#335 Deploying to a VPS pro

Great screencast, Ryan! I'm comfortable setting servers and, over time, I have created my own system to it but I'm always learning and some of your techniques might be added to my workflow… :)

Question: I always add an nginx conf file for the domain I'm hosting. You simply deleted the default. How does deleting a default automatically make it load the application from the deployer's home directory? Magic sauce there? :)

Thanks Ryan, this is superb.

Reassuring to see that it's pretty similar to what I bodged together after watching your Vagrant/Unicorn screencasts https://gist.github.com/1985879

in Linux,instead of using:
cat ~/.ssh/id_rsa.pub | ssh user@host 'cat >> ~/.ssh/authorized_keys'

i prefer to use:
ssh-copy-id user@host

I wonder if this could apply to services like EC2.

Okay, so I'm developing a web app that's going to have around 600 users - creating about 1500 records a day.

After watching this I'm tempted to check out a VPS, but I've only got experience deploying with Heroku ( <- total n00b ).

For my use case, would you recommend a VPS (such as Linode) or Heroku?

Most rails applications have more dependencies, like a sphinx server for example. How would you go about launching these (e.g. in case of a server restart)?
Can something like Foreman be used for this task?

If Linode is your thing another quick way to setup a VPS is to use a StackScript.

http://www.linode.com/stackscripts/view/?StackScriptID=1291

Includes:
*Rails 3
*Ruby 1.9.2
*Nginx with Passenger
*MySQL
*git
*Updates rubygems
*Install rails 3
*Install mysql gem
*Add deploy user

Nice work again Ryan but would love a full cast on deployment.

I've never commented here before, but felt like this episode was worth the comment.

Fantastic. I wish this had been done weeks ago prior to my deployment headaches.

Lastly, I've been a fan of the railsready script found here: https://github.com/joshfng/railsready

What are some security concerns for VPS (linode) vs managed solution (heroku)?

For some reason I'm getting this while doing a cold deploy, deploy:setup succeeds however, the repo is on the same server. Tried referencing it through the full ssh url or by giving a local absolute url, no avail.

$ cap deploy:cold 
  * executing `deploy:cold'
  * executing `deploy:update'
 ** transaction: start
  * executing `deploy:update_code'
    updating the cached checkout on all servers
    executing locally: "git ls-remote /home/deployer/projects/blog master"
fatal: '/home/deployer/projects/blog' does not appear to be a git repository
fatal: The remote end hung up unexpectedly
*** [deploy:update_code] rolling back
  * executing "rm -rf /home/deployer/apps/blog/releases/20120322194319; true"
    servers: ["192.168.1.49"]
    [192.168.1.49] executing command
    [192.168.1.49] sh -c 'rm -rf /home/deployer/apps/blog/releases/20120322194319; true'
    command finished in 186ms
Command git ls-remote /home/deployer/projects/blog master returned status code pid 9997 exit 128

Im getting this error with deploy:cold..any ideas?

failed: "sh -c 'cd /home/deployer/apps/burden_farming/releases/20120322200251 && bundle install --gemfile /home/deployer/apps/burden_farming/releases/20120322200251/Gemfile --path /home/deployer/apps/burden_farming/shared/bundle --deployment --quiet --without development test'" on 74.207.240.214

Excellent!! as usual...

Would like contribute a little tweek to this with installing gems on production.

create a gemrc (touch ~/.gemrc)

and then copy paste this to avoid fetching the gem docs on install, update and sources, plus some other defualts i use.

---
:verbose: true
:bulk_threshold: 1000
install: --no-ri --no-rdoc --env-shebang
:sources:
- http://gemcutter.org
- http://gems.rubyforge.org/
- http://gems.github.com
:benchmark: false
:backtrace: false
update: --no-ri --no-rdoc --env-shebang
:update_sources: true

Anybody know how to fire up the ruby console in the linode?

I prefer Heroku, so I can focus on development not setting up and maintaining servers. ;-)

This has got to be one of the most thoroughly researched and comprehensive Screencast produced by Ryan and he shines.

Nico Ritsche has a great point though. As one lone developer working on a big project (relatively speaking of course), I can simply rely on Heorku to do all my sys admin work and focus on the development. The Developer's docs are also of a high quality and very helpful.

Perhaps Ryan can do one screencast on Heroku?

Great job Ryan. Question: what if you're deploying to a test server first, then up to production? like so: Dev --> Test <--> Prod. The arrow goes both ways between test and prod because the prod database is sent to testing db and the testing code sent to prod server for releases.

Pushing from dev -> test is fine. But how about copying the code from test to prod? I don't want to ever push code from dev --> prod because multiple developers are working on the project and they're all at different points in the code.

I could simply scp the code over, but is there a more efficient way? Thanks all.

In Europe we smile about the pricing of VPS and RootServers Hosting's in USA ...

Linode: 512mb, 20Gb, 200Gb Traffice => $19.95
Hetzner: 512mb, 20Gb, 1Tb Traffic => €7.90
Heroku..unpayable..

When you work with clients you def. also need an Adminplatform for them to install their own email accounts etc..

Anyway thanks for this screencast..and all the others. So helpful.
Werner

I might be asking too much but where should seeds be run?

Im using carrier wave to store uploaded files, but when I do a deploy, those files get overwritten and are not longer accessible, cause they are in previous releases...I know I could use something like a3 for file storage, but I don't really require that right now...

Any ideas on what I might be able to do copy the previous deploys "upload" dir back into the "current" app?

Thanks!

anyone got a clue why this would happen to me?

sh -c 'cd /home/deployer/apps/appname/releases/20120324031224 && bundle exec rake RAILS_ENV=production RAILS_GROUPS=assets assets:precompile'
** [out :: 106.187.88.49] rake aborted!
** [out :: 106.187.88.49] FATAL: password authentication failed for user "deployer"

Can't figure it out to save my life. :)

Problem with assets:precompile
Thanks to ryan I can deploy my app with capistrano now but evertime come the assets:precompile command it takes longer and longer.

In fact assets subdirectories are added everytime :

app/public/assets/assets/assets/assets/assets etc .....

It looks like it compiles all assets recursively.
I thought it because I have : config.assets.precompile += ['*.js', '*.css'] but even if without the problem shows up.

Any idea ?

Fantastic screencast Ryan! So many rails developers are bewildered by the deployment process or have no idea how to setup the stack from scratch.

While capistrano is indeed a great tool and allows you to maintain very complex setups, but overtime found it to overkill for simple webapps with just single production server and maybe a staging server. Having tried heroku I really think git is the way to go in this case. Setting up post–receive hook on the the server is really very simple and all your deployment boils down to running 'git push production localbranch:master'.

With regards to unicorn I really see no reason you chose it as oposed to phusion passenger. It is an extra dependency that needs to be maintained. For me the beauty of using passenger lies in the fact that I don't need to maintain a separate service and make sure it is running at all times.

This is a great screencast, but I'm having an issue that I can't seem to get around.

I deployed my app, but it comes up with a 500 error every time (with my custom 500 error html page). I tried looking in the log, but it doesn't seem to be logging anything at all, so I'm at a loss to figure out what my error is.

Any ideas?

Thanks!

This screencast is brilliant -- much needed by many (obviously, from the comments!), and good to have out there.

I am a big fan of running the app as a special user, rather than root, so glad to see that.

It's also funny -- I love the occasional remarks like "this probably won't work the first time" for Capistrano. Understatement is the best form of humor ;-)

Finally a vote for a fantastic VPS host I have used since 2006 -- RimuHosting has great prices, many options, and best of all, really competent people to help when things go wrong.

How would you deploy to multiple stage environments with your unicorn configuration? Specifically, how to you change:

CMD="cd $APP_ROOT; bundle exec unicorn -D -c $APP_ROOT/config/unicorn.rb -E staging"

to

CMD="cd $APP_ROOT; bundle exec unicorn -D -c $APP_ROOT/config/unicorn.rb -E production"

when you run cap staging deploy or cap production deploy, respectively?

Troubleshooting tip:

If your cap deploy is failing, and you're logging into your server to troubleshoot manually, when you ssh to the server, if Agent forwarding is not already configured by default on your local machine, you'll need to enable it so you can communicate with github:

instead of:

ssh deployer@your-server.com

use:

ssh -A deployer@your-server.com

running postgresql fails. see error

perl: warning: Setting locale failed.
perl: warning: Please check that your locale settings:
        LANGUAGE = (unset),
        LC_ALL = (unset),
        LC_CTYPE = "et_EE.UTF-8",
        LANG = "et_EE.UTF-8"
    are supported and installed on your system.
perl: warning: Falling back to the standard locale ("C").
could not change directory to "/root"
psql: could not connect to server: No such file or directory
        Is the server running locally and accepting
        connections on Unix domain socket "/var/run/postgresql/.s.PGSQL.5432"?

Great episode Ryan. My only question was is there a reason you used rbenv over rvm to install ruby? Has rvm fallen out of favor? Are you using rbenv on OS X?

Ryan,

I followed these directions exactly but when I visit http://[ip] I get "Not Found: /". However, when I visit http://[ip]/[appname] it hits the app. Is it possible one of the paths in one of the configs is wrong?

Hey Ryan - it would truly be AWESOME if you could include your Linode server setup script and share it over there!

I have a problem, my nginx server is running with nginx user, and when the application try to access the /assets files i have a 403 http error. I changed the nginx user to deployer group but the issue persists. Can you help me to fix this? The only difference between your post and my server is i'm running in a centos distribution.

Hey, great video :)

Just a minor issue: I'm using branch other than master for deployment and in my opinion deploy.rb:44 should not include hard-coded branch.

I have an issue when I want to restart Unicorn, the pid and sock files aren't deleted when it is stopped so when Capistrano want to start Unicorn, it fails due to the files already present.

Anyone have an idea to fix this?

I'm trying to use Passenger and have now twice hosed my install.

I can successfully complete this line of the script in the show notes

# rbenv
curl -L https://raw.github.com/fesplugas/rbenv-installer/master/bin/rbenv-installer | bash
vim .bashrc # add rbenv to the top
. .bashrc
rbenv bootstrap-ubuntu-10-04
rbenv install 1.9.3-p125
rbenv global 1.9.3-p125
gem install bundler --no-ri --no-rdoc
rbenv rehash

But when I go to run

# Phusion Passenger (instead of Unicorn)
sudo apt-get -y install libcurl4-openssl-dev apache2-prefork-dev libapr1-dev libaprutil1-dev
sudo gem install passenger --no-ri --no-rdoc
sudo passenger-install-apache2-module
sudo vim /etc/apache2/apache2.conf # modify as instructed by installer

I get a sudo: gem: command not found error.

If I exit the deployer session and su deployer from root, I get this error: /home/deployer/.rbenv/bin/rbenv: line 20: cd: /root: Permission denied

Then running ruby -v gives me this error bash: ruby: command not found when previously that worked for deployer.

Any ideas what I'm doing wrong?

Ryan,

I wrote a very simple provisioning tool to help people setup a new instance on VPS. It has a bonus feature to setup / teardown Linode instances interactively, by choosing plans and datacenters, etc.

https://github.com/kenn/sunzi

I hope this helps to anyone who scratched their head dealing with chef / chef-solo!

Hi,

I got some problem about my app.
And Have no idea about where to look the development.log?

Great screencast :)

But I ran into a problem when Capistrano tries to build my assets. I have no idea what's going on, if I start rails locally, everything works fine!

  * executing `deploy:assets:precompile'
  * executing "cd /home/deployer/apps/app/releases/20120331112826 && bundle exec rake RAILS_ENV=production RAILS_GROUPS=assets assets:precompile"
    servers: ["example.com"]
    [example.com] executing command
    [example.com] sh -c 'cd /home/deployer/apps/app/releases/20120331112826 && bundle exec rake RAILS_ENV=production RAILS_GROUPS=assets assets:precompile'
 ** [out :: example.com] rake aborted!
 ** [out :: example.com]
 ** [out :: example.com] cannot load such file -- bcrypt
 ** [out :: example.com]
 ** [out :: example.com]
 ** [out :: example.com] (See full trace by running task with --trace)
 ** [out :: example.com]
    command finished in 5227ms
*** [deploy:update_code] rolling back

But in my Gemfile I have bccrypt-ruby set up.

gem 'bcrypt-ruby', :require => "bcrypt"

Is there a referral code for Linode we can use when we sign up, so Railscasts can receive a referral fee?

Ryan, Great show. One question though. Why Ubuntu 10.04 LTS and not Ubuntu 11.10? Is there any issues with Ruby or Rails on Ubuntu 11.10? The reason I'm asking is that I was going to try this with a spare machine I have lying around and it currently has 11.10 installed (nothing else). If there are issues then I will just re-roll it with 10.04 LTS

Hi Guys,

i don't really understood, how can deploy:setup successfully completed?
Because we can't create symlinks to non existing files.
Only after deploy:cold we get current folder with nginx.conf and unicorn_init.sh files.

 ** [out :: 172.17.11.10] creating symbolic link `/etc/nginx/sites-enabled/abc' to `/home/rails/apps/abc/current/config/nginx.conf'
 ** [out :: 172.17.11.10] : No such file or directory

Or I don't understand something.

How would I configure multiple websites on the same box? I know they would each be in a different directory by application name but how would I redirect using domain name. I know I would have to use apache virtual hosts but with Nginx I'm not sure.

Would it be better to have one IP address or multiple (looks like $1/IP on Linode).

Thanks Ryan!! Heroku and Engine Yard are great! But it is nice being able to test my applications live while keep cost in mind!

Hi, anyone got a clue why this would happen to me?

$ cap deploy:setup
./config/deploy.rb:3:in `load': compile error (SyntaxError)
./config/deploy.rb:3: syntax error, unexpected ':', expecting $end
server "176.34.96.44", :web, :app, :db , primary: true
                                                 ^
        from /usr/lib/ruby/gems/1.8/gems/capistrano-2.11.2/lib/capistrano/configuration/loading.rb:172:in `load_from_file'
        from /usr/lib/ruby/gems/1.8/gems/capistrano-2.11.2/lib/capistrano/configuration/loading.rb:89:in `load'
        from /usr/lib/ruby/gems/1.8/gems/capistrano-2.11.2/lib/capistrano/configuration/loading.rb:86:in `load'
        from /usr/lib/ruby/gems/1.8/gems/capistrano-2.11.2/lib/capistrano/configuration/loading.rb:86:in `each'
        from /usr/lib/ruby/gems/1.8/gems/capistrano-2.11.2/lib/capistrano/configuration/loading.rb:86:in `load'
        from Capfile:5:in `load'
        from /usr/lib/ruby/gems/1.8/gems/capistrano-2.11.2/lib/capistrano/configuration/loading.rb:172:in `load_from_file'
        from /usr/lib/ruby/gems/1.8/gems/capistrano-2.11.2/lib/capistrano/configuration/loading.rb:89:in `load'
        from /usr/lib/ruby/gems/1.8/gems/capistrano-2.11.2/lib/capistrano/configuration/loading.rb:86:in `load'
        from /usr/lib/ruby/gems/1.8/gems/capistrano-2.11.2/lib/capistrano/configuration/loading.rb:86:in `each'
        from /usr/lib/ruby/gems/1.8/gems/capistrano-2.11.2/lib/capistrano/configuration/loading.rb:86:in `load'
        from /usr/lib/ruby/gems/1.8/gems/capistrano-2.11.2/lib/capistrano/cli/execute.rb:65:in `load_recipes'
        from /usr/lib/ruby/gems/1.8/gems/capistrano-2.11.2/lib/capistrano/cli/execute.rb:65:in `each'
        from /usr/lib/ruby/gems/1.8/gems/capistrano-2.11.2/lib/capistrano/cli/execute.rb:65:in `load_recipes'
        from /usr/lib/ruby/gems/1.8/gems/capistrano-2.11.2/lib/capistrano/cli/execute.rb:31:in `execute!'
        from /usr/lib/ruby/gems/1.8/gems/capistrano-2.11.2/lib/capistrano/cli/execute.rb:14:in `execute'
        from /usr/lib/ruby/gems/1.8/gems/capistrano-2.11.2/bin/cap:4
        from /usr/bin/cap:19:in `load'
        from /usr/bin/cap:19

Thx :)

I followed this and am getting a 500 error complaining that application.css is not precompiled.

I logged into the server and ran rake assets:precompile without issues, but running the app gives the same error again.

Any ideas?

This is exactly the episode I was looking for. Ryan Bates is a leg-end!

One thing I'm not sure about:

Should there be a call to rbenv rehash inserted some where in one of the capistrano recipes. I tried adding paperclip to a project and deploying it but the application is not picking it up when it starts up. The recipe does run the bundle command to install new gems when doing a standard cap deploy but from what I can see from reading around I suspect the rbenv rehash needs to be called after the bundle command and I'm not sure if it is included in the recipe provided by Ryan in this example. If this is the case it would explain why my app is not picking up paperclip and is not starting. I'm not sure myself where or how I would make the call to rbenv rehash in the capistrano recipe.

Perhaps some would be kind enough to point me in the right direct?

Hello, i have questions about multi rails apps with unicorn. If we want to add two or more rails application with unicorn in nginx, how can we create the multi-unicorn-process startup/shutdown script?

I googled more of them somebody prefered god gem but i couldnt find proper solution for this.

Can you have any idea?

Great screencast! I'm wondering how I would configure the deployment to support an app that offers users basecamp style subdomains?

Any tips on getting hstore to work on Linode following these directions? I am seeing:

PG::Error: ERROR: could not open extension control file "/usr/share/postgresql/9.1/extension/hstore.control": No such file or directory

Hey guys, two questions. First, can anyone point me to a good tutorial that would help with adding a 2nd application on the same server with a configuration like this. Second, lets say I wanted my app on Heroku to act as a staging app. Could you use a setup like this to pull from Heroku rather than Github?

Hello!

May scripts from /etc/init.d/ be run by root at startup time?
If so, we have vulnerability allows "deployer" user change unicorn_init.sh and get root access, havent' we?

sudo "ln -nfs #{current_path}/config/unicorn_init.sh /etc/init.d/unicorn_#{application}"

Hello!

Why rbenv, not rvm ??

I am using Mongodb/mongoid with my Rails project. I followed the steps given. Most of things worked. But I am stuck at making mongo to work.

1. the screencasts makes heavy use of config/database.yml. Should I do the same with config/mongoid.yml and make copy to config/mongoid.example.yml? What other files should I modify?

2 I suppose I should install Mongodb on Linode server and just keep it runing and pointing the production server to localhost in my mongoid.yml file. Does that sound right?

3 I wanted to make things generally working without setting up mongoid.yml correctly. But then because I have database.yml there with unused sqlite3 reference, cap deploy:cold gives error about sqlite3 not in the gemfile. I suppose if I remove config/database.yml altogether, it should not ask me to include sqlite3 any more?

These questions might be too specific to my situation. But any links to relevant information that can help me would be really appreciated.

On 12.04, I ran into an error about the admin group not existing. All I had to do to fix it was run groupadd admin. The group is already listed in the sudoers file, it just wasn't created when the distro was installed.

Very helpful!! Thanks

I wanted to check with the community is this is still working for everybody.

I am having issues when using unicorn with a Capistrano deployment. From what I have been able to understand Capistrano uses a scheme in wich every release is deployed inside the releases directory under a unique name and if the transaction was successful, creates a symlink named current that will point to that release.

So I end up with a deployment directory such as:

/home/deployer/apps/sample_app/current

Then when I try to start unicorn from the binstubs directory all the unicorn methods look for things in the following path, particularly in the configurator.rb module:

/home/deployer/apps/sample_app

I haven't been able to fully understand how unicorn sets the working_directory from here:

https://github.com/defunkt/unicorn/raw/master/lib/unicorn/configurator.rb

But I wanted to check with the community if I am missing something evident due to the noob nature in me.

Is there a way to view a log output from Unicorm like WEBrick? My app is deploying correctly, but isn't running correctly off the sever. I'm just getting the useless error pages.

I'm getting an error when I first run cap deploy. I have put up the error I'm getting here but can't seem to find the soloution anywhere.

Error with cp

Any ideas?

Can anyone suggest an EC2 image to use?

My linode had to get restarted yesterday and I noticed that while the nginx service started properly, the unicorn socket didn't come back. I wonder how the deployment needs to be tweaked for events where the server has to get restarted. I assume it's a matter of running bundle exec unicorn on server startup.

for some reason, at the last step on cap deploy:cold it is not allowing me access. this is the output:

  * executing `deploy:start'
  * executing "/etc/init.d/unicorn_mock_app start"
    servers: ["MYIP"]
    [MYIP] executing command
 ** [out :: MYIP] sh: /etc/init.d/unicorn_mock_app: Permission denied
    command finished in 150ms
failed: "sh -c '/etc/init.d/unicorn_mock_app start'" on IP

I tried changing the permission to the symlink by running:

chmod +x /etc/init.d/unicorn_mock_app

But then it outputs

-bash: /etc/init.d/unicorn_mock_app: /bin/sh^M: bad interpreter: No such file or directory

any ideas?? on why it wont allow permissions for the app to start

I did all the instructions and successfully deployed, but then when i tried to access it, it was automatically redirect my browser to https://IP.

This is what Chrome said:

Error 102 (net::ERR_CONNECTION_REFUSED): The server refused the connection.

I though it was because i forced the SSL in production.rb and then i set it to false and redeployed. But nothing changed!

After that i set the ssl force back and try to install SSL using the instruction in http://railscasts.com/episodes/357-adding-ssl, and i had a problem in nginx when i tried to install:

Restarting nginx: nginx: [emerg] unknown directive "server" in /etc/nginx/nginx.conf:10
nginx: configuration file /etc/nginx/nginx.conf test failed

Can anyone give me a solution for this? Thanks in advance :)

Does anyone know how to install the unaccent extension for pg?

PG::Error: ERROR: could not open extension control file "/usr/share/postgresql/9.1/extension/unaccent.control": No such file or directory

it happened to me

servers: ["my.com"]
[my.com] executing command
** [out :: my.com] cp:
** [out :: my.com] cannot create directory `/home/rails-apps/my-beta/releases/20120620085906'
** [out :: my.com] : No such file or directory
** [out :: my.com]

For some of the strangest reason i cannot seem to solve this to save myself .

[106.187.36.187] executing command
** [out :: 106.187.36.187] rake aborted!
** [out :: 106.187.36.187] Invalid CSS after "...f.eot);src:url(": expected > expression (e.g. fr, 2n+1), was "/assets/"/asset..."
** out :: 106.187.36.187

Which leads to :
> failed: "sh -c 'cd /home/deployer/apps/app_name/releases/20120624151117 && bundle exec rake RAILS_ENV=production RAILS_GROUPS=assets assets:precompile'" on 106.187.36.187

This bug just appeared just out the blue . Have tried for hours so if anyone else knows how to fix , please save me.

Great tut!

But does anyone of you have any problem with openssl savon and the guide above?

I get

OpenSSL::SSL::SSLError (SSL_connect returned=1 errno=0 state=SSLv2/v3 read server hello A: sslv3 alert unexpected message):

Any tips? Do a need to generate a certificate on the server (guide does not state that).

Problem

I've been fighting with that one for a while now. That is the cry for help. As suggested I've going through all the commands and executed them on the remote server.
The command that fails on the server is:

sh -c "rm -rf /home/deployer/apps/blog/releases/20120726142915/public/assets &&\\\n        mkdir -p /home/deployer/apps/blog/releases/20120726142915/public &&\\\n        mkdir -p /home/deployer/apps/blog/shared/assets &&\\\n        ln -s /home/deployer/apps/blog/shared/assets /home/deployer/apps/blog/releases/20120726142915/public/assets"

It says: \n not found. As far as I can tell, it doesn't like the new lines. When I put them away manually - everything goes OK.

The interesting this is - the deployment script seem to execute this command OK, and it is the next one it stables upon (bundle install). Although when executing manually on the sever - bundles get installed OK.

What can it be? And how, and where, do I edit those new lines "\\n" out?

Complete Output:

λ config → λ git master → cap deploy:cold
  * executing `deploy:cold'
  * executing `deploy:update'
 ** transaction: start
  * executing `deploy:update_code'
    updating the cached checkout on all servers
    executing locally: "git ls-remote git@github.com:andremaha/blog.git master"
    command finished in 4916ms
  * executing "if [ -d /home/deployer/apps/blog/shared/cached-copy ]; then cd /home/deployer/apps/blog/shared/cached-copy && git fetch -q origin && git fetch --tags -q origin && git reset -q --hard 6fb7914a0c09159399ce85f00073692c4cd32a8a && git clean -q -d -x -f; else git clone -q git@github.com:andremaha/blog.git /home/deployer/apps/blog/shared/cached-copy && cd /home/deployer/apps/blog/shared/cached-copy && git checkout -q -b deploy 6fb7914a0c09159399ce85f00073692c4cd32a8a; fi"
    servers: ["50.112.249.58"]
    [50.112.249.58] executing command
    command finished in 12628ms
    copying the cached version to /home/deployer/apps/blog/releases/20120726142915
  * executing "cp -RPp /home/deployer/apps/blog/shared/cached-copy /home/deployer/apps/blog/releases/20120726142915 && (echo 6fb7914a0c09159399ce85f00073692c4cd32a8a > /home/deployer/apps/blog/releases/20120726142915/REVISION)"
    servers: ["50.112.249.58"]
    [50.112.249.58] executing command
    command finished in 706ms
  * executing `deploy:finalize_update'
    triggering before callbacks for `deploy:finalize_update'
  * executing `deploy:assets:symlink'
  * executing "rm -rf /home/deployer/apps/blog/releases/20120726142915/public/assets &&\\\n        mkdir -p /home/deployer/apps/blog/releases/20120726142915/public &&\\\n        mkdir -p /home/deployer/apps/blog/shared/assets &&\\\n        ln -s /home/deployer/apps/blog/shared/assets /home/deployer/apps/blog/releases/20120726142915/public/assets"
    servers: ["50.112.249.58"]
    [50.112.249.58] executing command
    command finished in 693ms
  * executing `bundle:install'
  * executing "cd /home/deployer/apps/blog/releases/20120726142915 && bundle install --gemfile /home/deployer/apps/blog/releases/20120726142915/Gemfile --path /home/deployer/apps/blog/shared/bundle --deployment --quiet --without development test"
    servers: ["50.112.249.58"]
    [50.112.249.58] executing command
 ** [out :: 50.112.249.58] sh: 1:
 ** [out :: 50.112.249.58] bundle: not found
 ** [out :: 50.112.249.58] 
    command finished in 679ms
*** [deploy:update_code] rolling back
  * executing "rm -rf /home/deployer/apps/blog/releases/20120726142915; true"
    servers: ["50.112.249.58"]
    [50.112.249.58] executing command
    command finished in 683ms
failed: "sh -c 'cd /home/deployer/apps/blog/releases/20120726142915 && bundle install --gemfile /home/deployer/apps/blog/releases/20120726142915/Gemfile --path /home/deployer/apps/blog/shared/bundle --deployment --quiet --without development test'" on 50.112.249.58

Hey guys, I have followed the instructions to the line.. I got to here;

deployer@ve:~$ rbenv install 1.9.3-p125

Downloading http://pyyaml.org/download/libyaml/yaml-0.1.4.tar.gz...
Installing yaml-0.1.4...

BUILD FAILED

Inspect or clean up the working tree at /tmp/ruby-build.20120728204839.7741
Results logged to /tmp/ruby-build.20120728204839.7741.log

Last 10 log lines:
yaml-0.1.4/win32/vs2008/run_parser.vcproj
yaml-0.1.4/win32/vs2008/yaml.vcproj
yaml-0.1.4/win32/vs2008/test_version.vcproj
yaml-0.1.4/win32/vs2008/run_emitter.vcproj
yaml-0.1.4/win32/vs2008/example_reformatter.vcproj
yaml-0.1.4/win32/vs2008/run_loader.vcproj
yaml-0.1.4/win32/vs2008/yamldll.vcproj
yaml-0.1.4/win32/config.h
/tmp/ruby-build.20120728204839.7741/yaml-0.1.4 /tmp/ruby-build.20120728204839.7741 ~
/home/deployer/.rbenv/plugins/ruby-build/bin/ruby-build: ./configure: /bin/sh: bad interpreter: Permission denied

... little help?

Thank you Ryan. After struggling for a few days with Linode/Debian6/rvm/capistrano/passenger/enginx I gave up and followed this screencast to success in a couple of hours. I decided to follow exactly so my stack is the same as yours. I am sure I was doing terrible things with rvm. My brain isn't quite as large as Wayne's and rvm does a lot of magic with system commands such as cd. Even rvm implode didn't seem to uninstall without side effects. Hence the reinstall of the OS. I think rvm is powerful & useful and would love to see an in depth railscast on it.

Great Railscast! I've always used Heroku. Is there an equivalent command to 'heroku logs' on VPS with Nginx?

Thanks!

Thanks Ryan. I am getting "500 Internal Server Error" error after installing nginx. Not sure what I am missing, even though I am following exact steps mentioned in the web cast.

Hey guys i am getting this error:

   command finished in 6676ms
  * executing `deploy:start'
  * executing "/etc/init.d/unicorn_default start"
    servers: ["10.0.14.17"]
    [10.0.14.17] executing command
 ** [out :: 10.0.14.17] sh: /etc/init.d/unicorn_default: not found
    command finished in 126ms
failed: "sh -c '/etc/init.d/unicorn_default start'" on 10.0.14.17

any ideas?

this is at the cap deploy:cold step

I'm getting a

failed: "sh -c 'cd /home/deployer/apps/angel_list_api_research/releases/20120827050009 && bundle install --gemfile /home/deployer/apps/angel_list_api_research/releases/20120827050009/Gemfile --path /home/deployer/apps/angel_list_api_research/shared/bundle --deployment --quiet --without development test'" on 198.58.96.39

Has anyone any idea what this is trying to tell me?

I have no issues compiling assets with my rvm setup on my current hosted env. Going thru these steps (using rbenv) I can't get the assets to compile. Really strange stacktrace

RAILS_ENV=production bundle exec rake assets:precompile --trace
** Invoke assets:precompile (first_time)
** Execute assets:precompile
/home/deployer/.rbenv/versions/1.9.3-p194/bin/ruby /home/deployer/apps/Merj-Web/shared/bundle/ruby/1.9.1/bin/rake assets:precompile:all RAILS_ENV=production RAILS_GROUPS=assets --trace
** Invoke assets:precompile:all (first_time)
** Execute assets:precompile:all
** Invoke assets:precompile:primary (first_time)
** Invoke assets:environment (first_time)
** Execute assets:environment
** Invoke environment (first_time)
** Execute environment
** Invoke tmp:cache:clear (first_time)
** Execute tmp:cache:clear
** Execute assets:precompile:primary
rake aborted!
Command failed with status (): [/home/deployer/.rbenv/versions/1.9.3-p194/...]
/home/deployer/apps/Merj-Web/shared/bundle/ruby/1.9.1/gems/rake-0.9.2.2/lib/rake/file_utils.rb:53:in block in create_shell_runner'
/home/deployer/apps/Merj-Web/shared/bundle/ruby/1.9.1/gems/rake-0.9.2.2/lib/rake/file_utils.rb:45:incall'
/home/deployer/apps/Merj-Web/shared/bundle/ruby/1.9.1/gems/rake-0.9.2.2/lib/rake/file_utils.rb:45:in sh'
/home/deployer/apps/Merj-Web/shared/bundle/ruby/1.9.1/gems/rake-0.9.2.2/lib/rake/file_utils_ext.rb:39:insh'
/home/deployer/apps/Merj-Web/shared/bundle/ruby/1.9.1/gems/rake-0.9.2.2/lib/rake/file_utils.rb:80:in ruby'
/home/deployer/apps/Merj-Web/shared/bundle/ruby/1.9.1/gems/rake-0.9.2.2/lib/rake/file_utils_ext.rb:39:inruby'
/home/deployer/apps/Merj-Web/shared/bundle/ruby/1.9.1/gems/actionpack-3.2.8/lib/sprockets/assets.rake:12:in ruby_rake_task'
/home/deployer/apps/Merj-Web/shared/bundle/ruby/1.9.1/gems/actionpack-3.2.8/lib/sprockets/assets.rake:21:ininvoke_or_reboot_rake_task'
/home/deployer/apps/Merj-Web/shared/bundle/ruby/1.9.1/gems/actionpack-3.2.8/lib/sprockets/assets.rake:29:in block (2 levels) in <top (required)>'
/home/deployer/apps/Merj-Web/shared/bundle/ruby/1.9.1/gems/rake-0.9.2.2/lib/rake/task.rb:205:incall'
/home/deployer/apps/Merj-Web/shared/bundle/ruby/1.9.1/gems/rake-0.9.2.2/lib/rake/task.rb:205:in block in execute'
/home/deployer/apps/Merj-Web/shared/bundle/ruby/1.9.1/gems/rake-0.9.2.2/lib/rake/task.rb:200:ineach'
/home/deployer/apps/Merj-Web/shared/bundle/ruby/1.9.1/gems/rake-0.9.2.2/lib/rake/task.rb:200:in execute'
/home/deployer/apps/Merj-Web/shared/bundle/ruby/1.9.1/gems/rake-0.9.2.2/lib/rake/task.rb:158:inblock in invoke_with_call_chain'
/home/deployer/.rbenv/versions/1.9.3-p194/lib/ruby/1.9.1/monitor.rb:211:in mon_synchronize'
/home/deployer/apps/Merj-Web/shared/bundle/ruby/1.9.1/gems/rake-0.9.2.2/lib/rake/task.rb:151:ininvoke_with_call_chain'
/home/deployer/apps/Merj-Web/shared/bundle/ruby/1.9.1/gems/rake-0.9.2.2/lib/rake/task.rb:144:in invoke'
/home/deployer/apps/Merj-Web/shared/bundle/ruby/1.9.1/gems/rake-0.9.2.2/lib/rake/application.rb:116:ininvoke_task'
/home/deployer/apps/Merj-Web/shared/bundle/ruby/1.9.1/gems/rake-0.9.2.2/lib/rake/application.rb:94:in block (2 levels) in top_level'
/home/deployer/apps/Merj-Web/shared/bundle/ruby/1.9.1/gems/rake-0.9.2.2/lib/rake/application.rb:94:ineach'
/home/deployer/apps/Merj-Web/shared/bundle/ruby/1.9.1/gems/rake-0.9.2.2/lib/rake/application.rb:94:in block in top_level'
/home/deployer/apps/Merj-Web/shared/bundle/ruby/1.9.1/gems/rake-0.9.2.2/lib/rake/application.rb:133:instandard_exception_handling'
/home/deployer/apps/Merj-Web/shared/bundle/ruby/1.9.1/gems/rake-0.9.2.2/lib/rake/application.rb:88:in top_level'
/home/deployer/apps/Merj-Web/shared/bundle/ruby/1.9.1/gems/rake-0.9.2.2/lib/rake/application.rb:66:inblock in run'
/home/deployer/apps/Merj-Web/shared/bundle/ruby/1.9.1/gems/rake-0.9.2.2/lib/rake/application.rb:133:in standard_exception_handling'
/home/deployer/apps/Merj-Web/shared/bundle/ruby/1.9.1/gems/rake-0.9.2.2/lib/rake/application.rb:63:inrun'
/home/deployer/apps/Merj-Web/shared/bundle/ruby/1.9.1/gems/rake-0.9.2.2/bin/rake:33:in <top (required)>'
/home/deployer/apps/Merj-Web/shared/bundle/ruby/1.9.1/bin/rake:23:inload'
/home/deployer/apps/Merj-Web/shared/bundle/ruby/1.9.1/bin/rake:23:in `'
Tasks: TOP => assets:precompile

hey guys tried the guide again and i am again getting the following error:

fighterpilot@JJDRORPROD01:~/jjdowns$ cap deploy:cold
  * executing `deploy:cold'
  * executing `deploy:update'
 ** transaction: start
  * executing `deploy:update_code'
    updating the cached checkout on all servers
    executing locally: "git ls-remote gitolite@10.0.14.14:jjdowns master"
    command finished in 166ms
  * executing "if [ -d /home/fighterpilot/jjdowns/shared/cached-copy ]; then cd /home/fighterpilot/jjdowns/shared/cached-copy && git fetch -q origin && git fetch --tags -q origin && git reset -q --hard 019cbef11c95698d7e8e7788e04fd77d07c47ae7 && git clean -q -d -x -f; else git clone -q gitolite@10.0.14.14:jjdowns /home/fighterpilot/jjdowns/shared/cached-copy && cd /home/fighterpilot/jjdowns/shared/cached-copy && git checkout -q -b deploy 019cbef11c95698d7e8e7788e04fd77d07c47ae7; fi"
    servers: ["10.0.14.17"]
    [10.0.14.17] executing command
 ** [10.0.14.17 :: out] Error writing to authentication socket.
 ** [10.0.14.17 :: out] Error writing to authentication socket.
    command finished in 1111ms
    copying the cached version to /home/fighterpilot/jjdowns/releases/20120909041546
  * executing "cp -RPp /home/fighterpilot/jjdowns/shared/cached-copy /home/fighterpilot/jjdowns/releases/20120909041546 && (echo 019cbef11c95698d7e8e7788e04fd77d07c47ae7 > /home/fighterpilot/jjdowns/releases/20120909041546/REVISION)"
    servers: ["10.0.14.17"]
    [10.0.14.17] executing command
 ** [out :: 10.0.14.17] cp: cannot create directory `/home/fighterpilot/jjdowns/releases/20120909041546': Permission denied
    command finished in 141ms
*** [deploy:update_code] rolling back
  * executing "rm -rf /home/fighterpilot/jjdowns/releases/20120909041546; true"
    servers: ["10.0.14.17"]
    [10.0.14.17] executing command
    command finished in 140ms
failed: "sh -c 'cp -RPp /home/fighterpilot/jjdowns/shared/cached-copy /home/fighterpilot/jjdowns/releases/20120909041546 && (echo 019cbef11c95698d7e8e7788e04fd77d07c47ae7 > /home/fighterpilot/jjdowns/releases/20120909041546/REVISION)'" on 10.0.14.17
fighterpilot@JJDRORPROD01:~/jjdowns$ 

any ideas?

Awesome screencast! I've used this multiple times for reference and works great. However, I recently tried following this example, but instead of installing rbenv I had to use rvm. And when I get to the point of running 'cap deploy:cold' I keep getting the following error: sh: 1: bundle: not found. I installed bundler correctly and confirmed it by running 'bundle -v'. One thing I did notice in previous apps when following this tutorial (when using rbenv) that under my_app_directory/shared there would be a bundle/ directory, however now that directory doesn't exist, so maybe that is the 'bundle not found'? Any ideas?

So i'm having a weird issue - followed the instructions to the tee on a new EC2 instance running Ubuntu 12.04 (only wrinkle was in the rbenv install process, found a good tutorial explaining the differences here: http://www.stehem.net/2012/05/08/how-to-install-ruby-with-rbenv-on-ubuntu-12-04.html

So i THINK i'm in good shape! UNTIL I try to cap deploy: the connection hangs AFTER connecting successfully..?! Turned on SSH logging, and the key exchange is successful, but then the script doesn't continue. I CAN login without a password from the terminal, so I know my key is set up correctly on my deployer user.

It's all a little odd.

I love this railscast. I have had to make some changes and such to the unicorn configurations to work with my platform, but overall it works well. Nice starting point.

Hi all...

I've tried to execute this railscast but i received this error.

failed: "sh -c 'cd /home/webadmin/apps/apptest/releases/20121002133503 && bundle install --gemfile /home/webadmin/apps/apptest/releases/20121002133503/Gemfile --path /home/webadmin/apps/apptest/shared/bundle --deployment --quiet --without development test'" on 10.X.X.X

It seems the copy cache is not copied in the release folder...

Thanks for your help

Hi,

I've a new question.

Which is the best way to completely remove an apps deployed by capristano.

nginx vhost unicorn_init script,....

Thanx

Bloody hell, that's complicated!

I am running on windows ,,,and I usually used to deploy to Heroku ,and I want to try linode . I have a question "deploying from different operating system like windows which I am using ,could work ?"
regards

Hi, thanks for this Screencast.

I have some troubles installing postgresql.

When typing in:
apt-get install postgresql libpq-dev

I get: could not find libpg-dev

Does somebody knows the problem?

Thanks

This is nice it solved my problem

Thanks

I've found a very similar article:

http://rubysource.com/deploying-a-rails-application/

(maybe copied from you ryan)

do you know how to solve these problem ..I searched a lot with no answer
/etc/init.d/unicorn_depot_2: Permission denied

It deployed correctly the first time but after second deploy it stopped working. I was getting problems with nginx and unicorn.

The problem was related to permission for unicorn_blog service. In deploy.rb we are starting the app with normal permissions.

run "/etc/init.d/unicorn_#{application} #{command}"

After changing it to

sudo "/etc/init.d/unicorn_#{application} #{command}"

everything started to work.

I don't know why it is necessary. I followed all instructions correctly.

Hi everyone,

First of all, many thanks Ryan for this amazing cast. I just wanted to know is there a way this can be applied in a scenario where I would like to use my VPN to host multiple rails applications?

For example if my domain is www.myrailsapps.com and I have three apps that I want to deploy on it which can be accessed at myrailsapps.com/app1, myrailsapps.com/app2, etc.

Anyone know if this can be accomplished with minor tweaks to this particular method?

Thanks again.

The screencast adn everyone's comments have really helped me through setting up the server. I feel a lot better about server setup, now.

Everything is working great - went through the whole screencast - the only difference for me is that I'm using mysql. Mysql is installed properly, and upon deploy, running the db migrations works and results in the tables being created in the database. So I know it's connected properly.

However, when I go to one of the URLs for that scaffolding (for me it was /users), I'm getting a "we're sorry, something went wrong" error.

The routes work perfectly locally. Not sure what the issues is because MySQL seems to be working and connected based on the migrations?

If anyone has any suggestions (or pointers to help debug these server errors), i would very much appreciate them.

This is an absolutely ASTOUNDING tutorial/runthrough - one of the best I've ever come across on anything Rails-related. I will admit it took me around 2 days involving a couple of attempts and a bit of debugging to get everything working, but I got there in the end.

From memory, the main issues I had to deal with were as follows:

1) As Steve Fox mentions above, you will probably need to edit the /etc/nginx/sites-enabled/default file as per Steve's instructions in order to get nginx to work properly.

2) For some reason, when I tried to install PostgreSQL, version 8 was installed on my server, whereas version 9 appears to be installed in the video for Ryan's installation. I had to make sure I added this to the relevant command in the terminal. From memory (so you might want to check), it was:

apt-get install postgresql-9.2 libpq-dev

(Or whatever the current version is at the point you are reading this :-) ). See also this link.

3) An obvious one, but make sure that you install the same versions of Rails, PostgreSQL etc. on your server that you have on your local machine!

4) If your gems fail to install properly on the server, you may need to install the relevant development files:

sudo apt-get install libxslt-dev libxml2-dev

See this StackOverflow question/response, and particularly the response with the most upvotes for further info.

HOWEVER, the most important piece of advice I can give based upon my experiences is as follows:

If you're intending to deploy your first ever app, do not build your complete app locally and then attempt to deploy it. There are probably minor differences in the different installations which are likely to cause your 500 page to appear, and which may end up being a nightmare to debug.

Instead, my advice would be to set up your server as soon as you write the first line of code for your app (even if it's just a "Hello World!") - and then deploy and debug as you go along. Believe me, this saves the frustration (particularly if you're a novice and not quite sure of what you're doing) of having a complete, working app on your local machine and one that refuses to work on a server. If you carry out a staged deployment, any issues should be identifiable and therefore rectifiable fairly quickly, rather than trying to debug a complete app at the end of the process. If you're worried about having an unfinished app publically available, I suppose you could set a username/password using the nginx equivalent of Apache's htpassword (do a Google search) to the root directory so that visitors to your site won't be able to access it.

Hope this helps others - and thanks again, Ryan!

Hi Ryan,
I followed this railscast and it helped me a lot in setting up a server recently however there is one weird issue I am encountering right now.

Whenever I enter my domain like www.example.com it gets redirected to the IP of the linode box having the rails application like xx.xxx.xx.xx. I have no idea why I am having this issue.I know it has something to do with Nginx configuration. Can you guide me what to modify to get this working ?

I've been struggling for a while now getting multiple apps online on one server so I hope someone can help me...

I've followed this tutorial to the letter apart from a few version changes (Ubuntu Server 12.04, rbenv 1.9.3p194).

Everything works fine with one website but as soon as I add another website things start to go wrong. At first nginx doesn't start and gives an error that I need to encrease my server_names_hash_bucket_size: to 64 which it's kinda weird because my domainnames for both sites aren't that long.

After changing that nginx starts again and the first site works but the second one gives an error. What's also weird is that my first sites stays the default site accessible by the ip of the server and all the domains pointing to it even though I have listen 80 default deferred; commented out.

I use this nginx.conf for both site with only the server_name different ofcourse:

upstream unicorn-<%= application %> {
  server unix:/tmp/unicorn.<%= application %>.sock fail_timeout=0;
}

server {
  # listen 80 default deferred;
  server_name slimmetelefoon.info;
  root <%= current_path %>/public;

  location ^~ /assets/ {
    gzip_static on;
    expires max;
    add_header Cache-Control public;
  }

  try_files $uri/index.html $uri @unicorn;
  location @unicorn {
    proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
    proxy_set_header Host $http_host;
    proxy_redirect off;
    proxy_pass http://unicorn-<%= application %>;
  }

  error_page 500 502 503 504 /500.html;
  client_max_body_size 4G;
  keepalive_timeout 10;
}

I was having trouble with the add-apt-repository command on Ubuntu 12.10 - Command not Found, even after doing sudo apt-get install python-software-properties

This was the error:
sudo: apt-add-repository: command not found

This was the solution:
sudo apt-get install software-properties-common

After that it started working finally.

Ignore my post 2 posts up. Turned my config was working fine.. Didn't realize they weren't being updated with normal cap deploys so my server still had the old ones in etc/nginx/sites-enabled...

To add to my original comment above, and to reiterate what a couple of other posters have suggested, I would love a follow-up tutorial to this showing how to deploy a new (separate) Rails app on the same server.

The reason I ask is that, given that I am a lone developer (and, probably like most folks here, my apps are generally quite small), my app only takes up around 5-10% of the allocated Linode so, when it comes round to deploying a second app (which will probably be just as tiny), it would appear to be more sensible to deploy it within the spare space rather than purchase an extra Linode.

There appear to be a few solutions above, but they are a bit daunting for a noob (particularly when one's current app just "works" and you're wary of breaking anything!), so a short screencast would be just dandy.

On a related note - Linode offer a backup service for around $5 a month (for the basic server option), so I'd recommend getting this if you're a noob like me. I think nightly backups are taken.

Have sometimes problem with "Cannot allocate memory" when mail sending or image uploading... but I have 1 Gb ram for not high loaded app :(

Thanks a lot Ryan for this and the other Railscasts that you have done on deployment.

On Capistrano's from-the-beginning wiki, there is a section on avoiding deploy:cold which says

First of all, deploy:cold was a bad idea. It attempted to do what I have described in this tutorial, but without enough opinions, and without enough options for configuration. I’ve come to the conclusion that setting up a new app on a new machine is not something that can be generically automated.

Ryan, would you suggest to avoid deploy:cold?

cap deploy:cold seems to break on executing deploy:start - followed every step along the way. Puzzled on this one. Anybody else have this issue?

  * 2012-12-11 09:13:18 executing `deploy:start'
  * executing "/etc/init.d/unicorn_blog start"
    servers: ["100.100.100.100"]
    [100.100.100.100] executing command
 ** [out :: 100.100.100.100] kill: 84: No such process
 ** [out :: 100.100.100.100] 
    command finished in 1436ms

For those of you having troubles when using routes containing subdomains in combination with Nginx (v1.2.5), you have to modify Ryan's Nginx configuration file like so:

Change this:

location @unicorn {
  #...
  proxy_set_header Host $http_host;
  #...
}

To this:

location @unicorn {
  #...
  proxy_set_header Host $host;
  #...
}

Restart, and you should be good to go!

The $http_host variable seems to only return the request subdomain.

I try to add admin user on Ubuntu 12.04 LTS:

$ adduser deployer --ingroup admin
adduser: The group `admin' does not exist.

Does anyone know how to fix that?

So let's start by saying Thanks, Ryan, great cast, really helpful and all that jazz.. (I mean it though)
Just one difference I'd like to share that I encountered on a fresh Ubuntu 12.04 VPS installation following the instructions of this screencast. After installing postgres, and adding a project specific user with password, I also had to modify the pg_hba.conf (/etc/postgresql/9.1/main/pg_hba.conf) file to enable the md5 authentication method, so that rails is able to connect to it with user and password specified in database.yml. Details are easily googleable, just wanted to point it out..

Hi Ryan et al.

Here's hoping someone can help. After setting up my linode acct & deploying, a couple links did not work (production only). So I made a few changes, thinking it was a cancan incompatibility with to_params, and now none of the link_to's work. In the production log it looks like it's trying to serve up a request to another host:

Started POST "/slingads/index.php" for 203.142.100.15 at 2013-02-08 16:50:22 -05
00
Error…500 error.
ActionController::RoutingError (No route matches [POST] "/slingads/index.php"):
actionpack (3.2.11)

Do I have a security breach & how should I handle this? Would it be best to just dump the database and start over? Any advice is much appreciated.
Thanks,
Cathy
PS) That ip address is for a company in HongKong!

Im struggling to get this solution to start with the environment set as staging.

I have tried updating the unicorn init file to pass the -E argument:

CMD="cd <%= current_path %>; bundle exec unicorn -D -c <%= unicorn_config %> -E <%= stage %>"

But Rails.env always says production.

Any ideas?

Hi Ryan!
It will be great if you can make a screencast showing us how to deploy an app on 2 servers, one for the Rails app itself and the other one for the database( Postgres). This is probably the first step I might do when I want to scale an app, but still, I can't find materials about how to do it right.

Thank you!

Hi can you help me
I'm trying to open rails console and getting this
deployer@microv:~$ rails console
-bash: rails: command not found

hey can anyone please help me with the following problem:

• executing "kill -s USR2 'cat /tmp/unicorn.sample.pid'" servers: ["server"] [server] executing command ** *** [err :: server] bash: line 0: kill: cat /tmp/unicorn.sample.pid: arguments must be process or job IDs command finished in 1023ms** failed: "rvm_path=/usr/local/rvm /usr/local/rvm/bin/rvm-shell '1.9.3' -c 'kill -s USR2 '\''cat /tmp/unicorn.sample.pid'\'''" on server

I also get the bundle install error:


failed: "sh -c 'cd /home/deployer/apps/burden_farming/releases/20120322200251 && bundle install --gemfile


But it does not tell much, how to debug? How can I find out which gem dependency is causing the error?

I'm finding Capistrano a little daunting, and am trying to set up NGINX and Unicorn on my VPS.

Unicorn_rails works like a charm right now when called from the app_root, but then I'm stuck with my terminal open.

Can't get the unicorn command to work, let alone start it up on VPS reboot. I get:

/usr/local/rvm/gems/ruby-1.9.3-p385/gems/unicorn-4.6.0/lib/unicorn/configurator.rb:634:in `parse_rackup_file': rackup file (config.ru) not readable (ArgumentError)
from /usr/local/rvm/gems/ruby-1.9.3-p385/gems/unicorn-4.6.0/lib/unicorn/configurator.rb:77:in `reload'
from /usr/local/rvm/gems/ruby-1.9.3-p385/gems/unicorn-4.6.0/lib/unicorn/configurator.rb:68:in `initialize'
from /usr/local/rvm/gems/ruby-1.9.3-p385/gems/unicorn-4.6.0/lib/unicorn/http_server.rb:108:in `new'
from /usr/local/rvm/gems/ruby-1.9.3-p385/gems/unicorn-4.6.0/lib/unicorn/http_server.rb:108:in `initialize'
from /usr/local/rvm/gems/ruby-1.9.3-p385/gems/unicorn-4.6.0/bin/unicorn:126:in `new'
from /usr/local/rvm/gems/ruby-1.9.3-p385/gems/unicorn-4.6.0/bin/unicorn:126:in `<top (required)>'
from /usr/local/rvm/gems/ruby-1.9.3-p385/bin/unicorn:19:in `load'
from /usr/local/rvm/gems/ruby-1.9.3-p385/bin/unicorn:19:in `<main>'
from /usr/local/rvm/gems/ruby-1.9.3-p385/bin/ruby_noexec_wrapper:14:in `eval'
from /usr/local/rvm/gems/ruby-1.9.3-p385/bin/ruby_noexec_wrapper:14:in `<main>'

Also, I keep hearing that unicorn is supposed to create a /tmp/unicorn.pid file, but I can only ever find /tmp/server.pid is that the right one?

Can Anyone help on that?

Hey all very green here, first time using Linode

(early on in the tutorial)
When running:

root@li511-230:~# service nginx start

I get:

nginx: [emerg] bind() to [::]:80 failed (98: Address already in use)
nginx: [emerg] bind() to [::]:80 failed (98: Address already in use)
nginx: [emerg] bind() to [::]:80 failed (98: Address already in use)
nginx: [emerg] bind() to [::]:80 failed (98: Address already in use)
nginx: [emerg] bind() to [::]:80 failed (98: Address already in use)
nginx: [emerg] still could not bind()

Any suggestions would be greatly appreciated. Thanks.

I was able to figure out my issue from 'vi /var/log/nginx/error.log'

(using ubuntu 12.10)

vim ~/.bashrc # add rbenv to the top
vim: command not found

So I used

sudo apt-get install vim

I have a question on redeployment. Say if I have updated my application? Must I redeploy the entire application or can I use a pull request?

Does "cap deploy" flush the db every time I redeploy?

Bitbucket hosts free private mercurial (hg) and git repositories and the capfile can just be changed to bitbuckets ssh link and it works like a charm. =)

As of time of posting, 7th April 2012 ,tutorial still works fantastically on Ubuntu 12.10.

A few updates:

Delete the line in /etc/nginx/sites-enabled/default if you get the error bind() to [::]:80 failed (98: Address already in use)

listen [::]:80 default_server

If using 12.10 or 12.04 use

rbenv bootstrap-ubuntu-12-04

Postgres can still be installed using the method noted in the tutorial, but it's being depreciated in lieu of:

https://wiki.postgresql.org/wiki/Apt

Thank you for this screencast, Ryan!

Hello

When i try to deploy VPS after running service nginx start command i got following error:

root@li356-76:~# service nginx start
 * Starting nginx nginx                                                                   nginx: [emerg] bind() to [::]:80 failed (98: Address already in use)
nginx: [emerg] bind() to [::]:80 failed (98: Address already in use)
nginx: [emerg] bind() to [::]:80 failed (98: Address already in use)
nginx: [emerg] bind() to [::]:80 failed (98: Address already in use)
nginx: [emerg] bind() to [::]:80 failed (98: Address already in use)
nginx: [emerg] still could not bind()

What should i do to fix this problem?

I am trying to deploy my rails app to Linode VPS but after i finished all steps when i did

xxx@xxxx-desktop:~/myrailsapp$ cap deploy:cold

i got following error:

failed: "sh -c 'cd /home/deployer/apps/myrailsapp/releases/20130422101521 && bundle install --gemfile /home/deployer/apps/myrailsapp/releases/20130422101521/Gemfile --path /home/deployer/apps/myrailsapp/shared/bundle --deployment --quiet --without development test'" on xxx.xxx.xxx.xxx

full log is below:

 xxx@xxx-desktop:~/myrailsapp$ cap deploy:cold
      * 2013-04-22 15:44:59 executing `deploy:cold'
      * 2013-04-22 15:44:59 executing `deploy:update'
     ** transaction: start
      * 2013-04-22 15:44:59 executing `deploy:update_code'
        updating the cached checkout on all servers
        executing locally: "git ls-remote git@bitbucket.org:myrailsapp/myrailsapp.git master"
        command finished in 5385ms
      * executing "if [ -d /home/deployer/apps/myrailsapp/shared/cached-copy ]; then cd /home/deployer/apps/myrailsapp/shared/cached-copy && git fetch -q origin && git fetch --tags -q origin && git reset -q --hard 88641306c92790d13f95beb7a2826a072ab72df4 && git clean -q -d -x -f; else git clone -q git@bitbucket.org:myrailsapp/myrailsapp.git maste /home/deployer/apps/myrailsapp/shared/cached-copy && cd /home/deployer/apps/myrailsapp/shared/cached-copy && git checkout -q -b deploy 88641306c92790d13f95beb7a2826a072ab72df4; fi"
        servers: ["xxx.xxx.xxx.xxx"]
    Password: 
        [xxx.xxx.xxx.xxx] executing command
        command finished in 5786ms
        copying the cached version to /home/deployer/apps/myrailsapp/releases/20130422101521
      * executing "cp -RPp /home/deployer/apps/myrailsapp/shared/cached-copy /home/deployer/apps/myrailsapp/releases/20130422101521 && (echo 88641306c92790d13f95beb7a2826a072ab72df4 > /home/deployer/apps/myrailsapp/releases/20130422101521/REVISION)"
        servers: ["xxx.xxx.xxx.xxx"]
        [xxx.xxx.xxx.xxx] executing command
        command finished in 951ms
      * 2013-04-22 15:45:22 executing `deploy:finalize_update'
        triggering before callbacks for `deploy:finalize_update'
      * 2013-04-22 15:45:22 executing `deploy:assets:symlink'
      * executing "rm -rf /home/deployer/apps/myrailsapp/releases/20130422101521/public/assets && mkdir -p /home/deployer/apps/myrailsapp/releases/20130422101521/public && mkdir -p /home/deployer/apps/myrailsapp/shared/assets && ln -s /home/deployer/apps/myrailsapp/shared/assets /home/deployer/apps/myrailsapp/releases/20130422101521/public/assets"
        servers: ["xxx.xxx.xxx.xxx"]
        [xxx.xxx.xxx.xxx] executing command
        command finished in 993ms
      * 2013-04-22 15:45:23 executing `bundle:install'
      * executing "cd /home/deployer/apps/myrailsapp/releases/20130422101521 && bundle install --gemfile /home/deployer/apps/myrailsapp/releases/20130422101521/Gemfile --path /home/deployer/apps/myrailsapp/shared/bundle --deployment --quiet --without development test"
        servers: ["xxx.xxx.xxx.xxx"]
        [xxx.xxx.xxx.xxx] executing command
     ** [out :: xxx.xxx.xxx.xxx] ERROR: Gem bundler is not installed, run `gem install bundler` first.
        command finished in 974ms
    *** [deploy:update_code] rolling back
      * executing "rm -rf /home/deployer/apps/myrailsapp/releases/20130422101521; true"
        servers: ["xxx.xxx.xxx.xxx"]
        [xxx.xxx.xxx.xxx] executing command
        command finished in 971ms
    failed: "sh -c 'cd /home/deployer/apps/myrailsapp/releases/20130422101521 && bundle install --gemfile /home/deployer/apps/myrailsapp/releases/20130422101521/Gemfile --path /home/deployer/apps/myrailsapp/shared/bundle --deployment --quiet --without development test'" on xxx.xxx.xxx.xxx

I tried all the ways and update following in deploy.rb
require "bundler/capistrano"

but problem still unresolved,Anyone can help me?

Problem got fixed using add this line in top of ~/.bashrc on remote server.

if [ -d $HOME/.rvm ]; then
  export PATH="$HOME/.rvm/bin:$PATH"
  eval "$(rvm init -)"
fi

Assuming that i am using rvm in my remote as well as in my local computer.

Thanks Ryan

For Ubuntu 12.04 LTS and later, create the deployer user in the sudo group instead of the admin group (the sudo group has replaced the admin group, source).