Adding collection of misc tools and scripts

ansible/LICENSE.txt

@@ -0,0 +1,19 @@
+ Copyright (c) 2014 Go Daddy Operating Company, LLC
+
+Permission is hereby granted, free of charge, to any person obtaining a 
+copy of this software and associated documentation files (the "Software"), 
+to deal in the Software without restriction, including without limitation 
+the rights to use, copy, modify, merge, publish, distribute, sublicense, 
+and/or sell copies of the Software, and to permit persons to whom the 
+Software is furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in 
+all copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR 
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, 
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL 
+THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER 
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING 
+FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER 
+DEALINGS IN THE SOFTWARE.

ansible/README.md

@@ -0,0 +1,101 @@
+ansible-playbooks
+=================
+
+Go Daddy Ansible playbooks for managing OpenStack infrastructure.
+
+Also available publically at https://github.com/godaddy/openstack-ansible
+
+This assumes your baseline Ansible config is at `/etc/ansible`, and this repo is cloned
+to `/etc/ansible/playbooks`  (Specifically, the playbooks assume the path to the tasks directory is `/etc/ansible/playbooks/tasks`,
+so if you are cloning this repo somewhere else, you'll need to adjust that.)
+
+Patches/comments/complaints welcomed and encouraged!  Create an issue or PR here.
+
+Usage Details
+-------------
+
+Playbooks are "shebanged" with `#!/usr/bin/env ansible-playbook --forks 50`, so you can actually 
+just run them directly from the command line.
+
+We have the concept of "worlds", which correspond to dev, test, prod, etc., servers.  We use
+the world teminology to avoid confusion with the Puppet environment setting (which, for us, 
+corresponds to a branch in our [openstack-puppet](https://github.com/godaddy/openstack-puppet) repo.)  So when you see references to the world
+variable, that's what it is.  Right now only a couple playbooks utilize that, so for the most
+part you can probably use these without worrying about defining a world variable.
+
+puppet-run.yaml and r10k-deploy.yaml are fairly specific to our environment, and are probably
+mostly irrelevant unless you're also using our openstack-puppet repo for Puppet configuration.
+
+Basic usage for the other playbooks follows.
+
+### template-prestage.yaml
+
+This one is really cool, complements of [krislindgren](http://github.com/krislindgren).  It sets up a 
+BitTorrent swarm, seeded by the machine running Glance, to distribute a Glance image out to any number
+of nova-compute nodes very quickly.  So when we roll new gold images each month, we can use this to
+"push" them out to all compute nodes, and avoid the first provision penalty of waiting for the image
+to transfer the first time.
+
+Caveats:
+* Only works when the Glance backend is on a traditional (local or network-based) filesystem.  Almost
+certainly this does not work, and may not even make sense, for Swift-backed Glance.
+* Firewalls or other traffic filters need to allow the BitTorrent ports through, and among, the Glance
+server and all compute nodes.
+* Assumes Glance images are stored at `/var/lib/glance/images` on the Glance server.
+* There are some situations where this cannot be run multiple times, if the tracker or othe BitTorrent
+processes are still running.  So use caution, and YMMV.
+* This is done completely outside the scope of Glance and Nova.  There is no Keystone authentication or
+access controls.  You must have ssh and sudo access to all machines involved for this to work.
+
+Usage:
+
+    ./template-prestage.yaml -k -K -e "image_uuid=<uuid> image_sha1=<sha1> image_md5=<md5> tracker_host=<glance server> hosts_to_update=<compute host group>"
+
+* _image_uuid_: UUID of the image to prestage (from `nova image-list` or `glance image-list`)
+* _image_sha1_: SHA1 sum of the image_uuid (this can be gotten by running: `echo -n "<image_uuid>" | sha1sum | awk '{print $1}'`  on any Linux box)
+* _image_md5_: MD5 sum of the image file itsemf (this can be gotten by running: `md5sum /var/lib/glance/images/<image_uuid> | awk '{print $1}'` on the Glance server
+* _tracker_host_: This is the Glance server host that runs the tracker
+* _hosts_to_update_: This is the host group to place the image onto (a list of compute nodes)
+
+### copy-hiera-eyaml-keys.yaml
+
+Copies public and private keys for hiera-eyaml from the source/ansible client machine, to hosts, at
+`/etc/pki/tls/private/hiera-eyaml-{public,private}_key.pkcs.pem`
+
+    ./copy-hiera-eyaml-keys.yaml -k -K -e "srcdir=<srcdir> hosts=<host group>"
+
+* _srcdir_: Source directory on the ansible client machine where the hiera-eyaml-public_key.pkcs7.pem and hiera-eyaml-private_key.pkcs7.pem keys can be found
+
+### patching.yaml
+
+This is a simple playbook to run `yum -y update --skip-broken` on all machines.
+
+    ./patching.yaml -k -K -e "hosts=<host group>"
+
+This also runs the remove-old-kernels.yaml task first, which removes any kernel packages from the
+system which are not 1) the currently running kernel, nor, 2) the default boot kernel in GRUB.
+
+### updatepackages.yaml
+
+Similar to patching.yaml, but this one allows for a specification of exactly which package(s) to update.
+
+    ./updatepackages.yaml -k -K -e "puppet=<true|false> package=<package spec> hosts=<host group>"
+
+* _puppet_: If true, will also run the puppet-run.yaml task after updating the packages.  Default false.
+* _package spec_: Specification for what packages to update, wildcards are valid.  Default '*'
+
+### restartworld.yaml
+
+Restarts some (or all) openstack services on hosts.  Note that this uses the `tools/restartworld.sh` script
+from the [godaddy/openstack-puppet](https://github.com/godaddy/openstack-puppet) repo, so you may want to look at that before trying to use this playbook.
+
+This is also somewhat specific to our environment, as far as how we group services together (most of 
+them run on the "app" class of server.)  So results and usefulness may vary.
+
+    ./restartworld.yaml -k -K -e "class=<server class> hosts=<host group> service=<service class>"
+
+* _class_: Server class to define what services to restart.  Recognized options are app, network, and compute.  See the `restartworld.sh` script referenced above for which services are on which class of server.  You may want to define a group var for this (that's what we did) to automatically map servers to their appropriate server class value.
+* _hosts_: Hosts on which to perform the service restarts
+* _service class_: Generally, the OpenStack project name for the services to restart.  Recognized options are nova,keystone,glance,ceilometer,heat,neutron,spice,els,world (where "world" means all services.)
+
+

ansible/playbooks/disable-glance-quota.yaml

@@ -0,0 +1,14 @@
+#!/usr/bin/env ansible-playbook -f 50
+---
+# This playbook requires the following variables:
+#   hosts - this is the host(s) that you are trying to run on. any host in the hosts file is valid
+#
+# Author: Mike Dorman <mdorman@godaddy.com>
+#
+# Usage: 
+#   ansible-playbook disable-glance-quota.yaml -k -K --extra-vars "hosts=glance-servers"
+
+- hosts: '{{ hosts }}'
+  sudo: yes
+  tasks:
+  - include: ../tasks/turn-off-glance-quota.yaml

ansible/playbooks/enable-glance-quota.yaml

@@ -0,0 +1,15 @@
+#!/usr/bin/env ansible-playbook -f 50
+---
+# This playbook requires the following variables:
+#   hosts - this is the host(s) that you are trying to run on. any host in the hosts file is valid
+#   value - value to set user_storage_quota to.  Default 21474836480 (20 GB)
+#
+# Author: Mike Dorman <mdorman@godaddy.com>
+#
+# Usage: 
+#   ansible-playbook enable-glance-quota.yaml -k -K --extra-vars "hosts=glance-servers value=21474836480"
+
+- hosts: '{{ hosts }}'
+  sudo: yes
+  tasks:
+  - include: ../tasks/turn-on-glance-quota.yaml

ansible/playbooks/orphaned-vm-clenaup.yaml

@@ -0,0 +1,18 @@
+#!/usr/bin/env ansible-playbook --forks 50
+---
+#
+# This playbook supports the following variables:
+#
+# Author: Kris Lindgren <klindgren@godaddy.com>
+#
+# hosts - host(s)/group(s) on which to run this playbook (REQUIRED)
+#
+# Example:
+#   To run without changing the plabook with differnt uuids run ansible-plabook with the following:
+#   ansible-playbook run-puppet.yaml -k -K --extra-vars "hosts=compute-servers"
+#
+
+- hosts: '{{ hosts }}'
+  sudo: yes
+  tasks:
+  - include: /etc/ansible/playbooks/tasks/orphaned-vms.yaml

ansible/playbooks/rolling-compute-reboot.yaml

@@ -0,0 +1,42 @@
+#!/usr/bin/env ansible-playbook --forks 50
+---
+# Reboot a series of compute nodes in a rolling fasion, verifying all VMs come back up on
+# each node before going on to reboot the next group.
+#
+# Author: Kris Lindgren <klindgren@godaddy.com>
+#
+# This playbook requires the following variables:
+#  api_server - this is the server that runs the nova-api instance that we will use to get a list of the running vm's on compute nodes
+#  hosts - this is the host group to perform the rolling reboot on typically would be: *-compute
+
+# This playbook also except the following *OPTIONAL* variables:
+#  reboot_parallelism (5) - How many hosts to reboot at once
+#  reboot_check_port - This is the port to check to see if the server has come back online (22)
+#  wait_delay - This is how long to wait between checks (120 seconds)
+#  wait_timeout - This is the maximum time to wait until we move on (1200 seconds)
+#  pause_for_host_boot - This is the time to wait for the host to fully restart  (3 minutes)
+ 
+# Example:
+#  ansible-playbook compute-rolling-reboot.yaml -k -K --extra-vars "api_server=api01 hosts=compute"
+
+- hosts: '{{ hosts }}'
+  sudo: yes
+  serial: "{{ reboot_parallelism | default('5') }}"
+  tasks:
+  - name: Gather list of all running vm's on the host
+    shell: source /root/keystonerc_admin; nova list --all-tenants --status Active --host {{inventory_hostname}} --fields host,OS-EXT-SRV-ATTR:instance_name,status | grep ACTIVE | awk -F" | " '{if(a[$4]){a[$4]=a[$4]","$2"+"$6} else { a[$4]=$2"+"$6}} END {for (i in a) { print i":"a[i]} }'
+    register: running_vms
+    delegate_to: '{{ api_server }}'
+
+  - include: ../tasks/rolling-reboot.yaml
+
+  - name: ensure that nova-compute is started
+    service: name=openstack-nova-compute state=started
+    register: novacompute
+
+  - name: Verify running vm's are still running
+    shell: rc=$(echo "0"); vmlist=$( echo "{{running_vms.stdout }}" | grep {{inventory_hostname }} |cut -d":" -f2,2 |awk -F"," '{for (i=1; i<=NF; i++) print $i}'); virshlist=$( virsh list | grep running | awk '{print $2}'); for i in $vmlist; do vm=$( echo $i | cut -d"+" -f2,2 ); tmp=$( echo "$virshlist" | grep $vm); if [ $? -eq 1 ]; then uuid=$( echo "$i" | cut -d"+" -f1,1); echo "$uuid"; rc=$(echo "1"); fi; done; if [ "$rc" == "1" ]; then false; else true; fi
+    register: vms_not_running
+    when: novacompute.state == "started"
+  - debug: msg="{{vms_not_running}}"
+    when: vms_not_running.rc == 1

ansible/playbooks/template-prestage.yaml

@@ -0,0 +1,112 @@
+#!/usr/bin/env ansible-playbook --forks 50
+---
+# Distribute/prestage Glance image out to many compute nodes at once using BitTorrent
+#
+# Author: Kris Lindgren <klindgren@godaddy.com>
+#
+# Sets up a BitTorrent swarm, seeded by the machine running Glance, to distribute a Glance image out to any number
+# of nova-compute nodes very quickly.  So when we roll new gold images each month, we can use this to
+# "push" them out to all compute nodes, and avoid the first provision penalty of waiting for the image
+# to transfer the first time.
+#
+# Caveats:
+#  * Only works when the Glance backend is on a traditional (local or network-based) filesystem.  Almost
+#    certainly this does not work, and may not even make sense, for Swift-backed Glance.
+#  * Firewalls or other traffic filters need to allow the BitTorrent ports through, and among, the Glance
+#    server and all compute nodes.
+#  * Assumes Glance images are stored at `/var/lib/glance/images` on the Glance server.
+#  * There are some situations where this cannot be run multiple times, if the tracker or othe BitTorrent
+#    processes are still running.  So use caution, and YMMV.
+#  * This is done completely outside the scope of Glance and Nova.  There is no Keystone authentication or
+#    access controls.  You must have ssh and sudo access to all machines involved for this to work.
+#
+# This playbook requires the following variables:
+#  image_uuid - this can be gotten from the output of either nova image-list or glance image-list for the image you want to prestage
+#  image_sha1 - this can be gotten by running: echo -n "<image_uuid" | sha1sum | awk '{print $1'}  on any linux box
+#  image_md5 - this can be gotten by running: md5sum /var/lib/glance/images/<image_uuid> | awk '{print $1}' on the glance server
+#  tracker_host - this is the host that runs the tracker (also this is the same host in the first and second play)
+#  hosts_to_update - this is the host group to place the image onto typically *-compute
+
+# To run without changing the plabook with differnt uuids run ansible-plabook with the following:
+#  ansible-playbook template-prestage.yaml -k -K --extra-vars "image_uuid=41009dbd-52f5-4972-b65f-c429b1d42f5f image_sha1=1b8cddc7825df74e19d0a621ce527a0272541c35 image_md5=41d45920d859a2d5bd4d1ed98adf7668 tracker_host=api01 hosts_to_update=compute"
+
+- hosts: '{{ tracker_host }}'
+  sudo: yes
+  vars:
+    image_uuid: '{{ image_uuid }}'
+    tracker_host: '{{ tracker_host }}'
+  tasks:
+  - name: install ctorrent client
+    yum: name=ctorrent state=present
+
+  - name: install opentracker-ipv4
+    yum: name=opentracker-ipv4 state=present
+
+  - name: make sane
+    shell: "killall -9 opentracker-ipv4 | true; killall -9 ctorrent | true;"
+
+  - name: Start Tracker
+    command: "{{item}}"
+    with_items:
+     - /usr/bin/opentracker-ipv4 -m -p 6969 -P 6969 -d /var/opentracker
+
+  - name: Create bittorrent file
+    command: "{{item}}"
+    with_items:
+     - mkdir -p /var/www/html/torrent
+     - rm -rf /var/www/html/torrent/{{ image_uuid }}.torrent
+     - /usr/bin/ctorrent -t -s /var/www/html/torrent/{{ image_uuid }}.torrent -u http://{{ tracker_host }}:6969/announce -c Testfile /var/lib/glance/images/{{ image_uuid }}
+
+  - name: Seed Bittorrent file
+    command: /usr/bin/ctorrent -d -U 50000 -s /var/lib/glance/images/{{ image_uuid }} /var/www/html/torrent/{{ image_uuid }}.torrent
+
+- hosts: '{{hosts_to_update}}'
+  sudo: yes
+  vars:
+    image_uuid: '{{ image_uuid }}'
+    image_sha1: '{{ image_sha1 }}'
+    image_md5: '{{ image_md5 }}'
+    tracker_host: '{{ tracker_host }}'
+  tasks:
+  - name: install ctorrent client
+    yum: name=ctorrent state=present
+
+  - name: Check if image exits
+    stat: path=/var/lib/nova/instances/_base/{{ image_sha1 }}
+    register: image
+
+  - name: make sane
+    shell: "killall -9 ctorrent | true; iptables -D INPUT -p tcp --dport 2704:2706 -j ACCEPT | true"
+    when: image.stat.exists == False
+
+  - name: Download Torrent File and run torrent
+    command: "{{item}}"
+    with_items:
+     - /sbin/iptables -I INPUT -p tcp --dport 2704:2706 -j ACCEPT
+     - /usr/bin/wget http://{{ tracker_host }}/torrent/{{ image_uuid }}.torrent
+     - /usr/bin/ctorrent -e 0 -m 10 -U 30000 -D 80000 -p 2706 -s /var/lib/elsprecachedir/{{ image_uuid }} {{ image_uuid }}.torrent
+    when: image.stat.exists == False
+
+  - name: insure md5sum matches
+    shell: "md5sum /var/lib/elsprecachedir/{{ image_uuid }} | grep {{ image_md5 }}"
+    when: image.stat.exists == False
+
+  - name: Convert image to raw file
+    command: "{{item}}"
+    with_items:
+     - /usr/bin/qemu-img convert -f qcow2 -O raw /var/lib/elsprecachedir/{{ image_uuid }} /var/lib/nova/instances/_base/{{ image_sha1 }}
+     - /bin/chown nova:qemu /var/lib/nova/instances/_base/{{ image_sha1 }}
+     - /bin/chmod 644 /var/lib/nova/instances/_base/{{ image_sha1 }}
+    when: image.stat.exists == False
+
+  - name: Cleanup
+    shell: "/sbin/iptables -D INPUT -p tcp --dport 2704:2706 -j ACCEPT | true; rm -rf {{ image_uuid }}*; rm -rf /var/lib/elsprecachedir/{{ image_uuid }}; killall -9 ctorrent | true"
+    when: image.stat.exists == False
+
+- hosts: '{{ tracker_host }}'
+  sudo: yes
+  vars:
+    image_uuid: '{{ image_uuid }}'
+  tasks:
+  - name: Kill tracker and ctorrent and remove torrent file
+    shell: "killall -9 ctorrent | true ; killall -9 opentracker-ipv4 | true; rm -rf /var/www/html/torrent/{{ image_uuid }}"

ansible/tasks/orphaned-vms.yaml

@@ -0,0 +1,3 @@
+  - name: Running puppet deployment script
+    shell: "tools/remove-deleted-orphans.sh"
+

ansible/tasks/rolling-reboot.yaml

@@ -0,0 +1,21 @@
+---
+
+  - name: Rebooting Server
+    shell: sleep 2 && /sbin/shutdown -r now &
+    tags: reboot
+  
+  - name: Waiting for port to go down from server reboot
+    wait_for: host={{ inventory_hostname }} port={{ reboot_check_port | default('22') }} timeout={{ wait_timeout | default('1200') }} state=stopped
+    connection: local
+    sudo: false
+    tags: reboot
+
+  - name: Waiting for port to come back after reboot
+    wait_for: host={{ inventory_hostname }} port={{ reboot_check_port | default('22') }} delay={{ wait_delay | default('120') }} timeout={{ wait_timeout | default('1200') }} state=started
+    connection: local
+    sudo: false
+    tags: reboot
+
+  - name: pausing to make sure host is fully booted
+    pause: minutes={{ pause_for_host_boot | default('3') }}
+    tags: reboot

ansible/tasks/turn-off-glance-quota.yaml

@@ -0,0 +1,4 @@
+  - name: Removing user_storage_quota setting from glance-api.conf
+    shell: "sed -r -i 's/^[[:space:]]*user_storage_quota/#user_storage_quota/g' /etc/glance/glance-api.conf"
+  - service: name=openstack-glance-api state=restarted
+

ansible/tasks/turn-on-glance-quota.yaml

@@ -0,0 +1,3 @@
+  - name: Adding user_storage_quota setting to glance-api.conf
+    shell: "sed -r -i '0,/^#?[[:space:]]*user_storage_quota/s/^#?[[:space:]]*user_storage_quota[[:space:]]*=[[:space:]]*[[:digit:]]+/user_storage_quota = {{ value | default('21474836480') }}/' /etc/glance/glance-api.conf"
+  - service: name=openstack-glance-api state=restarted

libvirt/cleanup-orphaned-vms.sh

@@ -0,0 +1,35 @@
+#!/usr/bin/env bash
+#
+# Run this script on a compute node to cleanup/remove any orphaned KVM VMs that were left behind by something.
+# Run with --noop to do a dry run and not actually delete anything
+#
+# To populate the UUIDS value below, run the following command on a Nova api server to get list of VM UUIDs that are known to OpenStack:
+#  nova list --all-tenants | awk '{print $2;}' | grep -E '^[0-9a-f]+' | tr '\n' '|' | sed -r 's/\|$/\n/'
+# Then paste in the results for UUIDS below OR define it in the environment before running this script.
+#
+# Author: Kris Lindgren <klindgren@godaddy.com>
+#
+
+#UUIDS=""
+
+if [ -z "$UUIDS" ]; then
+  echo "UUIDS value not defined"
+  exit 1
+fi
+
+for i in `virsh list --all | grep -E '^ [0-9-]+' | awk '{print $2;}'` ; do
+
+  virsh dumpxml $i | grep "source file" | grep -E "$UUIDS" >/dev/null
+  if [ $? -ne 0 ]; then
+    echo -n "+ $i is NOT known to OpenStack, removing managedsave info... "
+    [ ! -z "$1" ] && virsh managedsave-remove $i 1>/dev/null 2>&1
+    echo -n "destroying VM... "
+    [ ! -z "$1" ] && virsh destroy $i 1>/dev/null 2>&1
+    echo -n "undefining VM... "
+    [ ! -z "$1" ] && virsh undefine $i 1>/dev/null 2>&1
+    echo DONE
+  else
+    echo "* $i is known to OpenStack, not removing."
+  fi
+
+done

libvirt/remove-deleted-orphans.sh

@@ -0,0 +1,45 @@
+#!/bin/bash
+#
+# This script will look at the configured vm's and will check to make sure that their disk drive still exists,
+# If not then it will remove the vm from libvirt.  This fixes the nova errors about disks missing from VM's
+#
+# Author: Kris Lindgren <klindgren@godaddy.com>
+#
+
+removeorphan(){
+	local domain
+	local tmp
+	domain=$1
+ 
+	tmp=$( virsh destroy $domain )
+	tmp=$( virsh undefine $domain )
+	tmp=$(virsh list --all | grep $domain )
+	if [ $? -eq 1 ]
+	then
+		tmp=$( ps auxwwwf | grep $domain | grep -v grep )
+		if [ $? -eq 1 ]
+		then
+			return 0
+		fi
+	fi
+	return 1
+ 
+}
+ 
+for i in /etc/libvirt/qemu/*.xml
+do
+	disklocation=$( grep /var/lib/nova/instances $i | grep disk | cut -d"'" -f2,2)
+	if [ ! -e $disklocation ]
+	then
+		orphan=$(echo $i | cut -d"/" -f5,5 | cut -d"." -f1,1)
+		echo "$orphan does not have a disk located at: $disklocation"
+		echo "This is an orphan of openstack... stopping the orphaned vm."
+		removeorphan $orphan
+		if [ $? -eq 0 ]
+		then
+			echo "Domain $orphan has been shutdown and removed"
+		else
+			echo "Domain $orphan has *NOT* been shutdown and removed"
+		fi
+	fi
+done

nova/backfillconfigdrive.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+#
+# A quick and dirty script to backfill empty config drive
+# images to VMs that don't already have a config drive.
+#
+# This is a workaround for the config drive bug described
+# at https://bugs.launchpad.net/nova/+bug/1356534
+#
+# Author: Mike Dorman <mdorman@godaddy.com>
+#
+
+cd /root
+mkdir -p blank
+mkisofs -o blank.iso blank/ >/dev/null 2>&1
+rmdir blank
+
+for i in `ls /var/lib/nova/instances | grep -E '[a-f0-9]{8}-[a-f0-9]{4}-[a-f0-9]{4}-[a-f0-9]{4}-[a-f0-9]{12}'`; do
+  ls -l /var/lib/nova/instances/$i/disk.config
+  if [ ! -s /var/lib/nova/instances/$i/disk.config ]; then
+    echo "$i config drive doesn't exist, or is size zero."
+    cp -f /root/blank.iso /var/lib/nova/instances/$i/disk.config
+    chown qemu:qemu /var/lib/nova/instances/$i/disk.config
+  fi
+done
+

nova/list-vms-by-host.sh

@@ -0,0 +1,20 @@
+#!/bin/bash
+#
+# Outputs a tab-delimited list of all VMs with these fields:
+# [Hypervisor Host] [UUID] [Status] [IP Address] [Name]
+#
+# Author: Mike Dorman <mdorman@godaddy.com>
+#
+
+for i in `nova list --all-tenants | grep -v '^+-' | grep -v '^| ID' | awk '{print $2 "," $4 "," $6;}'`; do
+  ID=`echo $i | cut -d, -f 1`
+  NAME=`echo $i | cut -d, -f 2`
+  STATUS=`echo $i | cut -d, -f 3`
+
+  SHOW=`nova show ${ID}`
+  HV=`echo "${SHOW}" | grep OS-EXT-SRV-ATTR:host | awk '{print $4;}'`
+  IP=`echo "${SHOW}" | grep " network" | awk '{print $5;}'`
+
+  echo -e "${HV}\t${ID}\t${STATUS}\t${IP}\t${NAME}"
+
+done

nova/list-vms-without-tenants.sh

@@ -0,0 +1,32 @@
+#!/bin/bash
+#
+# Lists VMs which have been orphaned from their tenant (i.e. the tenant
+# was removed, but VMs were still in the tenant.)
+#
+# Author: Kris Lindgren <klindgren@godaddy.com>
+#
+
+echo "THIS SCRIPT NEED TO HAVE keystonerc sourced to work"
+sleep 5
+
+echo "Getting a list of vm's from nova..."
+novavmsraw=$( nova list --all-tenants --fields name,tenant_id,user_id )
+echo "done."
+echo "Getting a list of tenants from keystone...."
+keystoneraw=$( keystone tenant-list )
+echo "done."
+novatenants=$( echo "$novavmsraw" | awk '{print $6}' | sort | uniq | grep -v Tenant )
+echo "Starting to list vm's that are no longer attached to a tenant..."
+echo "Fields are:"
+echo "|  VM ID                               |     VM Name                               | Tenant Id                        | User Id        |"
+for i in $novatenants
+do
+	tmp=$( echo "$keystoneraw" | grep $i )
+	if [ $? -eq 0 ]
+	then
+		continue
+	else
+		vms=$( echo "$novavmsraw" | grep $i )
+		echo "$vms"
+	fi
+done