How to make Fabric continue running the next command after getting the exit status: 1?

Question

I'm going to install check_mk plugin by writing a simple fabfile like this:

from fabric.api import env, run, roles, execute, parallel

env.roledefs = {
    'monitoring': ['192.168.3.118'],
    'mk-agent': ['192.168.3.230', '192.168.3.231', '192.168.3.232']
}

@roles('monitoring')
def mk():
    run('[ -f check_mk-1.1.12p7.tar.gz ] || wget http://mathias-kettner.de/download/check_mk-1.1.12p7.tar.gz')
    run('[ -d check_mk-1.1.12p7 ] || tar zxvf check_mk-1.1.12p7.tar.gz')
    run('cd check_mk-1.1.12p7 && sudo ./setup.sh')

@parallel    
@roles('mk-agent')
def mk_agent():
    run('[ `rpm -qa | grep -c xinetd` -eq 0 ] && sudo yum -y install xinetd.x86_64')
    run('sudo rpm -ivh http://mathias-kettner.de/download/check_mk-agent-1.2.0b2-1.noarch.rpm') 

def check_mk():
    execute(mk)
    execute(mk_agent)

But, as you can guess, if the xinetd package is already installed, Fabric will be stopped with below errors:

Fatal error: run() received nonzero return code 1 while executing!

Requested: [ `rpm -qa | grep -c xinetd` -eq 0 ] && sudo yum -y install xinetd.x86_64
Executed: /bin/bash -l -c "[ \`rpm -qa | grep -c xinetd\` -eq 0 ] && sudo yum -y install xinetd.x86_64"

Aborting.

Is there any solution in this situation?

Answer 1

since stackoverflow doesn't let me upvote Morgan's answer without more rep, I'll contribute more detail from http://docs.fabfile.org/en/1.4.1/api/core/context_managers.html#fabric.context_managers.settings

Outside the 'with settings' in the code below, behaviour will return to normal :

def my_task():
    with settings(
        hide('warnings', 'running', 'stdout', 'stderr'),
        warn_only=True
    ):
        if run('ls /etc/lsb-release'):
            return 'Ubuntu'
        elif run('ls /etc/redhat-release'):
            return 'RedHat'

This is desirable since you can essentially 'catch' what would've been an error in one section without it being fatal, but leave errors fatal elsewhere.

Answer 2

Perhaps in 2020 it will be useful. In Fabric 2.5, you just need to add the warn=True to the command to avoid interruption.

For example: connection.run('test -f /path/to/file && tail /path/to/file, warn=True)

Answer 3

You simply need to add "env.warn_only = True" to the def mk_agent(): task.