Cronicle

It is a tool for execute script, and define cron on remote hosts.

Installation

Add this line to your application's Gemfile:

gem 'cronicle'

And then execute:

$ bundle

Or install it yourself as:

$ gem install cronicle

Usage

Commands:
  cronicle apply           # Apply cron jobs to remote hosts
  cronicle cleanup         # Clean up cron jobs on remote hosts
  cronicle exec JOB_NAME   # Execute a job on remote hosts
  cronicle help [COMMAND]  # Describe available commands or one specific command

Options:
  -f, [--file=FILE]                    # Job definition file
                                       # Default: Jobfile
  -h, [--hosts=HOSTS]                  # Hosts definition file
  -r, [--target-roles=one two three]   # Target host role list
  -p, [--sudo-password=SUDO-PASSWORD]  # Sudo password
      [--ssh-user=SSH-USER]            # SSH login user
      [--ask-pass], [--no-ask-pass]    # Ask sudo password
      [--dry-run], [--no-dry-run]      # Do not actually change
  -c, [--ssh-config=SSH-CONFIG]        # OpenSSH configuration file
                                       # Default: ssh_config
      [--ssh-options=SSH-OPTIONS]      # SSH options (JSON)
      [--connection-timeout=N]         # SSH connection timeout
      [--concurrency=N]                # SSH concurrency
                                       # Default: 10
      [--var-dir=VAR-DIR]              # Cronicle var dir path
                                       # Default: /var/lib/cronicle
  -v, [--verbose], [--no-verbose]      # Verbose mode
      [--debug], [--no-debug]          # Debug mode
      [--color], [--no-color]          # Colorize log
                                       # Default: true

Quick Start

$ cat Jobfile
on servers: :your_hostname do
  job :my_job, user: "ec2-user", schedule: "* * * * *" do
    puts "hello"
  end
end

$ cronicle exec my_job
my_job on your_hostname/ec2-user> Execute job
my_job on your_hostname/ec2-user> hello

$ cronicle apply --dry-run
my_job on your_hostname/ec2-user> Create job: schedule="* * * * *" (dry-run)

$ cronicle apply
my_job on your_hostname/ec2-user> Create job: schedule="* * * * *"

$ ssh your_hostname 'crontab -l'
* * * * * /var/lib/cronicle/libexec/ec2-user/my_job 2>&1 | logger -t cronicle/ec2-user/my_job

$ ssh your_hostname 'cat /var/lib/cronicle/libexec/ec2-user/my_job'
#!/usr/bin/env ruby
puts "hello"

Environment variables

• CRONICLE_SSH_USER
• CRONICLE_SSH_OPTIONS
• CRONICLE_SSH_CONFIG
• CRONICLE_SUDO_PASSWORD

Jobfile example

on servers: :your_hostname do
  job :my_job, user: "ec2-user", schedule: "* * * * *" do
    puts "hello"
  end

  job :my_job2, user: "ec2-user", schedule: "* * * * *", content: <<-EOS
    #!/bin/sh
    echo hello
  EOS
end

Use bundler

on servers: :your_hostname do
  job :my_job, user: "ec2-user", schedule: "* * * * *", bundle: 'ruby-mysql' do
    require 'mysql'
    my = Mysql.connect('hostname', 'username', 'password', 'dbname')
    my.query("select col1, col2 from tblname").each do |col1, col2|
      p col1, col2
    end
  end
end

Use variables

on servers: :your_hostname do
  job :my_job, user: "ec2-user", schedule: "* * * * *", locals: {foo: 'bar'} do
    puts foo
  end

  job :my_job2, user: :root, schedule: "* * * * *", extra: "FOO=bar\nZOO=baz", content <<-SH
    #!/bin/sh
    echo $FOO
    echo $ZOO
  SH
end

Hosts definition file

Any of the following formats:

server1,server2,...

server1
server2
...

{
  "servers": {
    "server1": ["web", "app"],
    "server2": ["db"]
  }
}

{
  "roles": {
    "web": ["server1"],
    "app": ["server1"],
    "db": ["server2"]
  }
}

Hosts definition file is not required.

If you pass --hosts option, can be defined as follows in Jobfile:

on servers: /any_host/ do # use regexp
  job :foo, user: "ec2-user", schedule: "* * * * *" do
    3.times do
      puts "hello"
    end
  end
end

on roles: "web" do # use role
  job :my_job, user: "ec2-user", schedule: "* * * * *" do
    puts "hello"
  end
end