During `vagrant provision`, Ansible fails when username & site domain are too long

[jdsteinbach]

Problem:

When the OS X username & site domain are too long, they cause an ansible error during vagrant provision:

fatal: [local.innovativemedianetwork.com] => SSH Error: unix_listener: "/Users/jamessteinbach/.ansible/cp/ansible-ssh-local.innovativemedianetwork.com-22-vagrant.tRI3QuUSacSXgiA6" too long for Unix domain socket
    while connecting to 192.168.161.53:22
It is sometimes useful to re-run the command using -vvvv, which prints SSH debug output to help diagnose the issue.

Solution:

Add the following to ansible.cfg:

[ssh_connection]
control_path = %(directory)s/%%h-%%p-%%r

Ansible issue: https://github.com/ansible/ansible/issues/11536

[jdsteinbach]

Also, if this file is written automatically by yo evolve wordpress then the Yeoman generator will need the same update.

[EvanK]

We already have a FAQ entry for this (see 6ef7b92). I suppose we could use the longer %h-%p-%r pattern that's widely supported across ssh versions, and then update the faq to suggest %C when the issue persists (and the user's ssh version supports it)

[EvanK]

This is fixed, as of 1.4.1