openstack-manuals/doc/install-guide/section_trove-install.xml

250 lines
13 KiB
XML

<?xml version="1.0" encoding="UTF-8"?>
<section xmlns="http://docbook.org/ns/docbook"
xmlns:xi="http://www.w3.org/2001/XInclude"
xmlns:xlink="http://www.w3.org/1999/xlink"
version="5.0"
xml:id="trove-install">
<title>Install the Database service</title>
<para>This procedure installs the Database module on the controller
node.</para>
<formalpara>
<title>Prerequisites</title>
<para>This chapter assumes that you already have a working
OpenStack environment with at least the following components
installed: Compute, Image Service, Identity.</para>
</formalpara>
<itemizedlist>
<listitem>
<para>If you want to do backup and restore, you also need Object Storage.</para>
</listitem>
<listitem>
<para>If you want to provision datastores on block-storage volumes, you also need Block Storage.</para>
</listitem>
</itemizedlist>
<para>To install the Database module on the controller:</para>
<procedure>
<step>
<para>Install required packages:</para>
<screen os="ubuntu;debian"><prompt>#</prompt> <userinput>apt-get install python-trove python-troveclient python-glanceclient \
trove-common trove-api trove-taskmanager</userinput></screen>
<screen os="rhel;centos;fedora"><prompt>#</prompt> <userinput>yum install openstack-trove python-troveclient</userinput></screen>
<screen os="opensuse;sles"><prompt>#</prompt> <userinput>zypper install openstack-trove python-troveclient</userinput></screen>
</step>
<step os="debian">
<para>Respond to the prompts for <link linkend="debconf-dbconfig-common">database management</link> and <link linkend="debconf-keystone_authtoken"><literal>[keystone_authtoken]</literal> settings</link>,
and <link linkend="debconf-api-endpoints">API endpoint</link>
registration. The <command>trove-manage db_sync</command>
command runs automatically.</para>
</step>
<step>
<para>Prepare OpenStack:</para>
<substeps>
<step>
<para>Source the <filename>admin-openrc.sh</filename> file.</para>
<screen><prompt>$</prompt> <userinput>source ~/admin-openrc.sh</userinput></screen>
</step>
<step os="rhel;centos;fedora;opensuse;sles;ubuntu">
<para>Create a <literal>trove</literal> user that Compute uses to
authenticate with the Identity service. Use the
<literal>service</literal> tenant and give the user the
<literal>admin</literal> role:
</para>
<screen><prompt>$</prompt> <userinput>keystone user-create --name trove --pass <replaceable>TROVE_PASS</replaceable></userinput>
<prompt>$</prompt> <userinput>keystone user-role-add --user trove --tenant service --role admin</userinput></screen>
<para>Replace <replaceable>TROVE_PASS</replaceable> with a
suitable password.</para>
</step>
</substeps>
</step>
<step>
<para>Edit the following configuration files, taking the below
actions for each file:</para>
<itemizedlist>
<listitem><para><filename>trove.conf</filename></para></listitem>
<listitem><para><filename>trove-taskmanager.conf</filename></para></listitem>
<listitem><para><filename>trove-conductor.conf</filename></para></listitem>
</itemizedlist>
<substeps>
<step><para>Edit the <literal>[DEFAULT]</literal> section of
each file and set appropriate values for the OpenStack service
URLs, logging and messaging configuration, and SQL
connections:</para>
<programlisting language="ini">[DEFAULT]
log_dir = /var/log/trove
trove_auth_url = http://<replaceable>controller</replaceable>:5000/v2.0
nova_compute_url = http://<replaceable>controller</replaceable>:8774/v2
cinder_url = http://<replaceable>controller</replaceable>:8776/v1
swift_url = http://<replaceable>controller</replaceable>:8080/v1/AUTH_
sql_connection = mysql://trove:<literal>TROVE_DBPASS</literal>@<replaceable>controller</replaceable>/trove
notifier_queue_hostname = <replaceable>controller</replaceable></programlisting>
</step>
<step>
<para>Configure the Database module to use the RabbitMQ message broker by
setting the following options in the <literal>[DEFAULT]</literal>
configuration group of each file:</para>
<programlisting language="ini">[DEFAULT]
...
rpc_backend = rabbit
rabbit_host = <replaceable>controller</replaceable>
rabbit_password = <replaceable>RABBIT_PASS</replaceable></programlisting>
</step>
</substeps>
</step>
<step os="opensuse;sles;fedora;rhel;centos;ubuntu">
<para os="fedora;rhel;centos">Get the <filename>api-paste.ini</filename> file and save it to <filename>/etc/trove</filename>. You can get the file from this location:</para>
<para os="fedora;rhel;centos"><link xlink:href="http://git.openstack.org/cgit/openstack/trove/plain/etc/trove/api-paste.ini?h=juno-eol">http://git.openstack.org/cgit/openstack/trove/plain/etc/trove/api-paste.ini?h=juno-eol</link></para>
<para>Edit the <literal>[filter:authtoken]</literal> section
of the <filename>api-paste.ini</filename> file so it matches the
listing shown below:</para>
<programlisting language="ini">[filter:authtoken]
auth_uri = http://<replaceable>controller</replaceable>:5000/v2.0
identity_uri = http://<replaceable>controller</replaceable>:35357
admin_user = trove
admin_password = <replaceable>ADMIN_PASS</replaceable>
admin_tenant_name = service
signing_dir = /var/cache/trove</programlisting>
</step>
<step><para>Edit the <filename>trove.conf</filename> file so it includes
appropriate values for the default datastore, network label
regex, and API information as shown below:</para>
<programlisting language="ini">[DEFAULT]
default_datastore = mysql
....
# Config option for showing the IP address that nova doles out
add_addresses = True
network_label_regex = ^NETWORK_LABEL$
....
api_paste_config = /etc/trove/api-paste.ini</programlisting>
</step>
<step>
<para>Edit the <filename>trove-taskmanager.conf</filename> file
so it includes the required settings to
connect to the OpenStack Compute service as shown below:</para>
<programlisting language="ini">[DEFAULT]
....
# Configuration options for talking to nova via the novaclient.
# These options are for an admin user in your keystone config.
# It proxy's the token received from the user to send to nova via this admin users creds,
# basically acting like the client via that proxy token.
nova_proxy_admin_user = admin
nova_proxy_admin_pass = <replaceable>ADMIN_PASS</replaceable>
nova_proxy_admin_tenant_name = service
taskmanager_manager = trove.taskmanager.manager.Manager
...</programlisting>
</step>
<step os="opensuse;sles;fedora;rhel;centos;ubuntu">
<para>Prepare the trove admin database:</para>
<screen><prompt>$</prompt> <userinput>mysql -u root -p</userinput>
<prompt>mysql&gt;</prompt> <userinput>CREATE DATABASE trove;</userinput>
<prompt>mysql&gt;</prompt> <userinput>GRANT ALL PRIVILEGES ON trove.* TO trove@'localhost' \
IDENTIFIED BY '<replaceable>TROVE_DBPASS</replaceable>';</userinput>
<prompt>mysql&gt;</prompt> <userinput>GRANT ALL PRIVILEGES ON trove.* TO trove@'%' \
IDENTIFIED BY '<replaceable>TROVE_DBPASS</replaceable>';</userinput></screen>
</step>
<step os="opensuse;sles;fedora;rhel;centos;ubuntu">
<para>Prepare the Database service:</para>
<substeps>
<step>
<para>Initialize the database:</para>
<screen><prompt>#</prompt> <userinput>su -s /bin/sh -c "trove-manage db_sync" trove</userinput></screen>
</step>
<step>
<para>Create a datastore. You need to create a separate datastore for
each type of database you want to use, for example, MySQL, MongoDB, Cassandra.
This example shows you how to create a datastore for a MySQL database:</para>
<screen><prompt>#</prompt> <userinput>su -s /bin/sh -c "trove-manage datastore_update mysql ''" trove</userinput></screen>
</step>
</substeps>
</step>
<step os="debian">
<para>Create a datastore. You need to create a separate datastore for
each type of database you want to use, for example, MySQL, MongoDB, Cassandra.
This example shows you how to create a datastore for a MySQL database:</para>
<screen><prompt>#</prompt> <userinput>su -s /bin/sh -c "trove-manage datastore_update mysql ''" trove</userinput></screen>
</step>
<step>
<para>Create a trove image.</para>
<para>Create an image for the type of database you want to use,
for example, MySQL, MongoDB, Cassandra.</para>
<para>This image must have the trove guest agent installed, and
it must have the <filename>trove-guestagent.conf</filename> file
configured to connect to your OpenStack environment. To
correctly configure the
<filename>trove-guestagent.conf</filename> file, follow these steps
on the guest instance you are using to build your image:</para>
<substeps>
<step>
<para>Add the following lines to <filename>trove-guestagent.conf</filename>:</para>
<programlisting language="ini">rabbit_host = <replaceable>controller</replaceable>
rabbit_password = <replaceable>RABBIT_PASS</replaceable>
nova_proxy_admin_user = admin
nova_proxy_admin_pass = <replaceable>ADMIN_PASS</replaceable>
nova_proxy_admin_tenant_name = service
trove_auth_url = http://<replaceable>controller</replaceable>:35357/v2.0</programlisting>
</step>
</substeps>
</step>
<step>
<para>Update the datastore to use the new image, using the
<command>trove-manage</command> command.</para>
<para>This example shows you how to create a MySQL 5.5 datastore:</para>
<screen><prompt>#</prompt> <userinput>trove-manage --config-file /etc/trove/trove.conf datastore_version_update \
mysql mysql-5.5 mysql <replaceable>glance_image_ID</replaceable> mysql-server-5.5 1</userinput></screen>
</step>
<step os="rhel;centos;fedora;opensuse;sles;ubuntu">
<para>You must register the Database module with the Identity service so
that other OpenStack services can locate it. Register the
service and specify the endpoint:</para>
<screen><prompt>$</prompt> <userinput>keystone service-create --name trove --type database \
--description "OpenStack Database Service"</userinput>
<prompt>$</prompt> <userinput>keystone endpoint-create \
--service-id $(keystone service-list | awk '/ trove / {print $2}') \
--publicurl http://<replaceable>controller</replaceable>:8779/v1.0/%\(tenant_id\)s \
--internalurl http://<replaceable>controller</replaceable>:8779/v1.0/%\(tenant_id\)s \
--adminurl http://<replaceable>controller</replaceable>:8779/v1.0/%\(tenant_id\)s \
--region regionOne</userinput></screen>
</step>
<step os="ubuntu;debian">
<para>Restart the Database services:</para>
<screen><prompt>#</prompt> <userinput>service trove-api restart</userinput>
<prompt>#</prompt> <userinput>service trove-taskmanager restart</userinput>
<prompt>#</prompt> <userinput>service trove-conductor restart</userinput></screen>
</step>
<step os="rhel;centos;fedora;sles;opensuse">
<para>Start the Database services and configure them to start when the
system boots:</para>
<screen os="centos;rhel;fedora"><prompt>#</prompt> <userinput>systemctl enable openstack-trove-api.service openstack-trove-taskmanager.service \
openstack-trove-conductor.service</userinput>
<prompt>#</prompt> <userinput>systemctl start openstack-trove-api.service openstack-trove-taskmanager.service \
openstack-trove-conductor.service</userinput></screen>
<para os="sles">On SLES:</para>
<screen os="sles"><prompt>#</prompt> <userinput>service openstack-trove-api start</userinput>
<prompt>#</prompt> <userinput>service openstack-trove-taskmanager start</userinput>
<prompt>#</prompt> <userinput>service openstack-trove-conductor start</userinput>
<prompt>#</prompt> <userinput>chkconfig openstack-trove-api on</userinput>
<prompt>#</prompt> <userinput>chkconfig openstack-trove-taskmanager on</userinput>
<prompt>#</prompt> <userinput>chkconfig openstack-trove-conductor on</userinput></screen>
<para os="opensuse">On openSUSE:</para>
<screen os="opensuse"><prompt>#</prompt> <userinput>systemctl enable openstack-trove-api.service openstack-trove-taskmanager.service \
openstack-trove-conductor.service</userinput>
<prompt>#</prompt> <userinput>systemctl start openstack-trove-api.service openstack-trove-taskmanager.service \
openstack-trove-conductor.service</userinput></screen>
</step>
</procedure>
</section>