# frozen_string_literal: true # Copyright 2021 Google LLC # # Licensed 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. require "shellwords" require "google/serverless/exec/gcloud" require "google/serverless/exec" module Google module Serverless class Exec ## # # Rake Task for serverless exec # # This module implements the `serverless:exec` rake task. To make it # available, add the line # # require "google/serverless/exec/tasks" # # to your Rakefile. If your app uses Ruby on Rails, then the gem provides # a railtie that adds this task automatically, so you don't have to do # anything beyond adding the gem to your Gemfile. # # ## Usage # # The `serverless:exec` a given command in the context of a serverless # application. See {Google::Serverless::Exec} for more information on # this capability. # # The command to be run may either be provided as a rake argument, or as # command line arguments, delimited by two dashes `--`. (The dashes are # needed to separate your command from rake arguments and flags.) # For example, to run a production database migration, you can run either # of the following equivalent commands: # # bundle exec rake "serverless:exec[bundle exec bin/rails db:migrate]" # bundle exec rake serverless:exec -- bundle exec bin/rails db:migrate # # To display usage instructions, provide two dashes but no command: # # bundle exec rake serverless:exec -- # # ## Parameters # # You may customize the behavior of the serverless execution through a few # enviroment variable parameters. These are set via the normal mechanism # at the end of a rake command line. For example, to set EXEC_APP_CONFIG: # # bundle exec rake serverless:exec EXEC_APP_CONFIG=myservice.yaml -- bundle exec bin/rails db:migrate # # Be sure to set these parameters before the double dash. Any arguments # following the double dash are interpreted as part of the command itself. # # The following environment variable parameters are supported: # # ### EXEC_PROJECT # # The ID of your Google Cloud project. If not specified, uses the current # project from gcloud. # # ### EXEC_PRODUCT # # The product used to deploy the app. Valid values are `app_engine` or # `cloud_run`. If omitted, the task will try to guess based on whether # an App Engine config file (e.g. `app.yaml`) is present. # # ### EXEC_REGION # # The Google Cloud region to which the app is deployed. Required for # Cloud Run apps. # # ### EXEC_STRATEGY # # The execution strategy to use. Valid values are `deployment` (which is # the default for App Engine Standard apps) and `cloud_build` (which is # the default for App Engine Flexible and Cloud Run apps). # # Normally you should leave the strategy set to the default. The main # reason to change it is if your app runs on the Flexible Environment and # talks to a database over a VPC (using a private IP address). The # `cloud_build` strategy used by default for Flexible apps cannot connect # to a VPC, so you should use `deployment` in this case. (But note that, # otherwise, the `deployment` strategy is significantly slower for apps # on the App Engine Flexible Environment.) # # ### EXEC_TIMEOUT # # Amount of time to wait before serverless:exec terminates the command. # Expressed as a string formatted like: `2h15m10s`. Default is `10m`. # # ### EXEC_SERVICE_NAME # # Name of the service to be used. If your app uses Cloud Run, this is # required. If your app uses App Engine, normally the service name is # obtained from the config file, but you can override it using this # parameter. # # ### EXEC_APP_CONFIG # # Path to the App Engine config file, used when your app has multiple # services, or the config file is otherwise not called `./app.yaml`. The # config file is used to determine the name of the App Engine service. # Unused for Cloud Run apps. # # ### EXEC_APP_VERSION # # The version of the App Engine service, used to identify which # application image to use to run your command. If not specified, uses # the most recently created version, regardless of whether that version # is actually serving traffic. Applies only to the `cloud_build` strategy # for App Engine apps; ignored for the `deployment` strategy or if your # app uses Cloud Run. # # ### EXEC_BUILD_LOGS_DIR # # GCS bucket name of the cloud build log when GAE_STRATEGY is # `cloud_build`. (ex. `gs://BUCKET-NAME/FOLDER-NAME`) # # ### EXEC_WRAPPER_IMAGE # # The fully-qualified name of the wrapper image to use. (This is a Docker # image that emulates the App Engine environment in Google Cloud Build # for the `cloud_build` strategy, and applies only to that strategy.) # Normally, you should not override this unless you are testing a new # wrapper. # module Tasks ## @private PROJECT_ENV = "EXEC_PROJECT" ## @private STRATEGY_ENV = "EXEC_STRATEGY" ## @private CONFIG_ENV = "EXEC_APP_CONFIG" ## @private SERVICE_ENV = "EXEC_SERVICE_NAME" ## @private VERSION_ENV = "EXEC_APP_VERSION" ## @private TIMEOUT_ENV = "EXEC_TIMEOUT" ## @private WRAPPER_IMAGE_ENV = "EXEC_WRAPPER_IMAGE" ## @private LOGS_DIR_ENV = "EXEC_BUILD_LOGS_DIR" ## @private PRODUCT_ENV = "EXEC_PRODUCT" ## @private REGION_ENV = "EXEC_REGION" @defined = false class << self ## # @private # Define rake tasks. # def define if @defined puts "Serverless rake tasks already defined." return end @defined = true setup_exec_task end private def setup_exec_task ::Rake.application.last_description = "Execute the given command in a Google serverless application." ::Rake::Task.define_task "serverless:exec", [:cmd] do |_t, args| command = extract_command args[:cmd], ::ARGV verify_gcloud_and_report_errors selected_product = extract_product ::ENV[PRODUCT_ENV] app_exec = Exec.new command, project: ::ENV[PROJECT_ENV], service: ::ENV[SERVICE_ENV], config_path: ::ENV[CONFIG_ENV], version: ::ENV[VERSION_ENV], timeout: ::ENV[TIMEOUT_ENV], wrapper_image: ::ENV[WRAPPER_IMAGE_ENV], strategy: ::ENV[STRATEGY_ENV], gcs_log_dir: ::ENV[LOGS_DIR_ENV], region: ::ENV[REGION_ENV], product: selected_product start_and_report_errors app_exec exit end end def extract_command cmd, argv if cmd ::Shellwords.split cmd else i = (argv.index { |a| a.to_s == "--" } || -1) + 1 if i.zero? report_error <<~MESSAGE No command provided for serverless:exec. Did you remember to delimit it with two dashes? e.g. bundle exec rake serverless:exec -- bundle exec ruby myscript.rb For detailed usage instructions, provide two dashes but no command: bundle exec rake serverless:exec -- MESSAGE end command = ::ARGV[i..-1] if command.empty? show_usage exit end command end end def extract_product product return unless product product = product.dup product.downcase! case product when "app_engine" APP_ENGINE when "cloud_run" CLOUD_RUN end end def show_usage puts <<~USAGE rake serverless:exec This Rake task executes a given command in the context of a serverless application. For more information, on this capability, see the Google::Serverless::Exec documentation at http://www.rubydoc.info/gems/google-serverless-exec/Google/Serverless/Exec The command to be run may either be provided as a rake argument, or as command line arguments delimited by two dashes "--". (The dashes are needed to separate your command from rake arguments and flags.) For example, to run a production database migration, you can run either of the following equivalent commands: bundle exec rake "serverless:exec[bundle exec bin/rails db:migrate]" bundle exec rake serverless:exec -- bundle exec bin/rails db:migrate To display these usage instructions, provide two dashes but no command: bundle exec rake serverless:exec -- You may customize the behavior of the serverless execution through a few enviroment variable parameters. These are set via the normal mechanism at the end of a rake command line but before the double dash. For example, to set EXEC_APP_CONFIG: bundle exec rake serverless:exec EXEC_APP_CONFIG=myservice.yaml -- \\ bundle exec bin/rails db:migrate Be sure to set these parameters before the double dash. Any arguments following the double dash are interpreted as part of the command itself. The following environment variable parameters are supported: EXEC_PROJECT The ID of your Google Cloud project. If not specified, uses the current project from gcloud. EXEC_PRODUCT The product used to deploy the app. Valid values are "app_engine" or "cloud_run". If omitted, the task will try to guess based on whether an App Engine config file (e.g. app.yaml) is present. EXEC_REGION The Google Cloud region to which the app is deployed. Required for Cloud Run apps. EXEC_STRATEGY The execution strategy to use. Valid values are "deployment" (which is the default for App Engine Standard apps) and "cloud_build" (which is the default for App Engine Flexible and Cloud Run apps). Normally you should leave the strategy set to the default. The main reason to change it is if your app runs on the Flexible Environment and talks to a database over a VPC (using a private IP address). The "cloud_build" strategy used by default for Flexible apps cannot connect to a VPC, so you should use "deployment" in this case. (But note that, otherwise, the "deployment" strategy is significantly slower for apps on the Flexible environment.) EXEC_TIMEOUT Amount of time to wait before serverless:exec terminates the command. Expressed as a string formatted like: "2h15m10s". Default is "10m". EXEC_SERVICE_NAME Name of the service to be used. If your app uses Cloud Run, this is required. If your app uses App Engine, normally the service name is obtained from the config file, but you can override it using this parameter. EXEC_APP_CONFIG Path to the App Engine config file, used when your app has multiple services, or the config file is otherwise not called "./app.yaml". The config file is used to determine the name of the App Engine service. Unused for Cloud Run apps. EXEC_APP_VERSION The version of the App Engine service, used to identify which application image to use to run your command. If not specified, uses the most recently created version, regardless of whether that version is actually serving traffic. Applies only to the "cloud_build" strategy for App Engine apps; ignored for the "deployment" strategy or if your app uses Cloud Run. EXEC_BUILD_LOGS_DIR GCS bucket name of the cloud build log when GAE_STRATEGY is "cloud_build". (ex. "gs://BUCKET-NAME/FOLDER-NAME") EXEC_WRAPPER_IMAGE The fully-qualified name of the wrapper image to use. (This is a Docker image that emulates the App Engine environment in Google Cloud Build for the "cloud_build" strategy, and applies only to that strategy.) Normally, you should not override this unless you are testing a new wrapper. USAGE end def verify_gcloud_and_report_errors Exec::Gcloud.verify! rescue Exec::Gcloud::BinaryNotFound report_error <<~MESSAGE Could not find the `gcloud` binary in your system path. This tool requires the Google Cloud SDK. To download and install it, visit https://cloud.google.com/sdk/downloads MESSAGE rescue Exec::Gcloud::GcloudNotAuthenticated report_error <<~MESSAGE The gcloud authorization has not been completed. If you have not yet initialized the Google Cloud SDK, we recommend running the `gcloud init` command as described at https://cloud.google.com/sdk/docs/initializing Alternately, you may log in directly by running `gcloud auth login`. MESSAGE rescue Exec::Gcloud::ProjectNotSet report_error <<~MESSAGE The gcloud project configuration has not been set. If you have not yet initialized the Google Cloud SDK, we recommend running the `gcloud init` command as described at https://cloud.google.com/sdk/docs/initializing Alternately, you may set the default project configuration directly by running `gcloud config set project `. MESSAGE end def start_and_report_errors app_exec app_exec.start rescue Exec::ConfigFileNotFound => e report_error <<~MESSAGE Could not determine which service should run this command because the App Engine config file "#{e.config_path}" was not found. Specify the config file using the EXEC_APP_CONFIG argument. e.g. bundle exec rake serverless:exec EXEC_APP_CONFIG=myapp.yaml -- myscript.sh Alternately, specify a service name directly with EXEC_SERVICE_NAME. e.g. bundle exec rake serverless:exec EXEC_SERVICE_NAME=myservice -- myscript.sh MESSAGE rescue Exec::BadConfigFileFormat => e report_error <<~MESSAGE Could not determine which service should run this command because the App Engine config file "#{e.config_path}" was malformed. It must be a valid YAML file. Specify the config file using the EXEC_APP_CONFIG argument. e.g. bundle exec rake serverless:exec EXEC_APP_CONFIG=myapp.yaml -- myscript.sh Alternately, specify a service name directly with EXEC_SERVICE_NAME. e.g. bundle exec rake serverless:exec EXEC_SERVICE_NAME=myservice -- myscript.sh MESSAGE rescue Exec::NoSuchVersion => e if e.version report_error <<~MESSAGE Could not find version "#{e.version}" of service "#{e.service}". Please double-check the version exists. To use the most recent version by default, omit the EXEC_APP_VERSION argument. MESSAGE else report_error <<~MESSAGE Could not find any versions of service "#{e.service}". Please double-check that you have deployed this service. If you want to run a command against a different service, you may provide a EXEC_APP_CONFIG argument pointing to your App Engine config file, or a EXEC_SERVICE_NAME argument to specify a service directly. MESSAGE end rescue Exec::NoDefaultProject report_error <<~MESSAGE Could not get the default project from gcloud. Please either set the current project using gcloud config set project my-project-id or specify the project by setting the EXEC_PROJECT argument. e.g. bundle exec rake serverless:exec EXEC_PROJECT=my-project-id -- myscript.sh MESSAGE rescue Exec::UsageError => e report_error e.message end def report_error str warn str exit 1 end end end end end end ::Google::Serverless::Exec::Tasks.define