openstack/trove
Code Issues Proposed changes

Improve guest image creation in devstack

Browse Source 
Change-Id: I0ef8cbc4babc5c6b9f2936923ffe2837ecc3da74
This commit is contained in:
Lingxian Kong 2019-09-06 00:19:19 +12:00
parent 6c50dec163
commit 3c09e6178a
7 changed files with 491 additions and 675 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

230
devstack/plugin.sh
View File

@ -75,9 +75,8 @@ function create_trove_accounts {
fi
}
# _cleanup_trove_apache_wsgi - Removes all the WSGI related files and
# restart apache.
function _cleanup_trove_apache_wsgi {
# Removes all the WSGI related files and restart apache.
function cleanup_trove_apache_wsgi {
sudo rm -rf $TROVE_WSGI_DIR
sudo rm -f $(apache_site_config_for trove-api)
restart_apache_server
@ -98,7 +97,7 @@ function cleanup_trove {
if [[ "${TROVE_USE_MOD_WSGI}" == "TRUE" ]]; then
echo "Cleaning up Trove's WSGI setup"
_cleanup_trove_apache_wsgi
cleanup_trove_apache_wsgi
fi
}
@ -165,9 +164,8 @@ function configure_nova_kvm {
echo "configure_nova_kvm: using virt_type: ${virt_type} for cpu: ${cpu}."
}
# _config_trove_apache_wsgi() - Setup WSGI config files for Trove and
# enable the site
function _config_trove_apache_wsgi {
# Setup WSGI config files for Trove and enable the site
function config_trove_apache_wsgi {
local trove_apache_conf
sudo mkdir -p ${TROVE_WSGI_DIR}
@ -258,7 +256,7 @@ function configure_trove {
# configure apache related files
if [[ "${TROVE_USE_MOD_WSGI}" == "TRUE" ]]; then
echo "Configuring Trove to use mod-wsgi and Apache"
_config_trove_apache_wsgi
config_trove_apache_wsgi
fi
# Use these values only if they're set
@ -317,54 +315,12 @@ function install_python_troveclient {
fi
}
# init_trove() - Initializes Trove Database as a Service
function init_trove {
function init_trove_db {
# (Re)Create trove db
recreate_database trove
# Initialize the trove database
$TROVE_MANAGE db_sync
# build and upload sample Trove mysql instance if not set otherwise.
# We recommend to use trovestack tooling for image build.
TROVE_DISABLE_IMAGE_SETUP=`echo ${TROVE_DISABLE_IMAGE_SETUP,,}`
if [[ ${TROVE_DISABLE_IMAGE_SETUP} != "true" ]]; then
echo "Setup datastore image."
_setup_minimal_image
else
echo "Skip datastore image building."
fi
[ -z "$TROVE_GUEST_IMAGE_URL" ] && return 0
# Find the glance id for the trove guest image
# The image is uploaded by stack.sh -- see $IMAGE_URLS handling
GUEST_IMAGE_NAME=$(basename "$TROVE_GUEST_IMAGE_URL")
GUEST_IMAGE_NAME=${GUEST_IMAGE_NAME%.*}
TOKEN=$(openstack token issue -c id -f value)
TROVE_GUEST_IMAGE_ID=$(openstack --os-token $TOKEN --os-url $GLANCE_SERVICE_PROTOCOL://$GLANCE_HOSTPORT image list | grep "${GUEST_IMAGE_NAME}" | get_field 1)
if [ -z "$TROVE_GUEST_IMAGE_ID" ]; then
# If no glance id is found, skip remaining setup
echo "Datastore ${TROVE_DATASTORE_TYPE} will not be created: guest image ${GUEST_IMAGE_NAME} not found."
return 1
fi
# Now that we have the guest image id, initialize appropriate datastores / datastore versions
$TROVE_MANAGE datastore_update "$TROVE_DATASTORE_TYPE" ""
$TROVE_MANAGE datastore_version_update "$TROVE_DATASTORE_TYPE" "$TROVE_DATASTORE_VERSION" "$TROVE_DATASTORE_TYPE" \
"$TROVE_GUEST_IMAGE_ID" "$TROVE_DATASTORE_PACKAGE" 1
$TROVE_MANAGE datastore_version_update "$TROVE_DATASTORE_TYPE" "inactive_version" "inactive_manager" "$TROVE_GUEST_IMAGE_ID" "" 0
$TROVE_MANAGE datastore_update "$TROVE_DATASTORE_TYPE" "$TROVE_DATASTORE_VERSION"
$TROVE_MANAGE datastore_update "Inactive_Datastore" ""
# Some datastores provide validation rules.
# if one is provided, configure it.
if [ -f "${TROVE_DIR}/trove/templates/${TROVE_DATASTORE_TYPE}"/validation-rules.json ]; then
echo "Configuring validation rules for ${TROVE_DATASTORE_TYPE}"
$TROVE_MANAGE db_load_datastore_config_parameters \
"$TROVE_DATASTORE_TYPE" "$TROVE_DATASTORE_VERSION" \
"${TROVE_DIR}/trove/templates/${TROVE_DATASTORE_TYPE}"/validation-rules.json
fi
}
function create_mgmt_subnet_v4 {
@ -380,7 +336,7 @@ function create_mgmt_subnet_v4 {
# Create private IPv6 subnet
# Note: Trove is not fully tested in IPv6.
function _create_subnet_v6 {
function create_subnet_v6 {
local project_id=$1
local net_id=$2
local name=$3
@ -426,13 +382,13 @@ function setup_mgmt_network() {
fi
# Trove doesn't support IPv6 for now.
# if [[ "$IP_VERSION" =~ .*6 ]]; then
# NEW_IPV6_SUBNET_ID=$(_create_subnet_v6 ${PROJECT_ID} ${network_id} ${IPV6_SUBNET_NAME})
# NEW_IPV6_SUBNET_ID=$(create_subnet_v6 ${PROJECT_ID} ${network_id} ${IPV6_SUBNET_NAME})
# openstack router add subnet $ROUTER_ID $NEW_IPV6_SUBNET_ID
# fi
}
# Set up Trove management network and make configuration change.
function finalize_trove_network {
function config_trove_network {
echo "Finalizing Neutron networking for Trove"
echo "Dumping current network parameters:"
echo " SERVICE_HOST: $SERVICE_HOST"
@ -531,117 +487,38 @@ function configure_tempest_for_trove {
fi
}
# _setup_minimal_image() - build and register in Trove a vm image with mysql
# - datastore can be set via env variables
# (lxkong): This function is deprecated in favor of trovestack script.
function _setup_minimal_image {
##### Prerequisites:
##### - SSH KEYS has to be created on controller
##### - trove will access controller ip to get trove source code by using HOST_SCP_USERNAME and an ssh key
##### - we assume tripleo elements and all other elements have been downloaded
echo "Exporting image-related environmental variables"
PRIMARY_IP=$(ip route get 8.8.8.8 | head -1 | awk '{print $7}')
export CONTROLLER_IP=${CONTROLLER_IP:-$PRIMARY_IP}
export HOST_USERNAME=${HOST_USERNAME:-'stack'}
export HOST_SCP_USERNAME=${HOST_SCP_USERNAME:-'stack'}
export GUEST_USERNAME=${GUEST_USERNAME:-'ubuntu'}
export PATH_TROVE=${PATH_TROVE:-'/opt/stack/trove'}
export ESCAPED_PATH_TROVE=$(echo $PATH_TROVE | sed 's/\//\\\//g')
export TROVESTACK_SCRIPTS=${TROVESTACK_SCRIPTS:-'/opt/stack/trove/integration/scripts'}
export TROVE_DATASTORE_TYPE=${TROVE_DATASTORE_TYPE:-'mysql'}
export TROVE_DATASTORE_VERSION=${TROVE_DATASTORE_VERSION:-'5.7'}
export TROVE_DATASTORE_PACKAGE=${TROVE_DATASTORE_PACKAGE:-"${TROVE_DATASTORE_TYPE}-${TROVE_DATASTORE_VERSION}"}
export SSH_DIR=${SSH_DIR:-'/opt/stack/.ssh'}
export GUEST_LOGDIR=${GUEST_LOGDIR:-'/var/log/trove/'}
export ESCAPED_GUEST_LOGDIR=$(echo $GUEST_LOGDIR | sed 's/\//\\\//g')
export DIB_CLOUD_INIT_DATASOURCES="ConfigDrive"
export DISTRO="ubuntu"
export VM=${VM:-"/opt/stack/images/${DISTRO}_${TROVE_DATASTORE_TYPE}/${DISTRO}_${TROVE_DATASTORE_TYPE}"}
if [ -d "$TROVESTACK_SCRIPTS/files/elements" ]; then
export ELEMENTS_PATH=$TROVESTACK_SCRIPTS/files/elements
else
export ELEMENTS_PATH=.
# Use trovestack to create guest image and register the image in the datastore.
function create_guest_image {
TROVE_DISABLE_IMAGE_SETUP=`echo ${TROVE_DISABLE_IMAGE_SETUP,,}`
if [[ ${TROVE_DISABLE_IMAGE_SETUP} == "true" ]]; then
echo "Skip creating guest image."
return 0
fi
if [ ! -z "$PATH_DISKIMAGEBUILDER" ]; then
export ELEMENTS_PATH+=:$PATH_DISKIMAGEBUILDER/elements
elif [ -d "/usr/local/lib/python2.7/dist-packages/diskimage_builder" ]; then
PATH_DISKIMG="/usr/local/lib/python2.7/dist-packages/diskimage_builder"
export ELEMENTS_PATH+=:$PATH_DISKIMG/elements
fi
echo "Starting to create guest image..."
if [ ! -z "$PATH_TRIPLEO_ELEMENTS" ]; then
export ELEMENTS_PATH+=:$PATH_TRIPLEO_ELEMENTS/elements
else
git_clone $TRIPLEO_IMAGES_REPO $TRIPLEO_IMAGES_DIR $TRIPLEO_IMAGES_BRANCH
setup_develop $TRIPLEO_IMAGES_DIR
$DEST/trove/integration/scripts/trovestack build-image ${TROVE_DATASTORE_TYPE} ${TROVE_IMAGE_OS} ${TROVE_IMAGE_OS_RELEASE} true
export ELEMENTS_PATH+=:$TRIPLEO_IMAGES_DIR/elements
fi
export QEMU_IMG_OPTIONS="--qemu-img-options compat=1.1"
export RELEASE=${RELEASE:-'xenial'}
export DIB_APT_CONF_DIR=/etc/apt/apt.conf.d
export DIB_CLOUD_INIT_ETC_HOSTS=true
export DIB_RELEASE=${RELEASE:-'xenial'}
# https://cloud-images.ubuntu.com/releases is more stable than the daily
# builds(https://cloud-images.ubuntu.com/xenial/current/),
# e.g. sometimes SHA256SUMS file is missing in the daily builds
declare -A releasemapping=( ["xenial"]="16.04" ["bionic"]="18.04")
export DIB_CLOUD_IMAGES="https://cloud-images.ubuntu.com/releases/${DIB_RELEASE}/release/"
export BASE_IMAGE_FILE="ubuntu-${releasemapping[${DIB_RELEASE}]}-server-cloudimg-amd64-root.tar.gz"
export TROVE_GUESTAGENT_CONF=${TROVE_GUESTAGENT_CONF:-'/etc/trove/trove-guestagent.conf'}
if [[ -d ${SSH_DIR} && -f ${SSH_DIR}/id_rsa.pub ]]; then
cat ${SSH_DIR}/id_rsa.pub >> ${SSH_DIR}/authorized_keys
sort ${SSH_DIR}/authorized_keys | uniq > ${SSH_DIR}/authorized_keys.uniq
mv ${SSH_DIR}/authorized_keys.uniq ${SSH_DIR}/authorized_keys
else
mkdir -p ${SSH_DIR}
/usr/bin/ssh-keygen -f ${SSH_DIR}/id_rsa -q -N ""
cat ${SSH_DIR}/id_rsa.pub >> ${SSH_DIR}/authorized_keys
chmod 600 ${SSH_DIR}/authorized_keys
fi
# Make sure the guest agent has permission to ssh into the devstack host
# in order to download trove code during the service initialization.
home_keys=$HOME/.ssh/authorized_keys
cat ${SSH_DIR}/id_rsa.pub >> ${home_keys}
sort ${home_keys} | uniq > ${home_keys}.uniq
mv ${home_keys}.uniq ${home_keys}
echo "Run disk image create to actually create a new image"
disk-image-create -a amd64 -o "${VM}" -x ${QEMU_IMG_OPTIONS} ${DISTRO} \
vm cloud-init-datasources ${DISTRO}-guest ${DISTRO}-${RELEASE}-guest \
${DISTRO}-${TROVE_DATASTORE_TYPE} ${DISTRO}-${RELEASE}-${TROVE_DATASTORE_TYPE}
QCOW_IMAGE="$VM.qcow2"
if [ ! -f $QCOW_IMAGE ]; then
echo "Image file was not found at $QCOW_IMAGE. Probably it was not created."
image_name=${TROVE_IMAGE_OS}-${TROVE_DATASTORE_TYPE}
image_file=$HOME/images/${image_name}.qcow2
if [ ! -f ${image_file} ]; then
echo "Image file was not found at ${image_file}. Probably it was not created."
return 1
fi
ACTIVE=1
INACTIVE=0
echo "Add image to glance"
GLANCE_OUT=$(openstack --os-url $GLANCE_SERVICE_PROTOCOL://$GLANCE_HOSTPORT \
image create $DISTRO-${TROVE_DATASTORE_TYPE}-${TROVE_DATASTORE_VERSION} \
--public --disk-format qcow2 --container-format bare --file $QCOW_IMAGE)
glance_image_id=$(echo "$GLANCE_OUT" | grep '| id ' | awk '{print $4}')
echo "Add the image to glance"
glance_image_id=$(openstack --os-region-name RegionOne --os-password ${SERVICE_PASSWORD} \
--os-project-name service --os-username trove \
image create ${TROVE_IMAGE_OS}-${TROVE_IMAGE_OS_RELEASE}-${TROVE_DATASTORE_TYPE} \
--disk-format qcow2 --container-format bare --property hw_rng_model='virtio' --file ${image_file} \
-c id -f value)
echo "Create datastore specific entry in Trove AFAIK one per datastore, do not need when changing image"
echo "Register the image in datastore"
$TROVE_MANAGE datastore_update $TROVE_DATASTORE_TYPE ""
echo "Connect datastore entry to glance image"
$TROVE_MANAGE datastore_version_update $TROVE_DATASTORE_TYPE $TROVE_DATASTORE_VERSION $TROVE_DATASTORE_TYPE $glance_image_id "" $ACTIVE
echo "Set default datastore version"
$TROVE_MANAGE datastore_update $TROVE_DATASTORE_TYPE $TROVE_DATASTORE_VERSION
# just for tests
@ -649,24 +526,22 @@ function _setup_minimal_image {
$TROVE_MANAGE datastore_update Test_Datastore_1 ""
echo "Add validation rules if available"
if [ -f "$PATH_TROVE"/trove/templates/$TROVE_DATASTORE_TYPE/validation-rules.json ]; then
if [ -f $DEST/trove/trove/templates/$TROVE_DATASTORE_TYPE/validation-rules.json ]; then
$TROVE_MANAGE db_load_datastore_config_parameters "$TROVE_DATASTORE_TYPE" "$TROVE_DATASTORE_VERSION" \
"$PATH_TROVE"/trove/templates/$TROVE_DATASTORE_TYPE/validation-rules.json
$DEST/trove/trove/templates/$TROVE_DATASTORE_TYPE/validation-rules.json
fi
echo "Generate cloudinit"
CLOUDINIT_PATH=/etc/trove/cloudinit/mysql.cloudinit
if [ ! -f $CLOUDINIT_PATH ]; then
sudo mkdir -p $(dirname $CLOUDINIT_PATH)
sudo echo "#!/usr/bin/env bash" | sudo tee $CLOUDINIT_PATH
PUBKEY=`cat ${SSH_DIR}/id_rsa.pub`
sudo echo "echo '${PUBKEY}' > /home/${GUEST_USERNAME}/.ssh/authorized_keys" | sudo tee --append $CLOUDINIT_PATH
fi
# To avoid 'Connection timed out' error of sudo command inside the guest agent.
CLOUDINIT_PATH=/etc/trove/cloudinit/${TROVE_DATASTORE_TYPE}.cloudinit
sudo mkdir -p $(dirname "$CLOUDINIT_PATH")
sudo touch "$CLOUDINIT_PATH"
sudo tee $CLOUDINIT_PATH >/dev/null <<'EOF'
#cloud-config
manage_etc_hosts: "localhost"
EOF
}
function _config_nova_keypair {
function config_nova_keypair {
export SSH_DIR=${SSH_DIR:-"$HOME/.ssh"}
if [[ ! -f ${SSH_DIR}/id_rsa.pub ]]; then
@ -689,7 +564,14 @@ function _config_nova_keypair {
iniset $TROVE_CONF DEFAULT nova_keypair ${TROVE_MGMT_KEYPAIR_NAME}
}
function _config_mgmt_security_group {
function config_cinder_volume_type {
volume_type=$(openstack --os-region-name RegionOne --os-password ${SERVICE_PASSWORD} \
--os-project-name service --os-username trove \
volume type list -c Name -f value | awk 'NR==1 {print}')
iniset $TROVE_CONF DEFAULT cinder_volume_type ${volume_type}
}
function config_mgmt_security_group {
local sgid
echo "Creating Trove management security group."
@ -719,17 +601,13 @@ if is_service_enabled trove; then
echo_summary "Configuring Trove"
configure_trove
elif [[ "$1" == "stack" && "$2" == "extra" ]]; then
# Initialize trove
init_trove
init_trove_db
config_nova_keypair
config_cinder_volume_type
config_mgmt_security_group
config_trove_network
create_guest_image
_config_nova_keypair
_config_mgmt_security_group
# finish the last step in trove network configuration
echo_summary "Finalizing Trove Network Configuration"
finalize_trove_network
# Start the trove API and trove taskmgr components
echo_summary "Starting Trove"
start_trove

2
devstack/settings
View File

@ -28,6 +28,8 @@ TROVE_LOCAL_CONF_DIR=${TROVE_LOCAL_CONF_DIR:-${TROVE_DIR}/etc/trove}
TROVE_LOCAL_API_PASTE_INI=${TROVE_LOCAL_API_PASTE_INI:-${TROVE_LOCAL_CONF_DIR}/api-paste.ini}
TROVE_LOCAL_POLICY_JSON=${TROVE_LOCAL_POLICY_JSON:-${TROVE_LOCAL_CONF_DIR}/policy.json}
TROVE_IMAGE_OS=${TROVE_IMAGE_OS:-"ubuntu"}
TROVE_IMAGE_OS_RELEASE=${TROVE_IMAGE_OS_RELEASE:-"xenial"}
TROVE_DATASTORE_TYPE=${TROVE_DATASTORE_TYPE:-"mysql"}
if [[ "$DISTRO" == "xenial" || "$DISTRO" == "bionic" ]]; then
TROVE_DATASTORE_VERSION=${TROVE_DATASTORE_VERSION:-"5.7"}

15
doc/source/install/index.rst
View File

@ -2,22 +2,19 @@
Database service
================
The Database service (Trove) provides scalable and reliable Cloud Database as a
Service provisioning functionality for both relational and non-relational
database engines.
.. toctree::
apache-mod-wsgi.rst
get_started.rst
install.rst
install-devstack.rst
install-manual.rst
install-obs.rst
install-rdo.rst
install-ubuntu.rst
manual_install.rst
dashboard.rst
verify.rst
next-steps.rst
The Database service (trove) provides cloud provisioning functionality
for database engines.
This chapter assumes a working setup of OpenStack following the
`OpenStack Installation Tutorial <http://docs.openstack.org/#install-guides>`_.

186
doc/source/install/install-devstack.rst Normal file
View File

@ -0,0 +1,186 @@
..
Copyright 2019 Catalyst Cloud
All Rights Reserved.
not use this file except in compliance with the License. You may obtain
a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
License for the specific language governing permissions and limitations
under the License.
Install Trove in DevStack
=========================
This page describes how to set up a working development
environment that can be used in deploying Trove on latest releases
of Ubuntu.
Following these instructions will allow you to have a fully functional Trove
environment using the DevStack on Ubuntu 16.04 or 18.04.
Config DevStack with Trove
~~~~~~~~~~~~~~~~~~~~~~~~~~
Trove can be enabled in devstack by using the plug-in based interface it
offers.
.. note::
The following steps have been fully verified both on Ubuntu 16.04 and 18.04.
Start by cloning the devstack repository using a non-root user(the default user
is ``ubuntu``) and change to devstack directory:
.. code-block:: console
git clone https://opendev.org/openstack/devstack
cd devstack/
Create the ``local.conf`` file with the following minimal devstack
configuration, change the ``HOST_IP`` to your own devstack host IP address:
.. code-block:: ini
[[local|localrc]]
RECLONE=False
HOST_IP=<your-host-ip-here>
enable_plugin trove https:/opendev.org/openstack/trove
LIBS_FROM_GIT+=,python-troveclient
DATABASE_PASSWORD=password
ADMIN_PASSWORD=password
SERVICE_PASSWORD=password
SERVICE_TOKEN=password
RABBIT_PASSWORD=password
LOGFILE=$DEST/logs/stack.sh.log
VERBOSE=True
LOG_COLOR=False
LOGDAYS=1
IPV4_ADDRS_SAFE_TO_USE=10.111.0.0/26
FIXED_RANGE=10.111.0.0/26
NETWORK_GATEWAY=10.111.0.1
FLOATING_RANGE=172.30.5.0/24
PUBLIC_NETWORK_GATEWAY=172.30.5.1
# Pre-requisites
ENABLED_SERVICES=rabbit,mysql,key
# Nova
enable_service n-api
enable_service n-cpu
enable_service n-cond
enable_service n-sch
enable_service n-api-meta
enable_service placement-api
enable_service placement-client
# Glance
enable_service g-api
enable_service g-reg
# Cinder
enable_service cinder
enable_service c-api
enable_service c-vol
enable_service c-sch
# Neutron
enable_service q-svc
enable_service q-agt
enable_service q-dhcp
enable_service q-l3
enable_service q-meta
# enable DVR
Q_PLUGIN=ml2
Q_ML2_TENANT_NETWORK_TYPE=vxlan
Q_DVR_MODE=legacy
# Swift
ENABLED_SERVICES+=,swift
SWIFT_HASH=66a3d6b56c1f479c8b4e70ab5c2000f5
SWIFT_REPLICAS=1
# Trove
TROVE_DISABLE_IMAGE_SETUP=False
Take a look at the
`options <https://opendev.org/openstack/trove/src/branch/master/devstack/settings>`_
you could use to customize the Trove installation.
Running devstack
~~~~~~~~~~~~~~~~
Run the ``stack.sh`` script:
.. code-block:: console
./stack.sh
After it completes, you can see there is a MySQL datastore available to create
Trove instance:
.. code-block:: console
$ openstack datastore version list mysql
+--------------------------------------+------------------+
| ID | Name |
+--------------------------------------+------------------+
| 9726354d-f989-4a68-9c5f-6e37b1bccc74 | 5.7 |
| f81a8448-2f6e-4746-8d97-866ab7dcccee | inactive_version |
+--------------------------------------+------------------+
Create your first Trove instance
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
#. Switch to a non-admin user, choose a Nova flavor ID for the Trove
instance.
.. code-block:: console
$ source ~/devstack/openrc demo demo
$ openstack flavor list
+----+---------------+-------+------+-----------+-------+-----------+
| ID | Name | RAM | Disk | Ephemeral | VCPUs | Is Public |
+----+---------------+-------+------+-----------+-------+-----------+
| 1 | m1.tiny | 512 | 1 | 0 | 1 | True |
| 2 | m1.small | 2048 | 20 | 0 | 1 | True |
| 3 | m1.medium | 4096 | 40 | 0 | 2 | True |
| 4 | m1.large | 8192 | 80 | 0 | 4 | True |
| 5 | m1.xlarge | 16384 | 160 | 0 | 8 | True |
| 6 | mysql-minimum | 512 | 5 | 0 | 1 | True |
| c1 | cirros256 | 256 | 1 | 0 | 1 | True |
| d1 | ds512M | 512 | 5 | 0 | 1 | True |
| d2 | ds1G | 1024 | 10 | 0 | 1 | True |
| d3 | ds2G | 2048 | 10 | 0 | 2 | True |
| d4 | ds4G | 4096 | 20 | 0 | 4 | True |
+----+---------------+-------+------+-----------+-------+-----------+
$ flavorid=6
#. Choose a private network on which the database service can be accessed.
.. code-block:: console
$ openstack network list --internal
+--------------------------------------+---------+----------------------------------------------------------------------------+
| ID | Name | Subnets |
+--------------------------------------+---------+----------------------------------------------------------------------------+
| a0f3cf12-3562-4064-aa34-61d37265e867 | private | 377e791f-2631-4d8e-93cd-036344b24b3f, 7e04abb4-7c16-4b92-8865-7831ecf3ee66 |
+--------------------------------------+---------+----------------------------------------------------------------------------+
# netid=a0f3cf12-3562-4064-aa34-61d37265e867
#. Create the Trove instance.
.. code-block:: console
$ openstack database instance create my-first-trove-instance $flavorid \
--size 1 \
--nic net-id=$netid \
--datastore mysql --datastore_version 5.7 \
--databases test --users test_user:password

243
doc/source/install/install-manual.rst Normal file
View File

@ -0,0 +1,243 @@
.. _install-manual:
Manual Trove Installation
=========================
Objectives
~~~~~~~~~~
This document provides a step-by-step guide for manual installation of Trove
with an existing OpenStack environment for development purposes.
This document will not cover OpenStack setup for other services.
Requirements
~~~~~~~~~~~~
A running OpenStack environment installed on Ubuntu 16.04 or 18.04 LTS is
required, including the following components:
- Compute (Nova)
- Image Service (Glance)
- Identity (Keystone)
- Network (Neutron)
- If you want to provision databases on block-storage volumes, you also need
Block Storage (Cinder)
- If you want to do backup/restore or replication, you also need Object Storage
(Swift)
- AMQP service (RabbitMQ or QPID)
- MySQL (SQLite, PostgreSQL) database
Networking requirements
-----------------------
Trove makes use of an "Management Network" that the controller uses to talk to
trove instance and vice versa. All the trove instance that Trove deploys will
have interfaces and IP addresses on this network. Therefore, it’s important
that the subnet deployed on this network be sufficiently large to allow for the
maximum number of trove instance and controllers likely to be deployed
throughout the lifespan of the cloud installation.
You must also create a Neutron security group which will be applied to trove
instance port created on the management network. The cloud admin has full
control of the security group, e.g it can be helpful to allow SSH access to the
trove instance from the controller for troubleshooting purposes (ie. TCP port
22), though this is not strictly necessary in production environments.
Finally, you need to add routing or interfaces to this network so that the
Trove controller is able to communicate with Nova servers on this network.
Trove Installation
~~~~~~~~~~~~~~~~~~
Required packages for Trove
---------------------------
List of packages to be installed:
.. code-block:: bash
$ sudo apt-get install -y build-essential python-dev libpython-dev \
python-setuptools libffi-dev libxslt1-dev libxml2-dev libyaml-dev \
libssl-dev zlib1g-dev mysql-client python-pymysql libmysqlclient-dev git
Python settings
---------------
Install pip:
.. code-block:: bash
curl -SO# https://bootstrap.pypa.io/get-pip.py && sudo python get-pip.py pip==9.0.3 && rm -f get-pip.py
Install virtualenv, create Trove environment and activate it:
.. code-block:: bash
pip install virtualenv --user
virtualenv --system-site-packages trove_env
source trove_env/bin/activate
Get Trove
---------
Obtain the Trove source components from OpenStack repositories:
.. code-block:: bash
cd ~
git clone https://opendev.org/openstack/trove.git
git clone https://opendev.org/openstack/python-troveclient.git
Install Trove
-------------
First, install the requirements:
.. code-block:: bash
cd ~/trove
sudo pip install -r requirements.txt -r test-requirements.txt
Then, install Trove:
.. code-block:: bash
sudo pip install -e .
Finally, install the Trove client:
.. code-block:: bash
cd ~/python-troveclient
sudo pip install -e .
cd ~
Other required OpenStack clients (python-novaclient, python-keystoneclient,
etc.) should already be installed as part of the Trove requirements.
Prepare Trove for OpenStack
---------------------------
.. note::
You need to run the following commands using OpenStack admin credentials.
#. Create Trove service user with admin role in the ``service`` project.
.. code-block:: bash
openstack user create trove --project service --password-prompt
openstack role add --user trove --project service admin
#. Register Trove in Keystone.
.. code-block:: bash
openstack service create --name trove --description "Database" database
openstack endpoint create --region RegionOne database public 'http://<EnvironmentPublicIP>:8779/v1.0/$(tenant_id)s'
openstack endpoint create --region RegionOne database admin 'http://<EnvironmentPublicIP>:8779/v1.0/$(tenant_id)s'
openstack endpoint create --region RegionOne database internal 'http://<EnvironmentPublicIP>:8779/v1.0/$(tenant_id)s'
Where <EnvironmentPublicIP> is the IP address of the server where Trove was
installed. This IP should be reachable from any hosts that will be used to
communicate with Trove.
Trove configuration
~~~~~~~~~~~~~~~~~~~
There are several configuration files for Trove, you can find samples of the
config files in ``etc/trove/`` of Trove repo:
- api-paste.ini and trove.conf — For trove-api service
- trove-guestagent.conf — For trove-guestagent service
- ``<datastore_manager>.cloudinit`` — Userdata for VMs during provisioning
Options in trove.conf
---------------------
#. Config service tenant model, change the values according to your own
environment.
.. code-block:: ini
nova_proxy_admin_user = admin
nova_proxy_admin_pass = password
nova_proxy_admin_tenant_name = admin
nova_proxy_admin_tenant_id = f472127c03f6410899225e26a3c1d22c
nova_proxy_admin_user_domain_name = default
nova_proxy_admin_project_domain_name = default
remote_nova_client = trove.common.single_tenant_remote.nova_client_trove_admin
remote_cinder_client = trove.common.single_tenant_remote.cinder_client_trove_admin
remote_neutron_client = trove.common.single_tenant_remote.neutron_client_trove_admin
os_region_name = RegionOne
#. Management config options.
management_networks
Trove management network ID list. Cloud admin needs to create the
networks.
management_security_groups
Security group IDs that applied to the management port in the trove
instance. Cloud admin needs to create the security groups.
nova_keypair
The Nova keypair used to create trove instance. Cloud admin needs to
create the keypair.
cinder_volume_type
The Cinder volume type name used to create volume that attached to the
trove instance, otherwise, users need to provide the volume type when
creating the instance.
Prepare Trove database
~~~~~~~~~~~~~~~~~~~~~~
Create the Trove database schema:
- Connect to the storage backend (MySQL, PostgreSQL)
- Create a database called `trove` (this database will be used for storing
Trove ORM)
- Compose connection string. Example:
``mysql+pymysql://<user>:<password>@<backend_host>:<backend_port>/<database_name>``
Initialize the database
-----------------------
Once the database for Trove is created, its structure needs to be populated.
.. code-block:: bash
$ trove-manage db_sync
Create and register Trove guest image
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
To build Trove guest image, refer to
`Build guest agent image <https://docs.openstack.org/trove/latest/admin/trovestack.html#build-guest-agent-image>`_
Run Trove
~~~~~~~~~
Starting Trove services
-----------------------
Run trove-api:
.. code-block:: bash
$ trove-api --config-file=${TROVE_CONF_DIR}/trove.conf &
Run trove-taskmanager:
.. code-block:: bash
$ trove-taskmanager --config-file=${TROVE_CONF_DIR}/trove.conf &
Run trove-conductor:
.. code-block:: bash
$ trove-conductor --config-file=${TROVE_CONF_DIR}/trove.conf &

24
doc/source/install/install.rst
View File

@ -1,24 +0,0 @@
.. _trove-install:
Install and configure
~~~~~~~~~~~~~~~~~~~~~
This section describes how to install and configure the
Database service, code-named trove, on the controller node.
This section assumes that you already have a working OpenStack
environment with at least the following components installed:
Compute, Image Service, Identity.
* If you want to do backup and restore, you also need Object Storage.
* If you want to provision datastores on block-storage volumes, you also
need Block Storage.
Note that installation and configuration vary by distribution.
.. toctree::
install-obs.rst
install-rdo.rst
install-ubuntu.rst

466
doc/source/install/manual_install.rst
View File

@ -1,466 +0,0 @@
.. _manual_install:
=========================
Manual Trove Installation
=========================
Objectives
==========
This document provides a step-by-step guide for manual installation of Trove with
an existing OpenStack environment for development purposes.
This document will not cover:
- OpenStack setup
- Trove service configuration
Requirements
============
A running OpenStack environment is required, including the following components:
- Compute (Nova)
- Image Service (Glance)
- Identity (Keystone)
- Network (Neutron)
- If you want to provision datastores on block-storage volumes, you also will need Block Storage (Cinder)
- If you want to do backup/restore and replication, you will also need Object Storage (Swift)
- An environment with a freshly installed Ubuntu 16.04 LTS to run Trove services.
This will be referred to as "local environment"
- AMQP service (RabbitMQ or QPID)
- MySQL (SQLite, PostgreSQL) database for Trove's internal needs, accessible from the local environment
- Certain OpenStack services must be accessible from VMs:
- Swift
- VMs must be accessible from local environment for development/debugging purposes
- OpenStack services must be accessible directly from the local environment, such as:
- Nova
- Cinder
- Swift
- Heat
Installation
============
-----------
Gather info
-----------
The following information about the existing environment is required:
- Keystone host and port(s)
- OpenStack administrator's username, tenant name and password
- Nova URL
- Cinder URL
- Swift URL
- Heat URL
- AMQP connection credentials (server URL, user, password)
- Trove's controller backend connection string (MySQL, SQLite, PostgreSQL)
--------------------
Install dependencies
--------------------
Required packages for Trove
---------------------------
List of packages to be installed:
.. code-block:: bash
$ sudo apt-get install build-essential libxslt1-dev qemu-utils mysql-client \
git python-dev python-pexpect python-pymysql libmysqlclient-dev
Python settings
---------------
To find out which setuptools version is latest please check out the `setuptools repo`_.
.. _setuptools repo: https://pypi.org/project/setuptools/
To find out which pip version is latest please visit the `pip repo`_.
.. _pip repo: https://pypi.org/project/pip/
Some packages in Ubuntu repositories are outdated. Please make sure to update to the latest versions from the appropriate sources.
Use latest setuptools:
Go https://pypi.org/project/setuptools, download the latest source setuptools, and move it under ~
.. code-block:: bash
$ cd ~
$ tar xfvz setuptools-{{latest}}.tar.gz
$ cd setuptools-{{latest}}
$ python setup.py install --user
Use latest pip:
Go https://pypi.org/project/pip, download the latest source pip, and move it under ~
.. code-block:: bash
$ tar xfvz pip-{{latest}}.tar.gz
$ cd pip-{{latest}}
$ python setup.py install --user
Note '--user' above -- we installed packages in user's home dir, in $HOME/.local/bin, so we need to add it to path:
.. code-block:: bash
$ echo PATH="$HOME/.local/bin:$PATH" >> ~/.profile
$ . ~/.profile
Install virtualenv, create environment and activate it:
.. code-block:: bash
$ pip install virtualenv --user
$ virtualenv --system-site-packages env
$ . env/bin/activate
Get Trove
---------
Obtain the Trove source components from OpenStack repositories:
.. code-block:: bash
$ cd ~
$ git clone https://opendev.org/openstack/trove.git
$ git clone https://opendev.org/openstack/python-troveclient.git
Install Trove
=============
First, install the requirements:
.. code-block:: bash
$ cd ~/trove
$ pip install -r requirements.txt -r test-requirements.txt
Then, install Trove:
.. code-block:: bash
$ sudo python setup.py develop
Finally, install the Trove client:
.. code-block:: bash
$ cd ~/python-troveclient
$ sudo python setup.py develop
$ cd ~
Other required OpenStack clients (python-novaclient, python-keystoneclient, etc.) should already be installed as part of the Trove requirements.
---------------------------
Prepare Trove for OpenStack
---------------------------
You will first need to create a tenant called 'trove_for_trove_usage'.
Next, create users called 'regular_trove_user' and 'admin_trove_user' —using 'trove' as the password. These are the accounts used by the Trove service.
Additionally, you will need to register Trove as an OpenStack service and its endpoints:
.. code-block:: bash
$ openstack --os-username <OpenStackAdminUsername> --os-password <OpenStackAdminPassword> --os-project-name <OpenStackAdminProject> --os-domain-name <OpenstackDomainName> --os-auth-url http://<KeystoneIP>/identity project create trove_for_trove_usage
$ openstack --os-username <OpenStackAdminUsername> --os-password <OpenStackAdminPassword> --os-project-name <OpenStackAdminProject> --os-domain-name <OpenstackDomainName> --os-auth-url http://<KeystoneIP>/identity user create regular_trove_user --project trove_for_trove_usage --password-prompt
$ openstack --os-username <OpenStackAdminUsername> --os-password <OpenStackAdminPassword> --os-project-name <OpenStackAdminProject> --os-domain-name <OpenstackDomainName> --os-auth-url http://<KeystoneIP>/identity user create admin_trove_user --project trove_for_trove_usage --password-prompt
$ openstack --os-username <OpenStackAdminUsername> --os-password <OpenStackAdminPassword> --os-project-name <OpenStackAdminProject> --os-domain-name <OpenstackDomainName> --os-auth-url http://<KeystoneIP>/identity role add --user admin_trove_user --project trove_for_trove_usage admin
$ openstack --os-username <OpenStackAdminUsername> --os-password <OpenStackAdminPassword> --os-project-name <OpenStackAdminProject> --os-domain-name <OpenstackDomainName> --os-auth-url http://<KeystoneIP>/identity service create --name trove --description "Database" database
$ openstack --os-username <OpenStackAdminUsername> --os-password <OpenStackAdminPassword> --os-project-name <OpenStackAdminProject> --os-domain-name <OpenstackDomainName> --os-auth-url http://<KeystoneIP>/identity endpoint create --region RegionOne database public 'http://<EnvironmentPublicIP>:<EnvironmentPort>/v1.0/$(tenant_id)s'
$ openstack --os-username <OpenStackAdminUsername> --os-password <OpenStackAdminPassword> --os-project-name <OpenStackAdminProject> --os-domain-name <OpenstackDomainName> --os-auth-url http://<KeystoneIP>/identity endpoint create --region RegionOne database admin 'http://<EnvironmentPublicIP>:<EnvironmentPort>/v1.0/$(tenant_id)s'
$ openstack --os-username <OpenStackAdminUsername> --os-password <OpenStackAdminPassword> --os-project-name <OpenStackAdminProject> --os-domain-name <OpenstackDomainName> --os-auth-url http://<KeystoneIP>/identity endpoint create --region RegionOne database internal 'http://<EnvironmentPublicIP>:<EnvironmentPort>/v1.0/$(tenant_id)s'
Where <EnvironmentPublicIP> and <EnvironmentPort> are the IP address and Port of the server where Trove was installed. This IP should be reachable from any hosts that will be used to communicate with Trove.
Prepare Trove configuration files
=================================
There are several configuration files for Trove:
- api-paste.ini and trove.conf — For trove-api service
- trove-guestagent.conf — For trove-guestagent service
- <datastore_manager>.cloudinit — Userdata for VMs during provisioning
Cloud-init scripts are userdata that is being used for different datastore types like mysql/percona, cassandra, mongodb, redis, couchbase while provisioning new compute instances.
Samples of the above are available in ~/trove/etc/trove/ as \*.conf.sample files.
If a clean Ubuntu image is used as the source image for Trove instances, the cloud-init script must install and run guestagent in the instance.
As an alternative, one may consider creating a custom image with pre-installed and pre-configured Trove in it.
Source images
=============
As the source image for Trove instances, we will use a Trove-compatible Ubuntu image.
Build your own Trove images, Then run the following steps:
.. code-block:: bash
$ export DATASTORE_TYPE="mysql"
$ glance --os-username admin_trove_user --os-password trove --os-tenant-name trove_for_trove_usage --os-auth-url http://<KeystoneIP>:<KeystoneAdminPort>/v2.0 image-create --name trove-image --is-public True --container-format ovf --disk-format qcow2 --file ${DATASTORE_TYPE}.qcow2
At this step please remember the image ID or store it in an environment variable (IMAGEID).
.. code-block:: bash
$ glance --os-username trove --os-password trove --os-tenant-name trove --os-auth-url http://<KeystoneIP>:<KeystoneAdminPort>/v2.0
image-create --name trove-image --is-public true --container-format ovf --disk-format qcow2 --owner trove < precise.qcow2
$ export IMAGEID=<glance_image_id>
Cloud-init scripts
==================
-------------------
Cloud-init location
-------------------
By default, trove-taskmanager will look at /etc/trove/cloudinit for <datastore_manager>.cloudinit.
------------------
Cloud-init content
------------------
Each cloud-init script for Trove-compatible images should contain:
- Trove installation
Custom images with Trove code inside
====================================
*To be added*
Prepare the database
====================
Create the Trove database schema:
- Connect to the storage backend (MySQL, PostgreSQL)
- Create a database called `trove` (this database will be used for storing Trove ORM)
- Compose connection string. Example: mysql+pymysql://<user>:<password>@<backend_host>:<backend_port>/<database_name>
Initialize the database
=======================
Once the database for Trove is created, its structure needs to be populated.
.. code-block:: bash
$ trove-manage db_sync
Setup Trove Datastores
======================
---------
Datastore
---------
A Datastore is a data structure that describes a set of Datastore Versions, which consists of::
- ID -- simple auto-generated UUID
- Name -- user-defined attribute, actual name of a datastore
- Datastore Versions
Example::
- mysql, cassandra, redis, etc.
-----------------
Datastore Version
-----------------
A Datastore Version is a data structure that describes a version of a specific database pinned to datastore, which consists of::
- ID — Simple auto-generated UUID
- Datastore ID — Reference to Datastore
- Name — User-defined attribute, actual name of a database version
- Datastore manager — trove-guestagent manager that is used for datastore management
- Image ID — Reference to a specific Glance image ID
- Packages — Operating system specific packages that would be deployed onto datastore VM
- Active — Boolean flag that defines if version can be used for instance deployment or not
Example::
- ID - edb1d22a-b66d-4e86-be60-756240439272
- Datastore ID - 9c3d890b-a2f2-4ba5-91b2-2997d0791502
- Name - mysql-5.7
- Datastore manager - mysql
- Image ID - d73a402-3953-4721-8c99-86fc72e1cb51
- Packages - mysql-server=5.7, percona-xtrabackup=2.4
- Active - True
--------------------------------------------
Datastore and Datastore Version registration
--------------------------------------------
To register a datastore, you must execute:
.. code-block:: bash
$ export DATASTORE_TYPE="mysql" # available options: mysql, mongodb, postgresql, redis, cassandra, couchbase, couchdb, db2, vertica, etc.
$ export DATASTORE_VERSION="5.7" # available options: for cassandra 2.0.x, for mysql: 5.x, for mongodb: 2.x.x, etc.
$ export PACKAGES="mysql-server-5.7" # available options: cassandra=2.0.9, mongodb=2.0.4, etc
$ export IMAGEID="9910350b-77e3-4790-86be-b971d0cf9175" # Glance image ID of the relevant Datastore version (see Source images section)
$ trove-manage datastore_update ${DATASTORE_TYPE} ""
$ trove-manage datastore_version_update ${DATASTORE_TYPE} ${DATASTORE_VERSION} ${DATASTORE_TYPE} ${IMAGEID} ${PACKAGES} 1
$ trove-manage datastore_update ${DATASTORE_TYPE} ${DATASTORE_VERSION}
=========
Run Trove
=========
Trove services configuration and tuning
=======================================
*To be added*
Starting Trove services
=======================
Run trove-api:
.. code-block:: bash
$ trove-api --config-file=${TROVE_CONF_DIR}/trove.conf &
Run trove-taskmanager:
.. code-block:: bash
$ trove-taskmanager --config-file=${TROVE_CONF_DIR}/trove.conf &
Run trove-conductor:
.. code-block:: bash
$ trove-conductor --config-file=${TROVE_CONF_DIR}/trove.conf &
=================
Trove interaction
=================
Keystonerc
==========
You need to build a `keystonerc` file that contains data to simplify the auth processes while using the Trove client:
.. code-block:: bash
export OS_PROJECT_NAME=trove
export OS_USERNAME=regular_trove_user
export OS_PASSWORD=<UserPassword>
export OS_AUTH_URL="http://<KeystoneIP>/identity"
export OS_USER_DOMAIN_NAME=Default
export OS_PROJECT_DOMAIN_NAME=Default
Trove deployment verification
=============================
First you need to execute:
.. code-block:: bash
$ . keystonerc
To see `help` for a specific command:
.. code-block:: bash
$ trove help <command>
To create an instance:
.. code-block:: bash
$ trove create <name> <flavor>
[--size <size>] [--volume_type <volume_type>]
[--databases <database> [<database> ...]]
[--users <user:password> [<user:password> ...]]
[--backup <backup>]
[--availability_zone <availability_zone>]
[--datastore <datastore>]
[--datastore_version <datastore_version>]
[--nic <net-id=<net-uuid>,v4-fixed-ip=<ip-addr>,port-id=<port-uuid>>]
[--configuration <configuration>]
[--replica_of <source_instance>] [--replica_count <count>]
[--module <module>] [--locality <policy>]
===============
Troubleshooting
===============
No instance IPs in the output of 'trove show <instance_id>'
===========================================================
If the Trove instance was successfully created, is showing ACTIVE state and working, yet there is no IP address for the instance shown in the output of 'trove show <instance_id>, then confirm the following lines are added to trove.conf ::
network_label_regex = ^NETWORK_NAME$
where NETWORK_NAME should be replaced with real name of the network to which the instance is connected to.
To decide which network would you like to attach a Trove instance to, run the following command:
.. code-block:: bash
$ openstack network list
One possible way to find the network name is to execute the 'nova list' command. The output will list all OpenStack instances for the tenant, including network information. Look for ::
NETWORK_NAME=IP_ADDRESS
Additional information
======================
Additional information can be found in the OpenStack installation guide for the trove project. This document can be found under the "Installation Tutorials and Guides" section of the OpenStack Documentation.
For the current documentation, visit:
http://docs.openstack.org/index.html#install-guides
Select the link for "Installation Tutorials and Guides"
The installation guides for trove (the Database Service) can be found under the appropriate operating system.
If you are interested in documentation for a specific OpenStack release, visit:
http://docs.openstack.org/<release-code-name>/
For example, the documentation for the Pike release is found at:
http://docs.openstack.org/pike/
and the documentation for the Queens release is found at:
http://docs.openstack.org/queens/