devstack/stack.sh

1537 lines
44 KiB
Bash
Executable File

#!/usr/bin/env bash
# ``stack.sh`` is an opinionated OpenStack developer installation. It
# installs and configures various combinations of **Cinder**, **Glance**,
# **Horizon**, **Keystone**, **Nova**, **Neutron**, and **Swift**
# This script's options can be changed by setting appropriate environment
# variables. You can configure things like which git repositories to use,
# services to enable, OS images to use, etc. Default values are located in the
# ``stackrc`` file. If you are crafty you can run the script on multiple nodes
# using shared settings for common resources (eg., mysql or rabbitmq) and build
# a multi-node developer install.
# To keep this script simple we assume you are running on a recent **Ubuntu**
# (16.04 Xenial or newer), **Fedora** (F24 or newer), or **CentOS/RHEL**
# (7 or newer) machine. (It may work on other platforms but support for those
# platforms is left to those who added them to DevStack.) It should work in
# a VM or physical server. Additionally, we maintain a list of ``deb`` and
# ``rpm`` dependencies and other configuration files in this repo.
# Learn more and get the most recent version at http://devstack.org
# Print the commands being run so that we can see the command that triggers
# an error. It is also useful for following along as the install occurs.
set -o xtrace
# Make sure custom grep options don't get in the way
unset GREP_OPTIONS
# NOTE(sdague): why do we explicitly set locale when running stack.sh?
#
# Devstack is written in bash, and many functions used throughout
# devstack process text coming off a command (like the ip command)
# and do transforms using grep, sed, cut, awk on the strings that are
# returned. Many of these programs are internationalized, which is
# great for end users, but means that the strings that devstack
# functions depend upon might not be there in other locales. We thus
# need to pin the world to an english basis during the runs.
#
# Previously we used the C locale for this, every system has it, and
# it gives us a stable sort order. It does however mean that we
# effectively drop unicode support.... boo! :(
#
# With python3 being more unicode aware by default, that's not the
# right option. While there is a C.utf8 locale, some distros are
# shipping it as C.UTF8 for extra confusingness. And it's support
# isn't super clear across distros. This is made more challenging when
# trying to support both out of the box distros, and the gate which
# uses diskimage builder to build disk images in a different way than
# the distros do.
#
# So... en_US.utf8 it is. That's existed for a very long time. It is a
# compromise position, but it is the least worse idea at the time of
# this comment.
#
# We also have to unset other variables that might impact LC_ALL
# taking effect.
unset LANG
unset LANGUAGE
LC_ALL=en_US.utf8
export LC_ALL
# Make sure umask is sane
umask 022
# Not all distros have sbin in PATH for regular users.
PATH=$PATH:/usr/local/sbin:/usr/sbin:/sbin
# Keep track of the DevStack directory
TOP_DIR=$(cd $(dirname "$0") && pwd)
# Check for uninitialized variables, a big cause of bugs
NOUNSET=${NOUNSET:-}
if [[ -n "$NOUNSET" ]]; then
set -o nounset
fi
# Set start of devstack timestamp
DEVSTACK_START_TIME=$(date +%s)
# Configuration
# =============
# Sanity Checks
# -------------
# Clean up last environment var cache
if [[ -r $TOP_DIR/.stackenv ]]; then
rm $TOP_DIR/.stackenv
fi
# ``stack.sh`` keeps the list of ``deb`` and ``rpm`` dependencies, config
# templates and other useful files in the ``files`` subdirectory
FILES=$TOP_DIR/files
if [ ! -d $FILES ]; then
die $LINENO "missing devstack/files"
fi
# ``stack.sh`` keeps function libraries here
# Make sure ``$TOP_DIR/inc`` directory is present
if [ ! -d $TOP_DIR/inc ]; then
die $LINENO "missing devstack/inc"
fi
# ``stack.sh`` keeps project libraries here
# Make sure ``$TOP_DIR/lib`` directory is present
if [ ! -d $TOP_DIR/lib ]; then
die $LINENO "missing devstack/lib"
fi
# Check if run in POSIX shell
if [[ "${POSIXLY_CORRECT}" == "y" ]]; then
set +o xtrace
echo "You are running POSIX compatibility mode, DevStack requires bash 4.2 or newer."
exit 1
fi
# OpenStack is designed to be run as a non-root user; Horizon will fail to run
# as **root** since Apache will not serve content from **root** user).
# ``stack.sh`` must not be run as **root**. It aborts and suggests one course of
# action to create a suitable user account.
if [[ $EUID -eq 0 ]]; then
set +o xtrace
echo "DevStack should be run as a user with sudo permissions, "
echo "not root."
echo "A \"stack\" user configured correctly can be created with:"
echo " $TOP_DIR/tools/create-stack-user.sh"
exit 1
fi
# OpenStack is designed to run at a system level, with system level
# installation of python packages. It does not support running under a
# virtual env, and will fail in really odd ways if you do this. Make
# this explicit as it has come up on the mailing list.
if [[ -n "$VIRTUAL_ENV" ]]; then
set +o xtrace
echo "You appear to be running under a python virtualenv."
echo "DevStack does not support this, as we may break the"
echo "virtualenv you are currently in by modifying "
echo "external system-level components the virtualenv relies on."
echo "We recommend you use a separate virtual-machine if "
echo "you are worried about DevStack taking over your system."
exit 1
fi
# Provide a safety switch for devstack. If you do a lot of devstack,
# on a lot of different environments, you sometimes run it on the
# wrong box. This makes there be a way to prevent that.
if [[ -e $HOME/.no-devstack ]]; then
set +o xtrace
echo "You've marked this host as a no-devstack host, to save yourself from"
echo "running devstack accidentally. If this is in error, please remove the"
echo "~/.no-devstack file"
exit 1
fi
# Prepare the environment
# -----------------------
# Initialize variables:
LAST_SPINNER_PID=""
# Import common functions
source $TOP_DIR/functions
# Import config functions
source $TOP_DIR/inc/meta-config
# Import 'public' stack.sh functions
source $TOP_DIR/lib/stack
# Determine what system we are running on. This provides ``os_VENDOR``,
# ``os_RELEASE``, ``os_PACKAGE``, ``os_CODENAME``
# and ``DISTRO``
GetDistro
# Global Settings
# ---------------
# Check for a ``localrc`` section embedded in ``local.conf`` and extract if
# ``localrc`` does not already exist
# Phase: local
rm -f $TOP_DIR/.localrc.auto
extract_localrc_section $TOP_DIR/local.conf $TOP_DIR/localrc $TOP_DIR/.localrc.auto
# ``stack.sh`` is customizable by setting environment variables. Override a
# default setting via export:
#
# export DATABASE_PASSWORD=anothersecret
# ./stack.sh
#
# or by setting the variable on the command line:
#
# DATABASE_PASSWORD=simple ./stack.sh
#
# Persistent variables can be placed in a ``local.conf`` file:
#
# [[local|localrc]]
# DATABASE_PASSWORD=anothersecret
# DATABASE_USER=hellaroot
#
# We try to have sensible defaults, so you should be able to run ``./stack.sh``
# in most cases. ``local.conf`` is not distributed with DevStack and will never
# be overwritten by a DevStack update.
#
# DevStack distributes ``stackrc`` which contains locations for the OpenStack
# repositories, branches to configure, and other configuration defaults.
# ``stackrc`` sources the ``localrc`` section of ``local.conf`` to allow you to
# safely override those settings.
if [[ ! -r $TOP_DIR/stackrc ]]; then
die $LINENO "missing $TOP_DIR/stackrc - did you grab more than just stack.sh?"
fi
source $TOP_DIR/stackrc
# write /etc/devstack-version
write_devstack_version
# Warn users who aren't on an explicitly supported distro, but allow them to
# override check and attempt installation with ``FORCE=yes ./stack``
if [[ ! ${DISTRO} =~ (xenial|artful|bionic|stretch|jessie|f28|f29|opensuse-42.3|opensuse-15.0|opensuse-tumbleweed|rhel7) ]]; then
echo "WARNING: this script has not been tested on $DISTRO"
if [[ "$FORCE" != "yes" ]]; then
die $LINENO "If you wish to run this script anyway run with FORCE=yes"
fi
fi
# Local Settings
# --------------
# Make sure the proxy config is visible to sub-processes
export_proxy_variables
# Remove services which were negated in ``ENABLED_SERVICES``
# using the "-" prefix (e.g., "-rabbit") instead of
# calling disable_service().
disable_negated_services
# Configure sudo
# --------------
# We're not as **root** so make sure ``sudo`` is available
is_package_installed sudo || install_package sudo
# UEC images ``/etc/sudoers`` does not have a ``#includedir``, add one
sudo grep -q "^#includedir.*/etc/sudoers.d" /etc/sudoers ||
echo "#includedir /etc/sudoers.d" | sudo tee -a /etc/sudoers
# Conditionally setup detailed logging for sudo
if [[ -n "$LOG_SUDO" ]]; then
TEMPFILE=`mktemp`
echo "Defaults log_output" > $TEMPFILE
chmod 0440 $TEMPFILE
sudo chown root:root $TEMPFILE
sudo mv $TEMPFILE /etc/sudoers.d/00_logging
fi
# Set up DevStack sudoers
TEMPFILE=`mktemp`
echo "$STACK_USER ALL=(root) NOPASSWD:ALL" >$TEMPFILE
# Some binaries might be under ``/sbin`` or ``/usr/sbin``, so make sure sudo will
# see them by forcing ``PATH``
echo "Defaults:$STACK_USER secure_path=/sbin:/usr/sbin:/usr/bin:/bin:/usr/local/sbin:/usr/local/bin" >> $TEMPFILE
echo "Defaults:$STACK_USER !requiretty" >> $TEMPFILE
chmod 0440 $TEMPFILE
sudo chown root:root $TEMPFILE
sudo mv $TEMPFILE /etc/sudoers.d/50_stack_sh
# Configure Distro Repositories
# -----------------------------
# For Debian/Ubuntu make apt attempt to retry network ops on it's own
if is_ubuntu; then
echo 'APT::Acquire::Retries "20";' | sudo tee /etc/apt/apt.conf.d/80retry >/dev/null
fi
# Some distros need to add repos beyond the defaults provided by the vendor
# to pick up required packages.
function _install_epel {
# NOTE: We always remove and install latest -- some environments
# use snapshot images, and if EPEL version updates they break
# unless we update them to latest version.
if sudo yum repolist enabled epel | grep -q 'epel'; then
uninstall_package epel-release || true
fi
# This trick installs the latest epel-release from a bootstrap
# repo, then removes itself (as epel-release installed the
# "real" repo).
#
# You would think that rather than this, you could use
# $releasever directly in .repo file we create below. However
# RHEL gives a $releasever of "6Server" which breaks the path;
# see https://bugzilla.redhat.com/show_bug.cgi?id=1150759
cat <<EOF | sudo tee /etc/yum.repos.d/epel-bootstrap.repo
[epel-bootstrap]
name=Bootstrap EPEL
mirrorlist=http://mirrors.fedoraproject.org/mirrorlist?repo=epel-7&arch=\$basearch
failovermethod=priority
enabled=0
gpgcheck=0
EOF
# Enable a bootstrap repo. It is removed after finishing
# the epel-release installation.
is_package_installed yum-utils || install_package yum-utils
sudo yum-config-manager --enable epel-bootstrap
yum_install epel-release || \
die $LINENO "Error installing EPEL repo, cannot continue"
sudo rm -f /etc/yum.repos.d/epel-bootstrap.repo
}
function _install_rdo {
# There are multiple options for this, including using CloudSIG
# repositories (centos-release-*), trunk versions, etc. Since
# we're not interested in the actual openstack distributions
# (since we're using git to run!) but only peripherial packages
# like kvm or ovs, this has been reliable.
# TODO(ianw): figure out how to best mirror -- probably use infra
# mirror RDO reverse proxy. We could either have test
# infrastructure set it up disabled like EPEL, or fiddle it here.
# Per the point above, it's a bunch of repos so starts getting a
# little messy...
if ! is_package_installed rdo-release ; then
yum_install https://rdoproject.org/repos/rdo-release.rpm
fi
# Also enable optional for RHEL7 proper. Note this is a silent
# no-op on other platforms.
sudo yum-config-manager --enable rhel-7-server-optional-rpms
# Enable the Software Collections (SCL) repository for CentOS.
# This repository includes useful software (e.g. the Go Toolset)
# which is not present in the main repository.
if [[ "$os_VENDOR" =~ (CentOS) ]]; then
yum_install centos-release-scl
fi
if is_oraclelinux; then
sudo yum-config-manager --enable ol7_optional_latest ol7_addons ol7_MySQL56
fi
}
# Configure Target Directories
# ----------------------------
# Destination path for installation ``DEST``
DEST=${DEST:-/opt/stack}
# Create the destination directory and ensure it is writable by the user
# and read/executable by everybody for daemons (e.g. apache run for horizon)
sudo mkdir -p $DEST
safe_chown -R $STACK_USER $DEST
safe_chmod 0755 $DEST
# Destination path for devstack logs
if [[ -n ${LOGDIR:-} ]]; then
mkdir -p $LOGDIR
fi
# Destination path for service data
DATA_DIR=${DATA_DIR:-${DEST}/data}
sudo mkdir -p $DATA_DIR
safe_chown -R $STACK_USER $DATA_DIR
safe_chmod 0755 $DATA_DIR
# Configure proper hostname
# Certain services such as rabbitmq require that the local hostname resolves
# correctly. Make sure it exists in /etc/hosts so that is always true.
LOCAL_HOSTNAME=`hostname -s`
if ! fgrep -qwe "$LOCAL_HOSTNAME" /etc/hosts; then
sudo sed -i "s/\(^127.0.0.1.*\)/\1 $LOCAL_HOSTNAME/" /etc/hosts
fi
# If you have all the repos installed above already setup (e.g. a CI
# situation where they are on your image) you may choose to skip this
# to speed things up
SKIP_EPEL_INSTALL=$(trueorfalse False SKIP_EPEL_INSTALL)
if [[ $DISTRO == "rhel7" ]]; then
# If we have /etc/ci/mirror_info.sh assume we're on a OpenStack CI
# node, where EPEL is installed (but disabled) and already
# pointing at our internal mirror
if [[ -f /etc/ci/mirror_info.sh ]]; then
SKIP_EPEL_INSTALL=True
sudo yum-config-manager --enable epel
fi
if [[ ${SKIP_EPEL_INSTALL} != True ]]; then
_install_epel
fi
# Along with EPEL, CentOS (and a-likes) require some packages only
# available in RDO repositories (e.g. OVS, or later versions of
# kvm) to run.
_install_rdo
fi
# Ensure python is installed
# --------------------------
is_package_installed python || install_package python
# Configure Logging
# -----------------
# Set up logging level
VERBOSE=$(trueorfalse True VERBOSE)
VERBOSE_NO_TIMESTAMP=$(trueorfalse False VERBOSE)
# Draw a spinner so the user knows something is happening
function spinner {
local delay=0.75
local spinstr='/-\|'
printf "..." >&3
while [ true ]; do
local temp=${spinstr#?}
printf "[%c]" "$spinstr" >&3
local spinstr=$temp${spinstr%"$temp"}
sleep $delay
printf "\b\b\b" >&3
done
}
function kill_spinner {
if [ ! -z "$LAST_SPINNER_PID" ]; then
kill >/dev/null 2>&1 $LAST_SPINNER_PID
printf "\b\b\bdone\n" >&3
fi
}
# Echo text to the log file, summary log file and stdout
# echo_summary "something to say"
function echo_summary {
if [[ -t 3 && "$VERBOSE" != "True" ]]; then
kill_spinner
echo -n -e $@ >&6
spinner &
LAST_SPINNER_PID=$!
else
echo -e $@ >&6
fi
}
# Echo text only to stdout, no log files
# echo_nolog "something not for the logs"
function echo_nolog {
echo $@ >&3
}
# Set up logging for ``stack.sh``
# Set ``LOGFILE`` to turn on logging
# Append '.xxxxxxxx' to the given name to maintain history
# where 'xxxxxxxx' is a representation of the date the file was created
TIMESTAMP_FORMAT=${TIMESTAMP_FORMAT:-"%F-%H%M%S"}
LOGDAYS=${LOGDAYS:-7}
CURRENT_LOG_TIME=$(date "+$TIMESTAMP_FORMAT")
if [[ -n "$LOGFILE" ]]; then
# Clean up old log files. Append '.*' to the user-specified
# ``LOGFILE`` to match the date in the search template.
LOGFILE_DIR="${LOGFILE%/*}" # dirname
LOGFILE_NAME="${LOGFILE##*/}" # basename
mkdir -p $LOGFILE_DIR
find $LOGFILE_DIR -maxdepth 1 -name $LOGFILE_NAME.\* -mtime +$LOGDAYS -exec rm {} \;
LOGFILE=$LOGFILE.${CURRENT_LOG_TIME}
SUMFILE=$LOGFILE.summary.${CURRENT_LOG_TIME}
# Redirect output according to config
# Set fd 3 to a copy of stdout. So we can set fd 1 without losing
# stdout later.
exec 3>&1
if [[ "$VERBOSE" == "True" ]]; then
_of_args="-v"
if [[ "$VERBOSE_NO_TIMESTAMP" == "True" ]]; then
_of_args="$_of_args --no-timestamp"
fi
# Set fd 1 and 2 to write the log file
exec 1> >( $TOP_DIR/tools/outfilter.py $_of_args -o "${LOGFILE}" ) 2>&1
# Set fd 6 to summary log file
exec 6> >( $TOP_DIR/tools/outfilter.py -o "${SUMFILE}" )
else
# Set fd 1 and 2 to primary logfile
exec 1> >( $TOP_DIR/tools/outfilter.py -o "${LOGFILE}" ) 2>&1
# Set fd 6 to summary logfile and stdout
exec 6> >( $TOP_DIR/tools/outfilter.py -v -o "${SUMFILE}" >&3 )
fi
echo_summary "stack.sh log $LOGFILE"
# Specified logfile name always links to the most recent log
ln -sf $LOGFILE $LOGFILE_DIR/$LOGFILE_NAME
ln -sf $SUMFILE $LOGFILE_DIR/$LOGFILE_NAME.summary
else
# Set up output redirection without log files
# Set fd 3 to a copy of stdout. So we can set fd 1 without losing
# stdout later.
exec 3>&1
if [[ "$VERBOSE" != "True" ]]; then
# Throw away stdout and stderr
exec 1>/dev/null 2>&1
fi
# Always send summary fd to original stdout
exec 6> >( $TOP_DIR/tools/outfilter.py -v >&3 )
fi
# Basic test for ``$DEST`` path permissions (fatal on error unless skipped)
check_path_perm_sanity ${DEST}
# Configure Error Traps
# ---------------------
# Kill background processes on exit
trap exit_trap EXIT
function exit_trap {
local r=$?
jobs=$(jobs -p)
# Only do the kill when we're logging through a process substitution,
# which currently is only to verbose logfile
if [[ -n $jobs && -n "$LOGFILE" && "$VERBOSE" == "True" ]]; then
echo "exit_trap: cleaning up child processes"
kill 2>&1 $jobs
fi
#Remove timing data file
if [ -f "$OSCWRAP_TIMER_FILE" ] ; then
rm "$OSCWRAP_TIMER_FILE"
fi
# Kill the last spinner process
kill_spinner
if [[ $r -ne 0 ]]; then
echo "Error on exit"
# If we error before we've installed os-testr, this will fail.
if type -p generate-subunit > /dev/null; then
generate-subunit $DEVSTACK_START_TIME $SECONDS 'fail' >> ${SUBUNIT_OUTPUT}
fi
if [[ -z $LOGDIR ]]; then
$TOP_DIR/tools/worlddump.py
else
$TOP_DIR/tools/worlddump.py -d $LOGDIR
fi
else
# If we error before we've installed os-testr, this will fail.
if type -p generate-subunit > /dev/null; then
generate-subunit $DEVSTACK_START_TIME $SECONDS >> ${SUBUNIT_OUTPUT}
fi
fi
exit $r
}
# Exit on any errors so that errors don't compound
trap err_trap ERR
function err_trap {
local r=$?
set +o xtrace
if [[ -n "$LOGFILE" ]]; then
echo "${0##*/} failed: full log in $LOGFILE"
else
echo "${0##*/} failed"
fi
exit $r
}
# Begin trapping error exit codes
set -o errexit
# Print the kernel version
uname -a
# Reset the bundle of CA certificates
SSL_BUNDLE_FILE="$DATA_DIR/ca-bundle.pem"
rm -f $SSL_BUNDLE_FILE
# Import common services (database, message queue) configuration
source $TOP_DIR/lib/database
source $TOP_DIR/lib/rpc_backend
# Configure Projects
# ==================
# Clone all external plugins
fetch_plugins
# Plugin Phase 0: override_defaults - allow plugins to override
# defaults before other services are run
run_phase override_defaults
# Import Apache functions
source $TOP_DIR/lib/apache
# Import TLS functions
source $TOP_DIR/lib/tls
# Source project function libraries
source $TOP_DIR/lib/infra
source $TOP_DIR/lib/libraries
source $TOP_DIR/lib/lvm
source $TOP_DIR/lib/horizon
source $TOP_DIR/lib/keystone
source $TOP_DIR/lib/glance
source $TOP_DIR/lib/nova
source $TOP_DIR/lib/placement
source $TOP_DIR/lib/cinder
source $TOP_DIR/lib/swift
source $TOP_DIR/lib/neutron
source $TOP_DIR/lib/ldap
source $TOP_DIR/lib/dstat
source $TOP_DIR/lib/etcd3
# Extras Source
# --------------
# Phase: source
run_phase source
# Interactive Configuration
# -------------------------
# Do all interactive config up front before the logging spew begins
# Generic helper to configure passwords
function read_password {
local xtrace
xtrace=$(set +o | grep xtrace)
set +o xtrace
var=$1; msg=$2
pw=${!var}
if [[ -f $RC_DIR/localrc ]]; then
localrc=$TOP_DIR/localrc
else
localrc=$TOP_DIR/.localrc.password
fi
# If the password is not defined yet, proceed to prompt user for a password.
if [ ! $pw ]; then
# If there is no localrc file, create one
if [ ! -e $localrc ]; then
touch $localrc
fi
# Presumably if we got this far it can only be that our
# localrc is missing the required password. Prompt user for a
# password and write to localrc.
echo ''
echo '################################################################################'
echo $msg
echo '################################################################################'
echo "This value will be written to ${localrc} file so you don't have to enter it "
echo "again. Use only alphanumeric characters."
echo "If you leave this blank, a random default value will be used."
pw=" "
while true; do
echo "Enter a password now:"
read -e $var
pw=${!var}
[[ "$pw" = "`echo $pw | tr -cd [:alnum:]`" ]] && break
echo "Invalid chars in password. Try again:"
done
if [ ! $pw ]; then
pw=$(generate_hex_string 10)
fi
eval "$var=$pw"
echo "$var=$pw" >> $localrc
fi
# restore previous xtrace value
$xtrace
}
# Database Configuration
# ----------------------
# To select between database backends, add the following to ``local.conf``:
#
# disable_service mysql
# enable_service postgresql
#
# The available database backends are listed in ``DATABASE_BACKENDS`` after
# ``lib/database`` is sourced. ``mysql`` is the default.
initialize_database_backends && echo "Using $DATABASE_TYPE database backend" || echo "No database enabled"
# Queue Configuration
# -------------------
# Rabbit connection info
# In multi node DevStack, second node needs ``RABBIT_USERID``, but rabbit
# isn't enabled.
if is_service_enabled rabbit; then
read_password RABBIT_PASSWORD "ENTER A PASSWORD TO USE FOR RABBIT."
fi
# Keystone
# --------
if is_service_enabled keystone; then
# Services authenticate to Identity with servicename/``SERVICE_PASSWORD``
read_password SERVICE_PASSWORD "ENTER A SERVICE_PASSWORD TO USE FOR THE SERVICE AUTHENTICATION."
# Horizon currently truncates usernames and passwords at 20 characters
read_password ADMIN_PASSWORD "ENTER A PASSWORD TO USE FOR HORIZON AND KEYSTONE (20 CHARS OR LESS)."
# Keystone can now optionally install OpenLDAP by enabling the ``ldap``
# service in ``local.conf`` (e.g. ``enable_service ldap``).
# To clean out the Keystone contents in OpenLDAP set ``KEYSTONE_CLEAR_LDAP``
# to ``yes`` (e.g. ``KEYSTONE_CLEAR_LDAP=yes``) in ``local.conf``. To enable the
# Keystone Identity Driver (``keystone.identity.backends.ldap.Identity``)
# set ``KEYSTONE_IDENTITY_BACKEND`` to ``ldap`` (e.g.
# ``KEYSTONE_IDENTITY_BACKEND=ldap``) in ``local.conf``.
# Only request LDAP password if the service is enabled
if is_service_enabled ldap; then
read_password LDAP_PASSWORD "ENTER A PASSWORD TO USE FOR LDAP"
fi
fi
# Swift
# -----
if is_service_enabled s-proxy; then
# We only ask for Swift Hash if we have enabled swift service.
# ``SWIFT_HASH`` is a random unique string for a swift cluster that
# can never change.
read_password SWIFT_HASH "ENTER A RANDOM SWIFT HASH."
if [[ -z "$SWIFT_TEMPURL_KEY" ]] && [[ "$SWIFT_ENABLE_TEMPURLS" == "True" ]]; then
read_password SWIFT_TEMPURL_KEY "ENTER A KEY FOR SWIFT TEMPURLS."
fi
fi
# Save configuration values
save_stackenv $LINENO
# Install Packages
# ================
# OpenStack uses a fair number of other projects.
# Bring down global requirements before any use of pip_install. This is
# necessary to ensure that the constraints file is in place before we
# attempt to apply any constraints to pip installs.
git_clone $REQUIREMENTS_REPO $REQUIREMENTS_DIR $REQUIREMENTS_BRANCH
# Install package requirements
# Source it so the entire environment is available
echo_summary "Installing package prerequisites"
source $TOP_DIR/tools/install_prereqs.sh
# Configure an appropriate Python environment
if [[ "$OFFLINE" != "True" ]]; then
PYPI_ALTERNATIVE_URL=${PYPI_ALTERNATIVE_URL:-""} $TOP_DIR/tools/install_pip.sh
fi
# Install subunit for the subunit output stream
pip_install -U os-testr
TRACK_DEPENDS=${TRACK_DEPENDS:-False}
# Install Python packages into a virtualenv so that we can track them
if [[ $TRACK_DEPENDS = True ]]; then
echo_summary "Installing Python packages into a virtualenv $DEST/.venv"
pip_install -U virtualenv
rm -rf $DEST/.venv
virtualenv --system-site-packages $DEST/.venv
source $DEST/.venv/bin/activate
$DEST/.venv/bin/pip freeze > $DEST/requires-pre-pip
fi
# Do the ugly hacks for broken packages and distros
source $TOP_DIR/tools/fixup_stuff.sh
fixup_all
if [[ "$USE_SYSTEMD" == "True" ]]; then
pip_install_gr systemd-python
# the default rate limit of 1000 messages / 30 seconds is not
# sufficient given how verbose our logging is.
iniset -sudo /etc/systemd/journald.conf "Journal" "RateLimitBurst" "0"
sudo systemctl restart systemd-journald
fi
# Virtual Environment
# -------------------
# Install required infra support libraries
install_infra
# Install bindep
$VIRTUALENV_CMD $DEST/bindep-venv
# TODO(ianw) : optionally install from zuul checkout?
$DEST/bindep-venv/bin/pip install bindep
# Extras Pre-install
# ------------------
# Phase: pre-install
run_phase stack pre-install
# NOTE(danms): Set global limits before installing anything
set_systemd_override DefaultLimitNOFILE ${ULIMIT_NOFILE}
install_rpc_backend
restart_rpc_backend
if is_service_enabled $DATABASE_BACKENDS; then
install_database
fi
if [ -n "$DATABASE_TYPE" ]; then
install_database_python
fi
if is_service_enabled neutron; then
install_neutron_agent_packages
fi
if is_service_enabled etcd3; then
install_etcd3
fi
# Check Out and Install Source
# ----------------------------
echo_summary "Installing OpenStack project source"
# Install additional libraries
install_libs
# Install uwsgi
install_apache_uwsgi
# Install client libraries
install_keystoneauth
install_keystoneclient
install_glanceclient
install_cinderclient
install_novaclient
if is_service_enabled swift glance horizon; then
install_swiftclient
fi
if is_service_enabled neutron nova horizon; then
install_neutronclient
fi
# Setup TLS certs
if is_service_enabled tls-proxy; then
configure_CA
init_CA
init_cert
fi
# Install middleware
install_keystonemiddleware
if is_service_enabled keystone; then
if [ "$KEYSTONE_AUTH_HOST" == "$SERVICE_HOST" ]; then
stack_install_service keystone
configure_keystone
fi
fi
if is_service_enabled swift; then
if is_service_enabled ceilometer; then
install_ceilometermiddleware
fi
stack_install_service swift
configure_swift
# swift3 middleware to provide S3 emulation to Swift
if is_service_enabled swift3; then
# Replace the nova-objectstore port by the swift port
S3_SERVICE_PORT=8080
git_clone $SWIFT3_REPO $SWIFT3_DIR $SWIFT3_BRANCH
setup_develop $SWIFT3_DIR
fi
fi
if is_service_enabled g-api n-api; then
# Image catalog service
stack_install_service glance
configure_glance
fi
if is_service_enabled cinder; then
# Block volume service
stack_install_service cinder
configure_cinder
fi
if is_service_enabled neutron; then
# Network service
stack_install_service neutron
fi
if is_service_enabled nova; then
# Compute service
stack_install_service nova
configure_nova
fi
if is_service_enabled placement; then
# placement api
stack_install_service placement
configure_placement
fi
# create a placement-client fake service to know we need to configure
# placement connectivity. We configure the placement service for nova
# if placement-api or placement-client is active, and n-cpu on the
# same box.
if is_service_enabled placement placement-client; then
if is_service_enabled n-cpu || is_service_enabled n-sch; then
configure_placement_nova_compute
fi
fi
if is_service_enabled horizon; then
# dashboard
stack_install_service horizon
fi
if is_service_enabled tls-proxy; then
fix_system_ca_bundle_path
if python3_enabled ; then
fix_system_ca_bundle_path python3
fi
fi
# Extras Install
# --------------
# Phase: install
run_phase stack install
# Install the OpenStack client, needed for most setup commands
if use_library_from_git "python-openstackclient"; then
git_clone_by_name "python-openstackclient"
setup_dev_lib "python-openstackclient"
else
pip_install_gr python-openstackclient
fi
# Installs alias for osc so that we can collect timing for all
# osc commands. Alias dies with stack.sh.
install_oscwrap
if [[ $TRACK_DEPENDS = True ]]; then
$DEST/.venv/bin/pip freeze > $DEST/requires-post-pip
if ! diff -Nru $DEST/requires-pre-pip $DEST/requires-post-pip > $DEST/requires.diff; then
echo "Detect some changes for installed packages of pip, in depend tracking mode"
cat $DEST/requires.diff
fi
echo "Ran stack.sh in depend tracking mode, bailing out now"
exit 0
fi
# Syslog
# ------
if [[ $SYSLOG != "False" ]]; then
if [[ "$SYSLOG_HOST" = "$HOST_IP" ]]; then
# Configure the master host to receive
cat <<EOF | sudo tee /etc/rsyslog.d/90-stack-m.conf >/dev/null
\$ModLoad imrelp
\$InputRELPServerRun $SYSLOG_PORT
EOF
else
# Set rsyslog to send to remote host
cat <<EOF | sudo tee /etc/rsyslog.d/90-stack-s.conf >/dev/null
*.* :omrelp:$SYSLOG_HOST:$SYSLOG_PORT
EOF
fi
RSYSLOGCONF="/etc/rsyslog.conf"
if [ -f $RSYSLOGCONF ]; then
sudo cp -b $RSYSLOGCONF $RSYSLOGCONF.bak
if [[ $(grep '$SystemLogRateLimitBurst' $RSYSLOGCONF) ]]; then
sudo sed -i 's/$SystemLogRateLimitBurst\ .*/$SystemLogRateLimitBurst\ 0/' $RSYSLOGCONF
else
sudo sed -i '$ i $SystemLogRateLimitBurst\ 0' $RSYSLOGCONF
fi
if [[ $(grep '$SystemLogRateLimitInterval' $RSYSLOGCONF) ]]; then
sudo sed -i 's/$SystemLogRateLimitInterval\ .*/$SystemLogRateLimitInterval\ 0/' $RSYSLOGCONF
else
sudo sed -i '$ i $SystemLogRateLimitInterval\ 0' $RSYSLOGCONF
fi
fi
echo_summary "Starting rsyslog"
restart_service rsyslog
fi
# Export Certificate Authority Bundle
# -----------------------------------
# If certificates were used and written to the SSL bundle file then these
# should be exported so clients can validate their connections.
if [ -f $SSL_BUNDLE_FILE ]; then
export OS_CACERT=$SSL_BUNDLE_FILE
fi
# Configure database
# ------------------
if is_service_enabled $DATABASE_BACKENDS; then
configure_database
fi
# Save configuration values
save_stackenv $LINENO
# Kernel Samepage Merging (KSM)
# -----------------------------
# Processes that mark their memory as mergeable can share identical memory
# pages if KSM is enabled. This is particularly useful for nova + libvirt
# backends but any other setup that marks its memory as mergeable can take
# advantage. The drawback is there is higher cpu load; however, we tend to
# be memory bound not cpu bound so enable KSM by default but allow people
# to opt out if the CPU time is more important to them.
if [[ $ENABLE_KSM == "True" ]] ; then
if [[ -f /sys/kernel/mm/ksm/run ]] ; then
sudo sh -c "echo 1 > /sys/kernel/mm/ksm/run"
fi
fi
# Start Services
# ==============
# Dstat
# -----
# A better kind of sysstat, with the top process per time slice
start_dstat
# Etcd
# -----
# etcd is a distributed key value store that provides a reliable way to store data across a cluster of machines
if is_service_enabled etcd3; then
start_etcd3
fi
# Keystone
# --------
# Rather than just export these, we write them out to a
# intermediate userrc file that can also be used to debug if
# something goes wrong between here and running
# tools/create_userrc.sh (this script relies on services other
# than keystone being available, so we can't call it right now)
cat > $TOP_DIR/userrc_early <<EOF
# Use this for debugging issues before files in accrc are created
# Set up password auth credentials now that Keystone is bootstrapped
export OS_IDENTITY_API_VERSION=3
export OS_AUTH_URL=$KEYSTONE_AUTH_URI
export OS_USERNAME=admin
export OS_USER_DOMAIN_ID=default
export OS_PASSWORD=$ADMIN_PASSWORD
export OS_PROJECT_NAME=admin
export OS_PROJECT_DOMAIN_ID=default
export OS_REGION_NAME=$KEYSTONE_REGION_NAME
EOF
if is_service_enabled tls-proxy; then
echo "export OS_CACERT=$INT_CA_DIR/ca-chain.pem" >> $TOP_DIR/userrc_early
start_tls_proxy http-services '*' 443 $SERVICE_HOST 80
fi
source $TOP_DIR/userrc_early
if is_service_enabled keystone; then
echo_summary "Starting Keystone"
if [ "$KEYSTONE_AUTH_HOST" == "$SERVICE_HOST" ]; then
init_keystone
start_keystone
bootstrap_keystone
fi
create_keystone_accounts
if is_service_enabled nova; then
create_nova_accounts
fi
if is_service_enabled glance; then
create_glance_accounts
fi
if is_service_enabled cinder; then
create_cinder_accounts
fi
if is_service_enabled neutron; then
create_neutron_accounts
fi
if is_service_enabled swift; then
create_swift_accounts
fi
fi
# Write a clouds.yaml file
write_clouds_yaml
# Horizon
# -------
if is_service_enabled horizon; then
echo_summary "Configuring Horizon"
configure_horizon
fi
# Glance
# ------
if is_service_enabled g-reg; then
echo_summary "Configuring Glance"
init_glance
fi
# Neutron
# -------
if is_service_enabled neutron; then
echo_summary "Configuring Neutron"
configure_neutron
# Run init_neutron only on the node hosting the Neutron API server
if is_service_enabled $DATABASE_BACKENDS && is_service_enabled neutron; then
init_neutron
fi
fi
# Nova
# ----
if is_service_enabled n-net q-dhcp; then
# Delete traces of nova networks from prior runs
# Do not kill any dnsmasq instance spawned by NetworkManager
netman_pid=$(pidof NetworkManager || true)
if [ -z "$netman_pid" ]; then
sudo killall dnsmasq || true
else
sudo ps h -o pid,ppid -C dnsmasq | grep -v $netman_pid | awk '{print $1}' | sudo xargs kill || true
fi
clean_iptables
if is_service_enabled n-net; then
rm -rf ${NOVA_STATE_PATH}/networks
sudo mkdir -p ${NOVA_STATE_PATH}/networks
safe_chown -R ${STACK_USER} ${NOVA_STATE_PATH}/networks
fi
# Force IP forwarding on, just in case
sudo sysctl -w net.ipv4.ip_forward=1
fi
# Storage Service
# ---------------
if is_service_enabled swift; then
echo_summary "Configuring Swift"
init_swift
fi
# Volume Service
# --------------
if is_service_enabled cinder; then
echo_summary "Configuring Cinder"
init_cinder
fi
# Placement Service
# ---------------
if is_service_enabled placement; then
echo_summary "Configuring placement"
init_placement
fi
# Compute Service
# ---------------
if is_service_enabled nova; then
echo_summary "Configuring Nova"
init_nova
# Additional Nova configuration that is dependent on other services
if is_service_enabled neutron; then
configure_neutron_nova
elif is_service_enabled n-net; then
create_nova_conf_nova_network
fi
init_nova_cells
fi
# Extras Configuration
# ====================
# Phase: post-config
run_phase stack post-config
# Local Configuration
# ===================
# Apply configuration from ``local.conf`` if it exists for layer 2 services
# Phase: post-config
merge_config_group $TOP_DIR/local.conf post-config
# Launch Services
# ===============
# Only run the services specified in ``ENABLED_SERVICES``
# Launch Swift Services
if is_service_enabled swift; then
echo_summary "Starting Swift"
start_swift
fi
# Launch the Glance services
if is_service_enabled glance; then
echo_summary "Starting Glance"
start_glance
fi
# Install Images
# ==============
# Upload an image to Glance.
#
# The default image is CirrOS, a small testing image which lets you login as **root**
# CirrOS has a ``cloud-init`` analog supporting login via keypair and sending
# scripts as userdata.
# See https://help.ubuntu.com/community/CloudInit for more on ``cloud-init``
if is_service_enabled g-reg; then
echo_summary "Uploading images"
for image_url in ${IMAGE_URLS//,/ }; do
upload_image $image_url
done
fi
# NOTE(lyarwood): By default use a single hardcoded fixed_key across devstack
# deployments. This ensures the keys match across nova and cinder across all
# hosts.
FIXED_KEY=${FIXED_KEY:-bae3516cc1c0eb18b05440eba8012a4a880a2ee04d584a9c1579445e675b12defdc716ec}
if is_service_enabled nova; then
iniset $NOVA_CONF key_manager fixed_key "$FIXED_KEY"
iniset $NOVA_CPU_CONF key_manager fixed_key "$FIXED_KEY"
fi
if is_service_enabled cinder; then
iniset $CINDER_CONF key_manager fixed_key "$FIXED_KEY"
fi
# Launch the nova-api and wait for it to answer before continuing
if is_service_enabled n-api; then
echo_summary "Starting Nova API"
start_nova_api
fi
if is_service_enabled neutron-api; then
echo_summary "Starting Neutron"
start_neutron_api
elif is_service_enabled q-svc; then
echo_summary "Starting Neutron"
configure_neutron_after_post_config
start_neutron_service_and_check
elif is_service_enabled $DATABASE_BACKENDS && is_service_enabled n-net; then
NM_CONF=${NOVA_CONF}
if is_service_enabled n-cell; then
NM_CONF=${NOVA_CELLS_CONF}
fi
# Create a small network
$NOVA_BIN_DIR/nova-manage --config-file $NM_CONF network create "$PRIVATE_NETWORK_NAME" $FIXED_RANGE 1 $FIXED_NETWORK_SIZE $NETWORK_CREATE_ARGS
# Create some floating ips
$NOVA_BIN_DIR/nova-manage --config-file $NM_CONF floating create $FLOATING_RANGE --pool=$PUBLIC_NETWORK_NAME
# Create a second pool
$NOVA_BIN_DIR/nova-manage --config-file $NM_CONF floating create --ip_range=$TEST_FLOATING_RANGE --pool=$TEST_FLOATING_POOL
fi
# Start placement before any of the service that are likely to want
# to use it to manage resource providers.
if is_service_enabled placement; then
echo_summary "Starting Placement"
start_placement
fi
if is_service_enabled neutron; then
start_neutron
fi
# Once neutron agents are started setup initial network elements
if is_service_enabled q-svc && [[ "$NEUTRON_CREATE_INITIAL_NETWORKS" == "True" ]]; then
echo_summary "Creating initial neutron network elements"
create_neutron_initial_network
fi
if is_service_enabled nova; then
echo_summary "Starting Nova"
start_nova
create_flavors
fi
if is_service_enabled cinder; then
echo_summary "Starting Cinder"
start_cinder
create_volume_types
fi
if is_service_enabled horizon; then
echo_summary "Starting Horizon"
init_horizon
start_horizon
fi
# Create account rc files
# =======================
# Creates source able script files for easier user switching.
# This step also creates certificates for tenants and users,
# which is helpful in image bundle steps.
if is_service_enabled nova && is_service_enabled keystone; then
USERRC_PARAMS="-PA --target-dir $TOP_DIR/accrc"
if [ -f $SSL_BUNDLE_FILE ]; then
USERRC_PARAMS="$USERRC_PARAMS --os-cacert $SSL_BUNDLE_FILE"
fi
$TOP_DIR/tools/create_userrc.sh $USERRC_PARAMS
fi
# Save some values we generated for later use
save_stackenv
# Wrapup configuration
# ====================
# local.conf extra
# ----------------
# Apply configuration from ``local.conf`` if it exists for layer 2 services
# Phase: extra
merge_config_group $TOP_DIR/local.conf extra
# Run extras
# ----------
# Phase: extra
run_phase stack extra
# local.conf post-extra
# ---------------------
# Apply late configuration from ``local.conf`` if it exists for layer 2 services
# Phase: post-extra
merge_config_group $TOP_DIR/local.conf post-extra
# Sanity checks
# =============
# Check that computes are all ready
#
# TODO(sdague): there should be some generic phase here.
if is_service_enabled n-cpu; then
is_nova_ready
fi
# Check the status of running services
service_check
# Configure nova cellsv2
# ----------------------
# Do this late because it requires compute hosts to have started
if is_service_enabled n-api; then
if is_service_enabled n-cpu; then
$TOP_DIR/tools/discover_hosts.sh
else
# Some CI systems like Hyper-V build the control plane on
# Linux, and join in non Linux Computes after setup. This
# allows them to delay the processing until after their whole
# environment is up.
echo_summary "SKIPPING Cell setup because n-cpu is not enabled. You will have to do this manually before you have a working environment."
fi
fi
# Run local script
# ----------------
# Run ``local.sh`` if it exists to perform user-managed tasks
if [[ -x $TOP_DIR/local.sh ]]; then
echo "Running user script $TOP_DIR/local.sh"
$TOP_DIR/local.sh
fi
# Bash completion
# ===============
# Prepare bash completion for OSC
openstack complete | sudo tee /etc/bash_completion.d/osc.bash_completion > /dev/null
# If cinder is configured, set global_filter for PV devices
if is_service_enabled cinder; then
if is_ubuntu; then
echo_summary "Configuring lvm.conf global device filter"
set_lvm_filter
else
echo_summary "Skip setting lvm filters for non Ubuntu systems"
fi
fi
# Run test-config
# ---------------
# Phase: test-config
run_phase stack test-config
# Apply late configuration from ``local.conf`` if it exists for layer 2 services
# Phase: test-config
merge_config_group $TOP_DIR/local.conf test-config
# Fin
# ===
set +o xtrace
if [[ -n "$LOGFILE" ]]; then
exec 1>&3
# Force all output to stdout and logs now
exec 1> >( tee -a "${LOGFILE}" ) 2>&1
else
# Force all output to stdout now
exec 1>&3
fi
# Dump out the time totals
time_totals
# Using the cloud
# ===============
echo ""
echo ""
echo ""
echo "This is your host IP address: $HOST_IP"
if [ "$HOST_IPV6" != "" ]; then
echo "This is your host IPv6 address: $HOST_IPV6"
fi
# If you installed Horizon on this server you should be able
# to access the site using your browser.
if is_service_enabled horizon; then
echo "Horizon is now available at http://$SERVICE_HOST$HORIZON_APACHE_ROOT"
fi
# If Keystone is present you can point ``nova`` cli to this server
if is_service_enabled keystone; then
echo "Keystone is serving at $KEYSTONE_SERVICE_URI/"
echo "The default users are: admin and demo"
echo "The password: $ADMIN_PASSWORD"
fi
# Warn that a deprecated feature was used
if [[ -n "$DEPRECATED_TEXT" ]]; then
echo
echo -e "WARNING: $DEPRECATED_TEXT"
echo
fi
# If USE_SYSTEMD is enabled, tell the user about using it.
if [[ "$USE_SYSTEMD" == "True" ]]; then
echo
echo "Services are running under systemd unit files."
echo "For more information see: "
echo "https://docs.openstack.org/devstack/latest/systemd.html"
echo
fi
# Useful info on current state
cat /etc/devstack-version
echo
# Indicate how long this took to run (bash maintained variable ``SECONDS``)
echo_summary "stack.sh completed in $SECONDS seconds."
# Restore/close logging file descriptors
exec 1>&3
exec 2>&3
exec 3>&-
exec 6>&-