/tmp directory does not exist (bash.exe) [Windows Guest & Mac Host]

[Geoff Low]

Hi all

First off, great tool.  Very much enjoying using it.

I have created a Windows 2008 R2 box using veewee and am trying to use this in vagrant.  Vagrant successfully creates the vm, and starts it up - but then it fails when trying to create /vagrant.  I'm assuming it's because of the trace shown here:

DEBUG ssh: stderr: "tty" option detected in CYGWIN environment variable.

CYGWIN=tty is no longer supported.  Please remove it from your

CYGWIN environment variable and use a terminal emulator like mintty,

xterm, or rxvt.

DEBUG ssh: Exit status: 0

DEBUG ssh: stderr: bash.exe: warning: could not find /tmp, please create!

DEBUG ssh: stdout: /vagrant

DEBUG linux: Shell expanded guest path: /vagrant

DEBUG ssh: Re-using SSH connection.

 INFO ssh: Execute: mkdir -p /vagrant (sudo=true)

DEBUG ssh: Exit status: 2

ERROR warden: Error occurred: The following SSH command responded with a non-zero exit status.

Vagrant assumes that this means the command failed!

If I ssh into the image (vagrant ssh app) I can see the /tmp directory.

Is my issue here with a) the veewee box I created or b) with the way I've configured my Vagrantfile

(for reference)

  config.vm.define :app do |app_config|

    app_config.vm.box = "winbase"

    app_config.vm.host_name = "app.mock"

    app_config.vm.provision :chef_solo do |chef|

      # use mdsol recipes (

      chef.provisioning_path = "/tmp/vagrant-chef"

      chef.roles_path = [File.join(__FILE__, "../chef-solo/roles")]

      chef.cookbooks_path = [File.join(__FILE__, "../chef-solo/cookbooks")]

      chef.add_role :app

    end

    app_config.vm.customize do |vm|

      vm.name = "com.mock.app"

    end

  end

Thanks in advance...

Geoff

[Geoff Low]

Ok, I may have missed something using the full DEBUG, with INFO I get the following:

[default] Mounting shared folders...

 INFO interface: info: -- v-root: /vagrant

[default] -- v-root: /vagrant

 INFO subprocess: Starting process: ["VBoxManage", "showvminfo", "772c797c-41d3-4455-9972-26274256f140", "--machinereadable"]

 INFO ssh: Execute: cat /etc/debian_version (sudo=false)

 INFO ssh: Execute: cat /etc/gentoo-release (sudo=false)

 INFO ssh: Execute: grep 'Fedora release 16' /etc/redhat-release (sudo=false)

 INFO ssh: Execute: cat /etc/redhat-release (sudo=false)

 INFO ssh: Execute: cat /etc/SuSE-release (sudo=false)

 INFO ssh: Execute: cat /etc/arch-release (sudo=false)

 INFO vm: Loading guest: linux

 INFO ssh: Execute: printf /vagrant (sudo=false)

 INFO ssh: Execute: mkdir -p /vagrant (sudo=true)

ERROR warden: Error occurred: The following SSH command responded with a non-zero exit status.

Vagrant assumes that this means the command failed!

mkdir -p /vagrant

 INFO warden: Calling recover: #<Vagrant::Action::VM::Import:0x007ff65237e808>

 INFO subprocess: Starting process: ["VBoxManage", "showvminfo", "772c797c-41d3-4455-9972-26274256f140", "--machinereadable"]

ERROR warden: Error occurred: The following SSH command responded with a non-zero exit status.

Vagrant assumes that this means the command failed!

And then I check the ssh as being up using;

Kauri-Geoff-Low:simulation glow$ VAGRANT_LOG=INFO vagrant ssh

 INFO global: Vagrant version: 1.0.1

 INFO vagrant: `vagrant` invoked: ["ssh"]

 INFO environment: Environment initialized (#<Vagrant::Environment:0x007f8e6315cba8>)

 INFO environment:   - cwd: /Users/glow/vagrant/simulation

 INFO environment: Home path: /Users/glow/.vagrant.d

 INFO plugin: Searching and loading any available plugins...

 INFO plugin: Loading plugin: veewee (0.2.3)

 INFO plugin: Loaded 1 plugins.

 INFO environment: Loading configuration...

 INFO provisioner: Provisioner class: Vagrant::Provisioners::ChefSolo

 INFO cli: CLI: [] "ssh" []

 INFO datastore: Created: /Users/glow/vagrant/simulation/.vagrant

 INFO virtualbox_base: VBoxManage path: VBoxManage

 INFO subprocess: Starting process: ["VBoxManage", "--version"]

 INFO virtualbox: Using VirtualBox driver: Vagrant::Driver::VirtualBox_4_1

 INFO virtualbox_base: VBoxManage path: VBoxManage

 INFO subprocess: Starting process: ["VBoxManage", "showvminfo", "772c797c-41d3-4455-9972-26274256f140"]

 INFO vm: Loading guest: linux

 INFO subprocess: Starting process: ["VBoxManage", "showvminfo", "772c797c-41d3-4455-9972-26274256f140", "--machinereadable"]

 INFO subprocess: Starting process: ["VBoxManage", "showvminfo", "772c797c-41d3-4455-9972-26274256f140", "--machinereadable"]

 INFO subprocess: Starting process: ["VBoxManage", "showvminfo", "772c797c-41d3-4455-9972-26274256f140", "--machinereadable"]

 INFO subprocess: Starting process: ["VBoxManage", "showvminfo", "772c797c-41d3-4455-9972-26274256f140", "--machinereadable"]

 INFO ssh: Invoking SSH: ["-p", "2222", "-o", "UserKnownHostsFile=/dev/null", "-o", "StrictHostKeyChecking=no", "-o", "IdentitiesOnly=yes", "-o", "LogLevel=ERROR", "-i", "/Users/glow/.vagrant.d/insecure_private_key", "vag...@127.0.0.1"]

Last login: Thu Mar  1 14:41:13 2012 from 10.0.2.2

"tty" option detected in CYGWIN environment variable.

CYGWIN=tty is no longer supported.  Please remove it from your

CYGWIN environment variable and use a terminal emulator like mintty,

xterm, or rxvt.

vagrant@vagrant-2008R2 ~

Anyone got any ideas?

Thanks

Geoff

[Mitchell Hashimoto]

I would try doing what this message says ^

Mitchell

[Mitchell Hashimoto]

On Wed, Jun 6, 2012 at 8:56 AM, Alex Rodionov <p0d...@gmail.com> wrote:
> Hello Mitchell,
>
> Can you please explain what you mean "doing what message says". If it's
> about CYGWIN=tty, then I fixed this problem but still get the same error.
>

In the above case I meant manually running "mkdir -p /vagrant" and
seeing what was going wrong.

[Alex Rodionov]

After "vagrant up" finishes, I can connect to th box with "vagrant ssh" without any errors - "mkdir -p /vagrant" works ok and returns zero exit code.