9/1/20 9/2/20
This README is for an example project that compares Ruby environments on Vercel and Render.com. The goal is to compare the capability to deploy Ruby functions as API endpoints on the Vercel and Render.com platforms.
This project is a work in progress and currently details some unresolved issues with Vercel. The code runs successfully on Render.com so I’m hoping to resolve the issues on Vercel.
Should the Gemfile be placed in the project root directory or in the api
directory?
What is the correct “Build Command” and “Output Directory” settings to run bundle update
(in order to install a gem from GitHub) and serve Ruby functions from an api
directory?
When “Build Command” and “Output Directory” are not set, how are Ruby gems installed? Does bundle install
run by default for a Ruby function? Why is there no logging of bundler output in the build logs during a default deploy?
Code for the example is on GitHub at https://github.com/DanielKehoe/ruby-trial. You can fork the project and deploy on either Vercel or Render.com.
Here’s the project structure:
$ tree
.
├── Gemfile
├── Gemfile.lock
├── api
│ ├── Gemfile
│ ├── Gemfile.lock
│ └── try-vercel.rb
├── render.yaml
└── try-render.rb
The top level folder contains code for Render.com.
render.yaml
provides settings for Render.com deploymenttry-render.rb
is a trial Ruby function implemented for Render.comGemfile
andGemfile.lock
specify required gems
The api
folder contains code for Vercel.
- Vercel doesn’t require a file for build settings
try-vercel.rb
is a trial Ruby function implemented for VercelGemfile
andGemfile.lock
specify required gems
The Gemfile is designed to test three aspects of Ruby support that can be problematic:
- installing an ordinary gem from rubygems.org
- installing a gem from a GitHub repository and branch
- installing a gem that has to be built with native extensions
Gemfile
source "https://rubygems.org"
gem 'cowsay'
gem 'fauna', github: 'igas/faunadb-ruby', branch: 'patch-1'
gem 'nokogiri'
The cowsay gem is an ordinary gem and installs successfully on either Vercel or Render.com.
The fauna gem is a fork of a gem that is not currently maintained. It is an example of a common situation in the Ruby world. In this case, a developer has forked a gem and patched it with an update. The update has not been published on the rubygems server so it must be installed from GitHub. Bundler documentation shows several formats for specifying a gem to be installed from GitHub. The gem installs successfully on Render.com but fails to install on Vercel.
The nokogiri gem requires native extensions. Nokogiri is a dependency for many other Ruby gems so it is important that it can be installed successfully. It installs successfully on either Vercel or Render.com.
This is a simple Ruby function that requires the gems described above. It doesn’t run successfully on Vercel.
try-vercelrb
require 'rubygems'
require 'cowsay'
require 'fauna'
require 'nokogiri'
puts("Cowsay gem version: " + Cowsay::VERSION)
puts("Fauna gem version: " + Fauna::VERSION)
puts("Nokogiri gem version: " + Nokogiri::VERSION)
puts("$LOAD_PATH: " + $LOAD_PATH.to_s)
Handler = Proc.new do |req, res|
name = req.query['name'] || 'World'
res.status = 200
res['Content-Type'] = 'text/text; charset=utf-8'
res.body = Cowsay.say("Hello #{name}", 'cow')
end
In the Ruby environment, Vercel provides a default WEBrick server with routing to handle any request to the api
directory.
The program loads the required gems and writes the gem version information to the standard output (it will appear in the function log files). Then it writes the $LOAD_PATH
which shows where the gems are installed.
In response to an HTTP request, the function responds with a “Hello World” message.
I’ve tried a similar function on Render.com for comparison to Vercel. It can be deployed and run successfully on Render.com.
try-render.rb
require 'rubygems'
require 'cowsay'
require 'fauna'
require 'nokogiri'
require 'webrick'
puts("Cowsay gem version: " + Cowsay::VERSION)
puts("WEBrick gem version: " + WEBrick::VERSION)
puts("Fauna gem version: " + Fauna::VERSION)
puts("Nokogiri gem version: " + Nokogiri::VERSION)
puts("$LOAD_PATH: " + $LOAD_PATH.to_s)
class Handler < WEBrick::HTTPServlet::AbstractServlet
def do_GET(req, res)
name = req.query['name'] || 'World'
res.status = 200
res['Content-Type'] = 'text/text; charset=utf-8'
res.body = Cowsay.say("Hello #{name}", 'cow')
end
end
server = WEBrick::HTTPServer.new(:Port => 8000)
server.mount("/", Handler)
['INT', 'TERM'].each {|signal|
trap(signal) {server.shutdown}
}
server.start()
Unlike Vercel, Render.com does not provide a default web server so the Ruby function must require the WEBrick gem and implement code to initialize a server and respond to a GET request..
The program loads the required gems and writes the gem version information to the standard output (it will appear in the function log files). Then it writes the $LOAD_PATH
which shows where the gems are installed.
In response to an HTTP request, the function responds with a “Hello World” message.
Before testing locally or deploying on Vercel or Render.com, you must install the needed gems. Remove Gemfile.lock
to create a pristine environment. Then run bundle install
in both the top level (Render.com) and api
(Vercel) folders to create a Gemfile.lock
for both hosting environments.
$ rm Gemfile.lock
$ rm ./api/Gemfile.lock
$ bundle install
$ cd api
$ bundle install
$ cd ../
Here is the local Gemfile.lock
:
GIT
remote: https://github.com/igas/faunadb-ruby.git
revision: f09176e27edb4efbb2996a2c12197fb06b17d3ac
branch: patch-1
specs:
fauna (3.0.1.pre.1)
faraday (~> 1.0)
json (~> 2.3)
net-http-persistent (~> 2.9)
GEM
remote: https://rubygems.org/
specs:
cowsay (0.3.0)
faraday (1.0.1)
multipart-post (>= 1.2, < 3)
json (2.3.1)
mini_portile2 (2.4.0)
multipart-post (2.1.1)
net-http-persistent (2.9.4)
nokogiri (1.10.10)
mini_portile2 (~> 2.4.0)
PLATFORMS
ruby
DEPENDENCIES
cowsay
fauna!
nokogiri
BUNDLED WITH
2.1.2
The try-render.rb
program can be run locally as it provides a complete web server implementation.
$ ruby ./try-render.rb
Open a browser and visit http://localhost:8000/. You’ll see:
_____________
| Hello World |
-------------
\ ^__^
\ (oo)\_______
(__)\ )\/\
||----w |
|| ||
The console shows the gem versions and load path:
Cowsay gem version: 0.3.0
WEBrick gem version: 1.6.0
Fauna gem version: 3.0.0
Nokogiri gem version: 1.10.10
$LOAD_PATH: ["/Users/danielkehoe/.gem/ruby/2.7.0/gems/cowsay-0.3.0/lib”,]
…
After verifying the program runs successfully locally, deploy and test on Vercel and Render.com.
I’m able to deploy and run the function with Vercel default build settings without the fauna gem. When I add the fauna gem, the function fails with an error
"cannot load such file -- fauna"
I’ve tried several scenarios to deploy on Vercel to try to determine what settings are needed to successfully deploy this project.
For the first trial, I removed the Fauna gem from the try-vercel.rb
file.
try-vercel.rb
require 'rubygems'
require 'cowsay'
# require 'fauna'
require 'nokogiri'
puts("Cowsay gem version: " + Cowsay::VERSION)
# puts("Fauna gem version: " + Fauna::VERSION)
puts("Nokogiri gem version: " + Nokogiri::VERSION)
puts("$LOAD_PATH: " + $LOAD_PATH.to_s)
Handler = Proc.new do |req, res|
name = req.query['name'] || 'World'
res.status = 200
res['Content-Type'] = 'text/text; charset=utf-8'
res.body = Cowsay.say("Hello #{name}", 'cow')
end
From the Vercel dashboard, import a project. Choose “Import Git Repository” and continue. Enter the URL of the Git repository https://github.com/DanielKehoe/ruby-trial and click “Deploy.” Select the root folder as the source for the project’s code. Use “ruby-trial” as the project name. Use “Other” as the “Framework Preset.” I’m not setting “Build and Output Settings.” The Build Command is set to “Override” with no settings (blank) and the Output Directory is set to “Override” with no settings (blank). Choose “Deploy”.
The log files show the build progress.
11:54:17.181 Cloning github.com/DanielKehoe/ruby-trial (Branch: master, Commit: 2dccea2)
11:54:17.673 Cloning completed in 492ms
11:54:17.674 Analyzing source code...
11:54:18.341 Uploading build outputs...
11:54:19.400 Installing build runtime...
11:54:20.781 Build runtime installed: 1380.154ms
11:54:21.189 Looking up build cache...
11:54:21.229 Build cache not found
11:55:00.145 Uploading build outputs...
11:55:03.768 Done with "api/try-vercel.rb"
After “Congratulations! Your project has been successfully deployed” open the dashboard. Open the function log page before trying the website. Then go to https://ruby-trial.vercel.app/api/try-vercel
The function runs successfully and the page displays:
_____________
| Hello World |
-------------
\ ^__^
\ (oo)\_______
(__)\ )\/\
||----w |
|| ||
My questions:
- The build log does not show any gems being installed. How are the gems installed? Does
bundle install
run by default for a Ruby function? Why is there no logging ofbundle install
output?
Next I will try restoring the Fauna gem to the try-vercel.rb
file.
try-vercel.rb
require 'rubygems'
require 'cowsay'
require 'fauna'
require 'nokogiri'
puts("Cowsay gem version: " + Cowsay::VERSION)
puts("Fauna gem version: " + Fauna::VERSION)
puts("Nokogiri gem version: " + Nokogiri::VERSION)
puts("$LOAD_PATH: " + $LOAD_PATH.to_s)
Handler = Proc.new do |req, res|
name = req.query['name'] || 'World'
res.status = 200
res['Content-Type'] = 'text/text; charset=utf-8'
res.body = Cowsay.say("Hello #{name}", 'cow')
end
Push to GitHub for an automatic deployment.
The log files will show the build progress.
14:44:10.703 Cloning github.com/DanielKehoe/ruby-trial (Branch: master, Commit: a5696a2)
14:44:11.378 Cloning completed in 675ms
14:44:11.379 Analyzing source code...
14:44:12.073 Uploading build outputs...
14:44:13.124 Installing build runtime...
14:44:14.719 Build runtime installed: 1594.935ms
14:44:15.143 Looking up build cache...
14:44:15.181 Build cache not found
14:44:55.844 Uploading build outputs...
14:44:59.209 Done with "api/try-vercel.rb"
Again the log files do not show any gems getting installed.
Open the function log page before trying the website. Then go to https://ruby-trial.vercel.app/api/try-vercel
The browser will show an application error:
502: BAD_GATEWAY
Code: NO_RESPONSE_FROM_FUNCTION
ID: sin1::n7mqj-1599029156692-f8d6cfe17096
The function log will show an error:
[GET] /api/try-vercel
14:46:50:76
Critical exception from handler{
"errorMessage": "cannot load such file -- fauna",
"errorType": "Function<LoadError>",
"stackTrace": [
"/var/lang/lib/ruby/site_ruby/2.7.0/rubygems/core_ext/kernel_require.rb:92:in `require'",
"/var/lang/lib/ruby/site_ruby/2.7.0/rubygems/core_ext/kernel_require.rb:92:in `require'",
"/var/task/api/try-vercel.rb:3:in `<top (required)>'",
"/var/task/now__handler__ruby.rb:29:in `require_relative'",
"/var/task/now__handler__ruby.rb:29:in `webrick_handler'",
"/var/task/now__handler__ruby.rb:96:in `now__handler'"
]
}
Unknown application error occurred
Function<LoadError>
This error originates from the code in try-vercelrb
at line 3:
require 'fauna'
Presumably the fauna gem was not loaded from GitHub during the default build process.
The Vercel Ruby documentation is limited. All it says about gems is:
This Runtime supports installing dependencies defined in the Gemfile. Alternatively, dependencies can be vendored with the bundler install --deployment command (useful for gems that require native extensions). In this case, dependencies are not built on deployment.
I will try with the “Build Command” set to bundle update
. Toggle “Override” and then enter
bundle update
The example shows bundle update
with backticks. I’m not sure backticks are required, so I will not enter backticks.
The documentation says, for “Output Directory”: “If you turn on the Override toggle and leave the field empty, the build step will be skipped.” That means bundle update
won’t run. So I won’t override the “Output Directory” setting.
“Contents in this output directory will be the only things that will be statically served by Vercel. By default, the output directory will be set as public if it exists, or . (current directory) otherwise. Therefore, as long as your project doesn’t have the public directory, it will serve the files in the root directory.”
I want files in the root directory to be served as static pages. And functions in the api
folder to be run as Ruby functions. I think I have the correct settings, if I understand the documentation correctly.
Redeploy. The build process shows bundle update
running. The build logs show output from bundler.
15:07:57.784 Cloning github.com/DanielKehoe/ruby-trial (Branch: master, Commit: 1d7ed70)
15:07:58.449 Cloning completed in 665ms
15:07:58.450 Analyzing source code...
15:07:58.987 Installing build runtime...
15:07:59.380 Build runtime installed: 392.074ms
15:07:59.786 Looking up build cache...
15:07:59.825 Build cache not found
15:08:00.384 [DEPRECATED] The `--no-prune` flag is deprecated because it relies on being remembered across bundler invocations, which bundler will no longer do in future versions. Instead please use `bundle config set no-prune 'true'`, and stop using this flag
15:08:01.629 Fetching gem metadata from https://rubygems.org/......
15:08:01.632 Fetching https://github.com/igas/faunadb-ruby.git
15:08:02.669 Using bundler 2.1.2
15:08:02.671 Fetching multipart-post 2.1.1
15:08:02.671 Fetching json 2.3.1
15:08:02.671 Fetching cowsay 0.3.0
15:08:02.730 Installing multipart-post 2.1.1
15:08:02.731 Installing cowsay 0.3.0
15:08:02.757 Installing json 2.3.1 with native extensions
15:08:02.771 Fetching net-http-persistent 2.9.4
15:08:02.792 Fetching mini_portile2 2.4.0
15:08:02.833 Installing net-http-persistent 2.9.4
15:08:02.842 Installing mini_portile2 2.4.0
15:08:03.886 Fetching faraday 1.0.1
15:08:03.886 Fetching nokogiri 1.10.10
15:08:03.964 Installing faraday 1.0.1
15:08:04.330 Installing nokogiri 1.10.10 with native extensions
15:08:37.280 Using fauna 3.0.1.pre.1 from https://github.com/igas/faunadb-ruby.git (at patch-1@f09176e)
15:08:37.285 Bundle complete! 3 Gemfile dependencies, 9 gems now installed.
15:08:37.285 Use `bundle info [gemname]` to see where a bundled gem is installed.
15:08:37.364 Installing dependencies...
15:08:37.596 yarn install v1.22.4
15:08:37.608 info No lockfile found.
15:08:37.611 [1/4] Resolving packages...
15:08:37.612 [2/4] Fetching packages...
15:08:37.613 [3/4] Linking dependencies...
15:08:37.617 [4/4] Building fresh packages...
15:08:37.620 success Saved lockfile.
15:08:37.621 Done in 0.03s.
15:08:37.907 Fetching https://github.com/igas/faunadb-ruby.git
15:08:38.929 Fetching gem metadata from https://rubygems.org/.......
15:08:38.961 Resolving dependencies...
15:08:38.972 Using bundler 2.1.2
15:08:38.973 Using cowsay 0.3.0
15:08:38.973 Using multipart-post 2.1.1
15:08:38.973 Using json 2.3.1
15:08:38.973 Using net-http-persistent 2.9.4
15:08:38.973 Using mini_portile2 2.4.0
15:08:38.973 Using faraday 1.0.1
15:08:38.973 Using nokogiri 1.10.10
15:08:38.974 Using fauna 3.0.1.pre.1 from https://github.com/igas/faunadb-ruby.git (at patch-1@f09176e)
15:08:38.976 Bundle updated!
15:08:38.991 Error: No Output Directory named "public" found after the Build completed. You can configure the Output Directory in your Project Settings. Learn More: https://vercel.link/missing-public-directory
15:08:40.015 Installing build runtime...
15:08:41.315 Build runtime installed: 1299.950ms
15:08:41.725 Looking up build cache...
15:08:41.798 Build cache not found
The build log shows that the Fauna gem is loaded from GitHub:
Using fauna 3.0.1.pre.1 from https://github.com/igas/faunadb-ruby.git (at patch-1@f09176e)
The build fails with this error:
15:08:38.991 Error: No Output Directory named "public" found after the Build completed. You can configure the Output Directory in your Project Settings. Learn More: https://vercel.link/missing-public-directory
This is confusing because the docs say, “By default, the output directory will be set as public if it exists, or . (current directory) otherwise. Therefore, as long as your project doesn’t have the public directory, it will serve the files in the root directory.”
If I understand the documentation correctly, there should be no need for a public
directory. In this case, the root directory should serve as the output directory, providing static files.
Checked the link https://vercel.link/missing-public-directory: “The build step will result in an error if the output directory is missing, empty, or invalid (if it’s actually not a directory).“ This statement seems to contradict the other statement , “By default, the output directory will be set as public if it exists, or . (current directory) otherwise.”
What happens if I explicitly set the output directory to .
?
I toggled “Override” and set the “Output Directory” to .
(dot).
I didn’t use backticks though the example surrounds .
in backticks.
Clicked “Save.” Then redeployed.
The build process showed output from bundler and completed successfully.
15:22:57.584 Cloning github.com/DanielKehoe/ruby-trial (Branch: master, Commit: 1d7ed70)
15:22:58.268 Cloning completed in 684ms
15:22:58.269 Analyzing source code...
15:22:58.882 Installing build runtime...
15:22:59.278 Build runtime installed: 395.054ms
15:23:00.227 [DEPRECATED] The `--no-prune` flag is deprecated because it relies on being remembered across bundler invocations, which bundler will no longer do in future versions. Instead please use `bundle config set no-prune 'true'`, and stop using this flag
15:23:01.413 Fetching gem metadata from https://rubygems.org/......
15:23:01.416 Fetching https://github.com/igas/faunadb-ruby.git
15:23:02.462 Using bundler 2.1.2
15:23:02.464 Fetching cowsay 0.3.0
15:23:02.465 Fetching multipart-post 2.1.1
15:23:02.465 Fetching json 2.3.1
15:23:02.521 Installing multipart-post 2.1.1
15:23:02.526 Installing cowsay 0.3.0
15:23:02.556 Installing json 2.3.1 with native extensions
15:23:02.560 Fetching net-http-persistent 2.9.4
15:23:02.584 Fetching mini_portile2 2.4.0
15:23:02.627 Installing net-http-persistent 2.9.4
15:23:02.650 Fetching faraday 1.0.1
15:23:02.650 Installing mini_portile2 2.4.0
15:23:02.694 Fetching nokogiri 1.10.10
15:23:02.726 Installing faraday 1.0.1
15:23:03.027 Installing nokogiri 1.10.10 with native extensions
15:23:36.637 Using fauna 3.0.1.pre.1 from https://github.com/igas/faunadb-ruby.git (at patch-1@f09176e)
15:23:36.642 Bundle complete! 3 Gemfile dependencies, 9 gems now installed.
15:23:36.642 Use `bundle info [gemname]` to see where a bundled gem is installed.
15:23:36.724 Installing dependencies...
15:23:36.953 yarn install v1.22.4
15:23:36.965 info No lockfile found.
15:23:36.968 [1/4] Resolving packages...
15:23:36.969 [2/4] Fetching packages...
15:23:36.970 [3/4] Linking dependencies...
15:23:36.974 [4/4] Building fresh packages...
15:23:36.976 success Saved lockfile.
15:23:36.977 Done in 0.03s.
15:23:37.256 Fetching https://github.com/igas/faunadb-ruby.git
15:23:38.288 Fetching gem metadata from https://rubygems.org/.......
15:23:38.324 Resolving dependencies...
15:23:38.336 Using bundler 2.1.2
15:23:38.337 Using cowsay 0.3.0
15:23:38.337 Using multipart-post 2.1.1
15:23:38.337 Using json 2.3.1
15:23:38.337 Using net-http-persistent 2.9.4
15:23:38.337 Using mini_portile2 2.4.0
15:23:38.337 Using faraday 1.0.1
15:23:38.338 Using nokogiri 1.10.10
15:23:38.338 Using fauna 3.0.1.pre.1 from https://github.com/igas/faunadb-ruby.git (at patch-1@f09176e)
15:23:38.340 Bundle updated!
15:23:38.521 Uploading build outputs...
15:23:40.055 Installing build runtime...
15:23:41.328 Build runtime installed: 1273.711ms
15:24:20.542 Uploading build outputs...
15:24:24.007 Build completed. Populating build cache...
15:24:24.018 Uploading build cache [207.00 B]...
15:24:24.086 Build cache uploaded: 68.546ms
15:24:24.092 Done with ".editorconfig"
15:24:24.501 Done with "api/try-vercel.rb"
Open the function log page before trying the website. Then go to https://ruby-trial.vercel.app/api/try-vercel
The request succeeds but returns a text file of the code itself (opening a window to download the file).
Tried https://ruby-trial.vercel.app/api/try-vercel.rb Same thing: a text file of the code itself.
I’ve set the “Build Command” to bundle update
. I want the project root to serve static files (no build needed) and the api
directory to serve Ruby functions. If I don’t override the “Output Directory” setting I get the error “No Output Directory named ‘public’ found after the Build completed” and the build fails. When I set the “Output Directory” setting to .
the scripts in the api
folder get served as static files not functions.
My question:
- What is the correct “Build Command” and “Output Directory” to run
bundle update
(in order to install a gem from GitHub) and serve Ruby functions from anapi
directory?
To be certain a gem can be loaded from GitHub and the function runs successfully, I deployed the project to Render.com (a service similar to Vercel).
From the Render.com dashboard, create a new web service. Choose the GitHub repository https://github.com/DanielKehoe/ruby-trial. Specify a project name (“ruby-trial) and a start command bundle exec ruby try-render.rb
.
Deployment will produce log output:
Sep 1 01:30:07 PM ==> Cloning from https://github.com/DanielKehoe/ruby-trial...
Sep 1 01:30:07 PM ==> Checking out commit 96f8c8dbd39cb192f59f7552de148212645580b4 in branch master
Sep 1 01:30:08 PM ==> Downloading cache...
Sep 1 01:30:11 PM ==> Running build command 'bundle install'...
Sep 1 01:30:13 PM Fetching gem metadata from https://rubygems.org/......
Sep 1 01:30:13 PM Fetching https://github.com/igas/faunadb-ruby.git
Sep 1 01:30:14 PM Using bundler 2.1.2
Sep 1 01:30:14 PM Fetching cowsay 0.3.0
Sep 1 01:30:14 PM Installing cowsay 0.3.0
Sep 1 01:30:14 PM Fetching multipart-post 2.1.1
Sep 1 01:30:14 PM Installing multipart-post 2.1.1
Sep 1 01:30:14 PM Fetching faraday 1.0.1
Sep 1 01:30:14 PM Installing faraday 1.0.1
Sep 1 01:30:14 PM Fetching json 2.3.1
Sep 1 01:30:14 PM Installing json 2.3.1 with native extensions
Sep 1 01:30:16 PM Fetching net-http-persistent 2.9.4
Sep 1 01:30:16 PM Installing net-http-persistent 2.9.4
Sep 1 01:30:16 PM Using fauna 3.0.1.pre.1 from https://github.com/igas/faunadb-ruby.git (at patch-1@f09176e)
Sep 1 01:30:16 PM Fetching mini_portile2 2.4.0
Sep 1 01:30:16 PM Installing mini_portile2 2.4.0
Sep 1 01:30:16 PM Fetching nokogiri 1.10.10
Sep 1 01:30:17 PM Installing nokogiri 1.10.10 with native extensions
Sep 1 01:30:54 PM Bundle complete! 3 Gemfile dependencies, 9 gems now installed.
Sep 1 01:30:54 PM Gems in the groups development and test were not installed.
Sep 1 01:30:54 PM Bundled gems are installed into `/opt/render/project/.gems`
Sep 1 01:30:56 PM ==> Uploading build...
Sep 1 01:30:59 PM ==> Build successful 🎉
Sep 1 01:30:59 PM ==> Deploying...
Sep 1 01:31:15 PM ==> Starting service with 'bundle exec ruby try-render.rb'
Sep 1 01:31:16 PM [2020-09-01 05:31:16] INFO WEBrick 1.4.2
Sep 1 01:31:16 PM [2020-09-01 05:31:16] INFO ruby 2.6.5 (2019-10-01) [x86_64-linux]
Sep 1 01:31:16 PM [2020-09-01 05:31:16] INFO WEBrick::HTTPServer#start: pid=54 port=8000
Sep 1 01:31:39 PM ==> Starting service with 'bundle exec ruby try-render.rb'
Sep 1 01:31:40 PM [2020-09-01 05:31:40] INFO WEBrick 1.4.2
Sep 1 01:31:40 PM [2020-09-01 05:31:40] INFO ruby 2.6.5 (2019-10-01) [x86_64-linux]
Sep 1 01:31:40 PM [2020-09-01 05:31:40] INFO WEBrick::HTTPServer#start: pid=54 port=8000
Sep 1 01:32:01 PM [2020-09-01 05:32:01] INFO going to shutdown ...
Sep 1 01:32:01 PM [2020-09-01 05:32:01] INFO WEBrick::HTTPServer#start done.
Sep 1 01:32:01 PM Cowsay gem version: 0.3.0
Sep 1 01:32:01 PM WEBrick gem version: 1.4.2
Sep 1 01:32:01 PM Fauna gem version: 3.0.1.pre.1
Sep 1 01:32:01 PM Nokogiri gem version: 1.10.10
Sep 1 01:32:01 PM $LOAD_PATH: ["/opt/render/project/.gems/gems/bundler-2.1.2/lib", "/opt/render/project/.gems/ruby/2.6.0/gems/nokogiri-1.10.10/lib", "/opt/render/project/.gems/ruby/2.6.0/extensions/x86_64-linux/2.6.0/nokogiri-1.10.10", "/opt/render/project/.gems/ruby/2.6.0/gems/mini_portile2-2.4.0/lib", "/opt/render/project/.gems/ruby/2.6.0/bundler/gems/faunadb-ruby-f09176e27edb/lib", "/opt/render/project/.gems/ruby/2.6.0/gems/net-http-persistent-2.9.4/lib", "/opt/render/project/.gems/ruby/2.6.0/gems/json-2.3.1/lib", "/opt/render/project/.gems/ruby/2.6.0/extensions/x86_64-linux/2.6.0/json-2.3.1", "/opt/render/project/.gems/ruby/2.6.0/gems/faraday-1.0.1/lib", "/opt/render/project/.gems/ruby/2.6.0/gems/faraday-1.0.1/spec/external_adapters", "/opt/render/project/.gems/ruby/2.6.0/gems/multipart-post-2.1.1/lib", "/opt/render/project/.gems/ruby/2.6.0/gems/cowsay-0.3.0/lib", "/opt/render/project/.gems/gems/bundler-2.1.2/lib/gems/bundler-2.1.2/lib", "/usr/local/lib/ruby/site_ruby/2.6.0", "/usr/local/lib/ruby/site_ruby/2.6.0/x86_64-linux", "/usr/local/lib/ruby/site_ruby", "/usr/local/lib/ruby/vendor_ruby/2.6.0", "/usr/local/lib/ruby/vendor_ruby/2.6.0/x86_64-linux", "/usr/local/lib/ruby/vendor_ruby", "/usr/local/lib/ruby/2.6.0", "/usr/local/lib/ruby/2.6.0/x86_64-linux"]
The Render.com build process shows the output from the bundle install
command. It’s easy to confirm that the FaunaDB gem was installed from GitHub:
Using fauna 3.0.1.pre.1 from https://github.com/igas/faunadb-ruby.git (at patch-1@f09176e)
Open a browser and visit https://ruby-trial.onrender.com You’ll see:
_____________
| Hello World |
-------------
\ ^__^
\ (oo)\_______
(__)\ )\/\
||----w |
|| ||
The log file shows the result of the request:
Sep 1 01:35:34 PM 10.104.78.7 - - [01/Sep/2020:05:35:34 UTC] "GET / HTTP/1.1" 200 161
Sep 1 01:35:34 PM - -> /
There is no problem on Render.com.