Capistrano Google Group

use_sudo option in multistage

btri...@gmail.com (Bryan Richardson) — Sat, 12 Dec 2009 23:02:56 UT

Hi All,

Does :use_sudo work in multistage? I have a production stage that runs
on a server requiring sudo, but a staging stage that runs on a
different server not requiring sudo. I tried "set :use_sudo, false" in
my config/deploy/staging.rb file, but it looks like Capistrano is
still trying to execute some commands on the staging server using

uninitialized constant DL::Importable on Windows Vista x64 after upgrading to Ruby 1.9.1

darthd...@gmail.com (darth) — Fri, 12 Dec 2009 17:48:59 UT

I've just upgraded to Ruby 1.9.1 and every time I do something with
Capistrano, like cap deploy, or even cap -v, I get this error
[link]

I tried reinstalling Capistrano gem, but it didn't help. I currently
have installed Capistrano 2.5.10 and ruby 1.9.1p243 (2009-07-16

Task execution based on :primary flag

justin.di...@gmail.com (jmadtech) — Fri, 12 Dec 2009 14:59:27 UT

Hello all,

I am currently trying to write a set of task commands that run based
on whether or not a particular server within the role attached to the
task is the primary server.

So for example, I have the following:

task :push_my_configs, :on_error => :continue do
run "my command 1", :roles => :app, :only => { :primary => true }

symlink doesn't work (403 forbidden)

christoph.jasin...@googlemail.com (christoph.jasinski@googlemail.com) — Fri, 12 Dec 2009 15:44:40 UT

You might hear it for the millionth time but I just spend some hours
on that issue. Search the internet, searched the capistrano mail etc.

My problem from apache's log:
Symbolic link not allowed or link target not accessible: /var/deploy/
deployments/myapp.com/current, referer: [link]

Executing a task once locally, but copying to all remote hosts

robin-li...@robinbowes.com (Robin Bowes) — Wed, 12 Dec 2009 13:24:48 UT

Hi,

I'm using railsless deploy, and have this task to generate a config file
locally, and copy it to all hosts:

desc <<-DESC
Create config file & copy to release path
DESC
task :create_copy_config_file, :roles => [:web] do
print " Creating config file.\n"
file = 'files/counter.php'

set :user remote_user_name not affecting SSH login

ja...@trallenya.net (Jamie) — Tue, 12 Dec 2009 04:51:56 UT

I'm attempting my first deployment using Capistrano. The recipe
includes

set :user, "remote_user"

When I attempt "cap deploy" I'm prompted for

local_u...@myserver.com's password:

Of course the local_user has no account on the remote server so I
can't get past this. The weird thing is I was successfully able to

Issue with specifying a machine as "spare"

bodarot...@gmail.com (Mike Papper) — Mon, 11 Nov 2009 21:43:30 UT

Hi all,

I wish to specify an app server as "spare" - to be out of the load
balancer rotation. I specified as so:

role :app, '192.168.10.99', :spare => true # machines out of load
balancer rotation

The task is like this:

namespace :loadbalancer do
desc 'Remove "spare" machines from load balancer rotation by

'stream' and multiple logs per server

pharri...@gmail.com (Pharrisee) — Mon, 11 Nov 2009 15:18:13 UT

I've got a situation where I have 2 physical servers, each of which
have 2 nodes of an IBM WebSphere Portal cluster on it.

I'm using capistrano for the deployment and staging (very
successfully), and am now trying to use cap for tailing the logs on
those servers, but because each node puts it's logs into a separate

How do I allow my teammate to deploy?

alexanderk...@gmail.com (alexanderkahn) — Mon, 11 Nov 2009 13:52:37 UT

Hi,

I'm working on a Rails project that I deploy using the basic
remote_cache strategy. I push to a bare git repo in my home folder on
the server, and then `cap deploy` checks out the latest code from the
repo, links it, and restarts Passenger.

This works well, but I have a new requirement: I want my designer to

Capistrano DSL for highly customized deployment

camp...@aol.com (campesr) — Sun, 11 Nov 2009 16:50:54 UT

Hi everybody,

I'm curious whether it is possible to use Capistrano's DSL and general
foundation (roles, tasks, #run, #sudo, #put etc.) to build a custom
Rakefile/Capfile?

No automagic deploy:migrate tasks (especially for non-rails
projects!), no automagic 'mkdir -p (..)/release; ln -nfs (..)/

Deploying Error: `split' called for 440:Fixnum

wwwjs...@gmail.com (jsoo) — Thu, 11 Nov 2009 19:13:09 UT

Hey

I'm pretty sure I've overlooked something obvious but I can't seem to
figure this out.

I'm trying to deploy and I receive the following error (x's inserted)
------------------------------ ------------------------------ ------------------------------ -----
* executing `production'
* executing `deploy'

Re: [capistrano] Abridged summary of capistrano@googlegroups.com - 1 Message in 1 Topic

jr...@appriss.com (John Rice) — Thu, 11 Nov 2009 14:55:09 UT

I am out of the office and will return on November 30th.

Today's Topic Summary
Group: [link]
• Abridged summary of capistrano@googlegroups.com - 4 Messages in 2 Topics [1 Update]
Topic: Abridged summary of capistrano@googlegroups.com - 4 Messages in 2 Topics

Re: [capistrano] Abridged summary of capistrano@googlegroups.com - 4 Messages in 2 Topics

jr...@appriss.com (John Rice) — Wed, 11 Nov 2009 17:21:15 UT

I am out of the office and will return on November 30th.

Today's Topic Summary
Group: [link]
• Specifying alternate prerequisites in rake [2 Updates]
• Abridged summary of capistrano@googlegroups.com - 3 Messages in 2 Topics [2 Updates]

Re: [capistrano] Abridged summary of capistrano@googlegroups.com - 3 Messages in 2 Topics

jr...@appriss.com (John Rice) — Tue, 11 Nov 2009 14:59:18 UT

I am out of the office and will return on November 30th.

Today's Topic Summary
Group: [link]
• Specifying alternate prerequisites in rake [2 Updates]
• Question about running a piece of task only once [1 Update]
Topic: Specifying alternate prerequisites in rake

Specifying alternate prerequisites in rake

myli...@polite.se (Oivvio Polite) — Mon, 11 Nov 2009 20:40:20 UT

Is there anyway of specifying alternate prerequisites for a task, ie at
least one of the prerequisites must be run before the task itself is
run.

The way I'm doing it now is by using a "proxy task" that picks one of
the prerequisite tasks based on a global settings variable.

It looks something like this:

Question about running a piece of task only once

joseph.e.mcdon...@gmail.com (Joe McDonagh) — Mon, 11 Nov 2009 19:51:48 UT

Say i have a task on like 6 servers (deploying the web code) but at the
end i want only one email to go out that says "Website updated to
revision whatever" how do i make a run or local statement only run once
and not once per whole task and not once per host in the role?

I'm thinking an after deploy hook that runs a task that has just one

capfile syntax help

robin-li...@robinbowes.com (Robin Bowes) — Thu, 11 Nov 2009 16:28:08 UT

Hi,

I've got the following block in my capfile:

namespace :remote do

desc <<-DESC
Create directory required by copy_remote_dir.
DESC
task :create_copy_remote_dir, :roles => :web do
print " creating #{copy_remote_dir}.\n"
run "mkdir -p #{copy_remote_dir}"
end

end

installing mongrel

saud.prak...@gmail.com (prakash saud) — Wed, 11 Nov 2009 07:06:21 UT

hello, everybody
Actually i am new to capistrano. I have installed ruby 1.8.6, Rails
2.3.4 and my OS is ubuntu . I have made a rails application. I went
through the getting started and everything went fine upto check
dependency. when i tried to install mongrel, it gives message
"Successfully installed mongrel-1.1.5" but when i tried mongrel -v it

Webistrano stuck on update_code only when trying to deploy to multiple hosts

burga...@gmail.com (burg) — Sun, 11 Nov 2009 17:22:32 UT

I'm trying to set up Webistrano + Capistrano to deploy to multiple
servers:
When trying to deploy to each server separately, everything works
great. However when I select multiple hosts to deploy it just gets
stuck in the update_code stage (if I run 'restart' for example on
multiple hosts - it works just fine).

More than one project on a server

richard.iba...@gmail.com (ribarra) — Sun, 11 Nov 2009 00:25:08 UT

Hi guys!
I would like to deploy more than one project on a server. For example,
the project A is running on foo.bar.com and I need foo.bar.com:1234
running other project (B). Can I deploy something like that using
capistrano? How the deploy.rb must be configured?

Thanks in advance!

-- Richard

commandline options

bea...@benle.de (Benjamin Fritsch) — Fri, 11 Nov 2009 00:30:53 UT

hey guys,

first of all, thanks for capistrano!

i use capistrano primary for server management and not for deploying
my code.
i want to change the behavior of how the servers are selected. i want
to add an commandline options, with this option a server must apply to
_all_ roles that are given.

how to specify environment variables and chdir

maurizio.demag...@gmail.com (maurizio de magnis) — Wed, 11 Nov 2009 12:02:24 UT

Hello everybody, this is my first post here :)

My environment:
$ ruby --version
ruby 1.8.7 (2008-08-11 patchlevel 72) [universal-darwin10.0]
$ cap --version
Capistrano v2.5.9
I'm on macosx 10.6.1.

I'd like to avoid using the "set_dir" and "set_path" trick for each
command.
I needed "set_path" because on :my_dest_host, "hg" is not in the usual

run command with resiliency to failure

spike....@gmail.com (spike grobstein) — Wed, 11 Nov 2009 01:27:31 UT

I'm working on some cap tasks where it's not that big of a deal if
they fail, but I'd like to be able to let the user know that they did.

What I'm trying to do is create a new user on a bunch of servers, but
if it fails on one, I don't want the entire capistrano session to
fail.

is there any way to do this? right now, if the user exists on one

NOOB question on first deploy attempt

j...@whidbey.com (blueHandTalking) — Tue, 11 Nov 2009 19:42:40 UT

Capistrano, git, ruby, rails: Just getting started on all of them. I
am attempting
to deploy from my remote git repository, on the master branch in
Aptana Cloud.

My remote depository is in place, now I am trying:

cap staging deploy

I get the error message:

C:/Ruby/lib/ruby/gems/1.8/gems /capistrano-2.5.5/lib/capistra no/recipes/

check Capistrano on local-host

saud.prak...@gmail.com (prakash saud) — Tue, 11 Nov 2009 09:56:10 UT

Actually, I am new to Capistrano. Currently i am using ruby version
1.8.6 and Capistrano version 2.5.9. I tried to test a file on local-
host but i am getting error message like " connection failed for:
[link] (Socket Error: getaddrinfo: Name or service not
known)". I made a cap file in home directory where wrote the code

No releases.last crashes first deploy

dusan.malia...@gmail.com (skrat) — Thu, 11 Nov 2009 15:20:37 UT

Hi all,

I'm running into problem where I am unable to do first time deploy.
Capistrano crashes with following trace ([link]).
Looking at the source file, it's looking for last release, which
obviously isn't there.

I found some other posts on this, but none with working solution. This

Any advice using capistrano and git to deploy sites with heavy flash assets?

calb...@neomantic.com (Chad A.) — Wed, 11 Nov 2009 18:31:01 UT

Hi,

We are currently experimenting with using both git and capistrano to
deploy our websites. Here's the situation. Both our flash developers
and backend developers use the same git repository (don't really
believe in submodules). Now only certain flash pieces need to be
deployed - the rest are binary assets used to build the flash

creating tasks for multiple environments (and gateways)

jjwilcox...@gmail.com (Joe Wilcoxson) — Tue, 11 Nov 2009 21:31:12 UT

Hi, I'm looking for a way to manage systems in multiple
datacenters/environments, each with their own gateway.
gateway 1 -> app1 app2 app3
gateway 2 -> app1 app2 app3
I get stuck when trying to set "gateway" within the same capfile after its
already been set and was looking for a recommendation on how to approach

Rails app in a subfolder of a git clone - possible ?

thibaut.barr...@gmail.com (Thibaut Barrère) — Mon, 11 Nov 2009 15:56:45 UT

Hi,

is there a clean way to deploy a rails app that would not sit in the
root of the git repository, but in a sub-folder instead ?

any hint will be most welcome,

Thibaut
--
[link]

New feature deployment

amugn...@gmail.com (xymbol) — Sat, 10 Oct 2009 23:26:21 UT

Hi,

On previous projects, I've used separate staging and production stages
which would both check out source code from a same master branch.

Now, for a project where static content gets added frequently through
the repository, I'm feeling the need for either:a separate staging
branch or, an experimental stage + branch to deploy long running work

SUDO Problem

partydr...@gmail.com (partydrone) — Fri, 10 Oct 2009 20:32:42 UT

When trying to deploy to a new staging server, I get the following
error:

sudo: no tty present and no askpass program specified

Anyone know what's up with this and how do I fix it?

Thanks!

tasks that refuse to run after they fail twice

patc...@gmail.com (patcoll) — Wed, 10 Oct 2009 18:27:21 UT

I have two custom "restart" tasks that restart apache and nginx,
respectively. I put :on_error => :continue on both.

After the Apache "restart" throws an error twice in the same script
(my fault), any subsequent calls to that task seem to get ignored...

Has anyone else experienced this? I can see how this may be a feature

Needed a single revision: what does it mean?

collimarc...@gmail.com (collimarco) — Tue, 10 Oct 2009 19:38:48 UT

Capistrano and Git worked properly till I moved my site to another
server. My first attempt to deploy went wrong because of an SSH-known-
hosts related problem.

Even if I managed to solve this problem, another one raised and I
can't deploy any more:

[link]

My git repository seems intact: why do I get "Needed a single

Is there a way to have capistrano ignore timed-out machines?

tomri...@gmail.com (Tom Riley) — Tue, 10 Oct 2009 18:25:38 UT

Hi there,

I'm trying to use Capistrano to run periodic tasks over a bunch of
machines and I don't really care if a machine goes down (i set quite a
short timeout in ssh_options) but by default, capistrano won't attempt
to run any commands on any machine if one is inaccessible. Does anyone
know of a way to easily change this behaviour so machines that aren't

cap deploy failed, why is capistrano expecting to find database.yml in my deploy_to folder ?

octimi...@gmail.com (octimizer) — Mon, 10 Oct 2009 23:22:59 UT

Hi everyone,

I'm new to Rails and Capistrano,
I'm trying to deploy my first app (just a hello world app, no database
work yet), from an iMac to a Linux server, using git for SCM.

cap deploy:setup went fine
cap deploy:check went fine
cap deploy or cap deploy:cold or cap:deploy:migrations they all fail

using multistage: cap staging deploy:migrations doin' stuff on the production server

robert.h...@gmail.com (Robert Head) — Thu, 10 Oct 2009 19:29:13 UT

I'm having trouble getting multistage working. Thank you so much in
advance for any insight.

For some reason, asking for a staging deployment tries to deploy on
both the production server and the staging server.
The deploy output includes (see pastie below):
servers: ["myproductionserver.com", "stage.myapp.com"]

SSH Timeouts for Long Running Migrations

bod...@gmail.com (Mike) — Thu, 10 Oct 2009 00:44:20 UT

We have a long running migration. After the migration finishes on one
box, when Cap starts the next task, the other ssh connections to the
other servers timeout and the deploy fails.

I have read a number of items on the net and from this group regarding
this issue. However, there seems to be nothing that resolves the

SVN and Capistrano for designers?

andym...@gmail.com (Andy) — Wed, 10 Oct 2009 15:57:31 UT

I often work with designers and HTML/CSS coders, and MANY MANY times,
I just need them to make quick changes to the CSS, images, or HTML.

They usually have to email me the files, I overwrite my local
versions, check them in w/ SVN, and then cap deploy them.

Is there some tool out there that would make it easy for my designers

Patch: use :current_release in --dry-run

maciekp.li...@gmail.com (Maciej Pasternacki) — Mon, 10 Oct 2009 10:33:39 UT

Hello,

For some time I noticed some commands failing with --dry-run. I
managed to trace it back to usage of :current_release, and to patch it
with a default that is used when network is not available.

Patch to deploy.rb, made against capistrano-2.5.9 gem, is available at
[link]

Capistrano has ceased to start and stop mongrels

danfdonald...@gmail.com (omnivore) — Sat, 10 Oct 2009 21:22:35 UT

development machine: Capistrano v2.5.9, Rails 2.2.2, OS X Snow Leopard
deployments machine: Capistrano v1.4.1, Rails 1.2.3, Solaris (Joyent)
Accelerator

I changed development machines, had a couple of rough spots that
caused me to upgrade a substantial part of my RoR setup. Now, running
cap deploy, or variants that should (and always have) start, stop or

NOOB Error in git.rb when doing a cap deploy

j...@whidbey.com (blueHandTalking) — Wed, 10 Oct 2009 22:12:31 UT

I am new to Capistrano, have just set up a repository I wish to deploy
from.

The following is error I am getting:

capistrano/recipes/deploy/scm/ git.rb:215:in `query_revision':
private method `split' called for nil:NilClass (NoMethodError)

The code leading up to that error is :

def query_revision(revision)

Rake migrate challenge, MediaTemple

brianma...@gmail.com (Brian Maggi) — Tue, 10 Oct 2009 15:16:42 UT

We're trying to host a demonstration application on MediaTemple and
I'm about to pull my hair out trying to get it deployed. After finally
getting all the right dependencies removed / installed Capistrano
appears to be functioning but not deploying.

cap deploy:check comes back okay.
cap deploy:setup works too.

Multiple Checkouts?

pheg...@gmail.com (Kiran) — Mon, 10 Oct 2009 18:05:49 UT

Hi all,
I have a rails app that i have setup to deploy using capistrano but
i have another set of files that are in a different repository that
are neccessary to have the rails app function. They are JS/HTML/CSS
files that need to end up in the public directory or at least symlink
into the directory. I need to keep the files separate in the

Re: [capistrano] Error in git.rb:215: in `query_revision' when doing a deploy:cold

lee.hamb...@gmail.com (Lee Hambley) — Sun, 10 Oct 2009 19:06:02 UT

Please paste your deploy.rb so I can take a look at your configuration.

Easily limit deployment to only hosts in a certain role

nagarebo...@gmail.com (Kang) — Fri, 10 Oct 2009 22:06:57 UT

Pardon my newbieness but is there a way to easily limit deploy code to
only hosts in a certain role? I am thinking deploy:update_code to
default to app and migrations to run only on db.

Thanks

Kang

Best way to trim duplicate hosts in a given role

joseph.e.mcdon...@gmail.com (Joe McDonagh) — Fri, 10 Oct 2009 17:29:47 UT

Anybody think of a good way to trim duplicate hosts in a role?
I realize this probably never happens to most of you, but I fill in
dynamic roles via two DB's and there is a possibility of duplicating a
host. Part of this is an architectural problem that I will fix in due
time, however this would be nice to figure out.

Start JBoss as background process using capistrano (nohup & ampersand)

zahi...@gmail.com (zahidal) — Fri, 10 Oct 2009 13:25:27 UT

I am trying nohup and ampersand (&) but no luck.

If i use:

run "nohup /opt/jboss/jboss-eap-4.3/jboss -as/bin/run.sh > /tmp/
my_jboss.start.log", :pty => true

Then it is creating the my_jboss.start.log file and starting the
server but it doesn't return the capistrano console back to me. If I

Setting up symlinks for multiple stages

cschal...@gmail.com (cswebgrl) — Fri, 10 Oct 2009 14:17:47 UT

Hi,

I'm sure I'm missing it somewhere in the documentation, but I can't
seem to get the syntax in the correct place to make this work. I am
using multi stage and currently have 3 staging environments that I
deploy to. The staging and production servers have the same directory
structure, but a third temporary testing environment has a different

Setting symlinks in core

j...@whidbey.com (blueHandTalking) — Thu, 10 Oct 2009 19:47:26 UT

The below chunk was taken just about directly from a video Scott
Chacon
made on setting up Capistrano tasks in deploy.rb:

namespace :deploy do
task :restart do
run "touch #{current_path/tmp/restart.txt }"
end
task :symlink_shared do
run "ln -nfs #{shared_path}/config/database .yml #

Deploying from WinXP: Connecting with Putty, Pageant, and OpenSSH to a git repository

j...@whidbey.com (blueHandTalking) — Thu, 10 Oct 2009 05:49:17 UT

I have tried about everything to get this working, but still no luck.

I am using public and private keys I generated with puttygen,
uploaded, then converted on my Solaris
server before putting into authorized_keys.

I set Agent forwarding in Putty before I run:

cap staging deploy.

I get the following error: