to copy it + # to the right locations before running publish. + find /home/ec2-user/results/distributed/aws-ec2-c4.xlarge -type d -exec cp /home/ec2-user/results/distributed/aws-ec2-c4.xlarge/machine.json {} \; + asv --config $DISTRIBUTED_CONFIG publish +fi + +# exit on error otherwise it might still commit +set -e + +STATUS=$(($DASK_STATUS + $DISTRIBUTED_STATUS)) +if [ "$STATUS" -lt "4" ]; then + echo "Publishing results to github..." + cd $BENCHMARK_REPO + git checkout gh-pages + cp -r $HOME/html . + rm -rf results + mv html results + git add results + git commit -am "Auto-committed by benchmark script" + git push +else + echo "No updates to publish..." +fi