I was using Postgres universal (postgresql package managed by brew) and I had installed postgresql@14 managed version by brew at the same time. This instruction covers what to do if you use postgresql@14 managed by brew or the universal version.
This instruction is inspired by the issue.
You can change the version to the one that you want to upgrade to. It's recommended to upgrade one version at a time.
| - run: | |
| name: Run tests | |
| command: | | |
| export CIRCLE_TEST_REPORTS=/tmp/test-results | |
| mkdir -p $CIRCLE_TEST_REPORTS | |
| # split slow spec files by test examples | |
| # https://docs.knapsackpro.com/ruby/split-by-test-examples/ | |
| export KNAPSACK_PRO_RSPEC_SPLIT_BY_TEST_EXAMPLES=true |
| before_suite_block = proc { | |
| puts '+'*100 | |
| puts 'RSpec before suite hook called only once' | |
| } | |
| unless KnapsackPro::Config::Env.queue_recording_enabled? | |
| RSpec.configure do |config| | |
| config.before(:suite) do | |
| before_suite_block.call | |
| puts 'Run this only when not using Knapsack Pro Queue Mode' |
| # repalce REDIS_CONNECTION with your Redis instance | |
| # | |
| # run this script in Rails console. | |
| # For example start Heroku Dyno with Rails console to measure the Redis latency from inside of the Heroku Dyno | |
| # | |
| # Do you like it? Check what we do: https://knapsackpro.com | |
| rtts = [] | |
| 100.times do | |
| a = b = 0; a = ::Process.clock_gettime(::Process::CLOCK_MONOTONIC, :microsecond); REDIS_CONNECTION.ping; b = ::Process.clock_gettime(::Process::CLOCK_MONOTONIC, :microsecond); rtt = b - a; |
| KnapsackPro::Hooks::Queue.after_subset_queue do |queue_id, subset_queue_id| | |
| puts '\\'*100 | |
| puts 'Executed tests for a batch of tests fetched from Queue API' | |
| Dir.glob(".knapsack_pro/queue/#{queue_id}/*.json").each do |file| | |
| report = JSON.parse(File.read(file)) | |
| puts report.inspect | |
| end | |
| puts '/'*100 | |
| end |
Retry tests 3 times on CI only (when CI environment variable is present). In development run tests only once for easier development.
How to retry failed tests (flaky tests)? https://knapsackpro.com/faq/question/how-to-retry-failed-tests-flaky-tests
Fix intermittently failing CI builds due to flaky tests in RSpec
OpenSSL::SSL::SSLError SSL_connect returned=1 errno=0 state=error: certificate verify failed
Let's Encrypt certificates may be affected by the recent expiry (2021-09-30) of their old root certificate. https://letsencrypt.org/docs/dst-root-ca-x3-expiration-september-2021/
Requirements:
| # bitbucket-pipelines.yml | |
| # this is example using cypress/base docker image | |
| image: cypress/base:10 | |
| options: | |
| max-time: 30 | |
| # job definition for running E2E tests in parallel with KnapsackPro.com | |
| e2e: &e2e | |
| name: Run E2E tests with @knapsack-pro/cypress | |
| caches: |
A commercially supported version of Knapsack that provides a hosted service for test timing data and additional job distribution modes for Ruby, JavaScript, etc. See the documentation and step-by-step tutorial for Ruby setup instructions and example pipelines. For other programming languages please check integrations.
| # This is example how to rescue from exception ActionDispatch::Http::MimeNegotiation::InvalidType | |
| # and show nice JSON error in your API | |
| module API | |
| class BaseController < ActionController::API | |
| def process_action(*args) | |
| super | |
| rescue ActionDispatch::Http::MimeNegotiation::InvalidType => exception | |
| # set valid Content-Type to be able to call render method below | |
| request.headers['Content-Type'] = 'application/json' | |
| render status: 400, json: { errors: [exception.message] } |