circleci

问题 Is it possible to cache gems, so that bundle install will not install bunch of gems for every build? This takes 5 minutes on every build, that is too much. I've added this to the circle.yml config: dependencies: cache_directories: - "/home/ubuntu/.rvm/gems/ruby-2.1.2/gems/" This is the common directory which is provided by bundle show gem_name where all system gems are put. After this addition system writes such log: restoring cache v4/company/repo_name/dependency/circle-ci/42

问题 Is it possible to cache gems, so that bundle install will not install bunch of gems for every build? This takes 5 minutes on every build, that is too much. I've added this to the circle.yml config: dependencies: cache_directories: - "/home/ubuntu/.rvm/gems/ruby-2.1.2/gems/" This is the common directory which is provided by bundle show gem_name where all system gems are put. After this addition system writes such log: restoring cache v4/company/repo_name/dependency/circle-ci/42

问题 I'm running the cypress-example-kitchen sink app on CircleCI. This is my yaml config script: version: 2.1 orbs: cypress: cypress-io/cypress@1.0.1 workflows: build: jobs: - cypress/install: build: 'npm run build' - cypress/run: requires: - cypress/install start: 'npm start' This kicks off and passes just fine when I make a commit to my fork of the repo above. However, when I try to execute a CircleCI build programmatically, using curl -X POST https://circleci.com/api/v1.1/project/github

问题 I have test and deploy paths into the workflow and when I sequentially execute them deploy overrides artifact . Is it possible to keep it, so it will be passed from stage to stage sequentially? Store artifact into Slack or any other system is not an option due to limitations in the development environment. 回答1: The Circle CI provides the solution through workspace. To share artifact or any set of files, you need to persist_to_workspace into the test stage and attach_to_workspace on the deploy

问题 I'm trying to use CircleCI on a Go project on GitHub with the default template for Go projects. For reference, here's what the default .circleci/config.yml looks like : version: 2 jobs: build: docker: - image: circleci/golang:1.12.7-buster working_directory: /go/src/github.com/Permaweb/Host steps: - checkout - run: go get -t -u -v ./... - run: go test -v ./... When the job runs, I get an error that's totally unrelated to the code itself. agent key RSA SHA256:L1iUIhjfKejEONe0TFU3TaWkNwE2O0xWF

问题 I am trying to setup CircleCI with Elasticsearch in my Rails application. I think have the image configured, but how do I connect to it in CI? So far I have tried... https://github.com/elastic/elasticsearch/issues/23866 Error Message Elasticsearch::Transport::Transport::Errors::Unauthorized: [401] Circle YAML Config version: 2 jobs: build: working_directory: ~/gathrly-smartforms docker: - image: circleci/ruby:2.4.1-node environment: RAILS_ENV: continous_integration PGHOST: 127.0.0.1 PGUSER:

问题 I've managed to create a CircleCI build that triggers a subsequent build using their API using curl . I've added this to my circle.yml : test: override: - mvn test -s settings.xml - mvn deploy -Prun-its -s settings.xml - curl -v -X POST https://circleci.com/api/v1/project/alexec/docker-maven-plugin/tree/master?circle-token=$CIRCLE_TOKEN How do I trigger only if all of the previous steps are green? 回答1: I think you should do this in the deployment section: Since this is - by definition - only

问题 Since upgrading to Laravel 5.2 from 5.1, when running artisan optimize on CircleCI, running PHP 5.6.14, I am getting [ErrorException] php_strip_whitespace(/var/laravel/project/root): failed to open stream: No child processes Where /var/laravel/project/root is the directory where composer.json and vendor reside. The command runs fine on my dev box running PHP 5.6.11-1ubuntu3.1. I followed the official 5.1 to 5.2 upgrade guide. Exception trace: () at /var/laravel/project/root/vendor

问题 I am trying to build a CircleCI configuration file that only installs a specific set of gems via the environment parameter. In this case let's call that environment continuous_integration and this environment matches the test environment. So far I have tried a number of things and this is my current configuration in this spike. Could anyone point me in the right direction? Is this possible? machine: timezone: America/Los_Angeles ruby: version: 2.4.1 services: - redis environment: RAILS_ENV:

问题 I have a Dockerfile that works fine: FROM node:10 RUN npm set unsafe-perm true RUN npm install -g '@oresoftware/r2g@0.0.132' but the same CircleCI config.yml file that mirrors the above Dockerfile does not work: { "version": 2, "jobs": { "build": { "docker": [ { "image": "circleci/node:10" } ], "steps": [ { "run": "npm set unsafe-perm true" }, { "run": "npm install -g --loglevel=warn '@oresoftware/r2g@0.0.132'" } ] } } } I get the following error on CircleCI using the above config.yml file: #