openstack-manuals/doc/install-guide/section_swift-controller-node.xml

205 lines
9.8 KiB
XML
Raw Normal View History

Update swift content I updated the swift content in the installation guide as follows: 1) Consolidated Identity service configuration into proxy service configuration section. 2) Changed initial example architecture to install the proxy service on the controller node instead of a separate node. However, I used wording that supports separate and/or multiple proxy services. 3) The steps to configure memcached break installations with horizon because the latter references 127.0.0.1 instead of the management network interface IP address. After consolidating the proxy service to a single instance on the controller node, it now references 127.0.0.1. 4) Changed initial example architecture to install two storage nodes rather than five storage nodes to reduce the barrier to entry for users that want to try swift. However, I used wording that supports additional storage nodes and also references the swift documentation for larger deployments. 5) Changed the initial example architecture to use only the management network instead of a separate storage/replication network because the original instructions didn't actually use the latter. Fully implementing a separate storage/replication network requires a considerably more complex configuration and eliminating it reduces the barrier to entry for users that want to try swift. However, I used wording that mentions a separate storage/replication network and also references the swift documentation for larger deployments. 6) The Ubuntu and RHEL/CentOS/Fedora packages include ancient upstream/example configuration files that contain deprecated/defunct options and lack options that enable useful features. Packages take a while to fix, so the instructions now pull the upstream configuration files. 7) Removed the steps to create disk labels and partitions on the storage nodes because a variety of methods exist and users should choose their own method. 8) Changed rsync to run as a standard service rather than use the xinetd wrapper on RHEL/CentOS/Fedora. 9) Separated account, container, and object ring creation steps into separate sections to improve clarity and reduce typos. 10) Reduced partition power from 18 to 10 during ring creation because the example architecture shouldn't scale to the higher value using the example architecture. 11) Changed storage node service management steps on Ubuntu to use 'swift-init' rather than manually starting a large number of services. 12) Modified file names and IDs to increase consistency with other content in the guide. I recommend backporting to Juno as this chapter wasn't functional in time for release. Also, I'm planning to improve the diagrams in a future patch. Change-Id: I26710efe16e8cb6ed1a20ebc70b3f6d5962536d1 Partial-Bug: #1389382 Implements: blueprint installation-guide-improvements backport: juno
2014-11-06 10:51:07 -06:00
<?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="swift-install-controller-node">
<title>Install and configure the controller node</title>
<para>This section describes how to install and configure the proxy
service that handles requests for the account, container, and object
services operating on the storage nodes. For simplicity, this
Update swift content I updated the swift content in the installation guide as follows: 1) Consolidated Identity service configuration into proxy service configuration section. 2) Changed initial example architecture to install the proxy service on the controller node instead of a separate node. However, I used wording that supports separate and/or multiple proxy services. 3) The steps to configure memcached break installations with horizon because the latter references 127.0.0.1 instead of the management network interface IP address. After consolidating the proxy service to a single instance on the controller node, it now references 127.0.0.1. 4) Changed initial example architecture to install two storage nodes rather than five storage nodes to reduce the barrier to entry for users that want to try swift. However, I used wording that supports additional storage nodes and also references the swift documentation for larger deployments. 5) Changed the initial example architecture to use only the management network instead of a separate storage/replication network because the original instructions didn't actually use the latter. Fully implementing a separate storage/replication network requires a considerably more complex configuration and eliminating it reduces the barrier to entry for users that want to try swift. However, I used wording that mentions a separate storage/replication network and also references the swift documentation for larger deployments. 6) The Ubuntu and RHEL/CentOS/Fedora packages include ancient upstream/example configuration files that contain deprecated/defunct options and lack options that enable useful features. Packages take a while to fix, so the instructions now pull the upstream configuration files. 7) Removed the steps to create disk labels and partitions on the storage nodes because a variety of methods exist and users should choose their own method. 8) Changed rsync to run as a standard service rather than use the xinetd wrapper on RHEL/CentOS/Fedora. 9) Separated account, container, and object ring creation steps into separate sections to improve clarity and reduce typos. 10) Reduced partition power from 18 to 10 during ring creation because the example architecture shouldn't scale to the higher value using the example architecture. 11) Changed storage node service management steps on Ubuntu to use 'swift-init' rather than manually starting a large number of services. 12) Modified file names and IDs to increase consistency with other content in the guide. I recommend backporting to Juno as this chapter wasn't functional in time for release. Also, I'm planning to improve the diagrams in a future patch. Change-Id: I26710efe16e8cb6ed1a20ebc70b3f6d5962536d1 Partial-Bug: #1389382 Implements: blueprint installation-guide-improvements backport: juno
2014-11-06 10:51:07 -06:00
guide installs and configures the proxy service on the controller node.
However, you can run the proxy service on any node with network
connectivity to the storage nodes. Additionally, you can install and
configure the proxy service on multiple nodes to increase performance
and redundancy. For more information, see the
<link xlink:href="http://docs.openstack.org/developer/swift/deployment_guide.html"
>Deployment Guide</link>.</para>
<procedure>
<title>To configure prerequisites</title>
<para>The proxy service relies on an authentication and authorization
mechanism such as the Identity service. However, unlike other services,
it also offers an internal mechanism that allows it to operate without
any other OpenStack services. However, for simplicity, this guide
references the Identity service in <xref linkend="ch_keystone"/>. Before
you configure the Object Storage service, you must create service
credentials and an API endpoint.</para>
Update swift content I updated the swift content in the installation guide as follows: 1) Consolidated Identity service configuration into proxy service configuration section. 2) Changed initial example architecture to install the proxy service on the controller node instead of a separate node. However, I used wording that supports separate and/or multiple proxy services. 3) The steps to configure memcached break installations with horizon because the latter references 127.0.0.1 instead of the management network interface IP address. After consolidating the proxy service to a single instance on the controller node, it now references 127.0.0.1. 4) Changed initial example architecture to install two storage nodes rather than five storage nodes to reduce the barrier to entry for users that want to try swift. However, I used wording that supports additional storage nodes and also references the swift documentation for larger deployments. 5) Changed the initial example architecture to use only the management network instead of a separate storage/replication network because the original instructions didn't actually use the latter. Fully implementing a separate storage/replication network requires a considerably more complex configuration and eliminating it reduces the barrier to entry for users that want to try swift. However, I used wording that mentions a separate storage/replication network and also references the swift documentation for larger deployments. 6) The Ubuntu and RHEL/CentOS/Fedora packages include ancient upstream/example configuration files that contain deprecated/defunct options and lack options that enable useful features. Packages take a while to fix, so the instructions now pull the upstream configuration files. 7) Removed the steps to create disk labels and partitions on the storage nodes because a variety of methods exist and users should choose their own method. 8) Changed rsync to run as a standard service rather than use the xinetd wrapper on RHEL/CentOS/Fedora. 9) Separated account, container, and object ring creation steps into separate sections to improve clarity and reduce typos. 10) Reduced partition power from 18 to 10 during ring creation because the example architecture shouldn't scale to the higher value using the example architecture. 11) Changed storage node service management steps on Ubuntu to use 'swift-init' rather than manually starting a large number of services. 12) Modified file names and IDs to increase consistency with other content in the guide. I recommend backporting to Juno as this chapter wasn't functional in time for release. Also, I'm planning to improve the diagrams in a future patch. Change-Id: I26710efe16e8cb6ed1a20ebc70b3f6d5962536d1 Partial-Bug: #1389382 Implements: blueprint installation-guide-improvements backport: juno
2014-11-06 10:51:07 -06:00
<note>
<para>The Object Storage service does not use a SQL database on
the controller node. Instead, it uses distributed SQLite databases
on each storage node.</para>
Update swift content I updated the swift content in the installation guide as follows: 1) Consolidated Identity service configuration into proxy service configuration section. 2) Changed initial example architecture to install the proxy service on the controller node instead of a separate node. However, I used wording that supports separate and/or multiple proxy services. 3) The steps to configure memcached break installations with horizon because the latter references 127.0.0.1 instead of the management network interface IP address. After consolidating the proxy service to a single instance on the controller node, it now references 127.0.0.1. 4) Changed initial example architecture to install two storage nodes rather than five storage nodes to reduce the barrier to entry for users that want to try swift. However, I used wording that supports additional storage nodes and also references the swift documentation for larger deployments. 5) Changed the initial example architecture to use only the management network instead of a separate storage/replication network because the original instructions didn't actually use the latter. Fully implementing a separate storage/replication network requires a considerably more complex configuration and eliminating it reduces the barrier to entry for users that want to try swift. However, I used wording that mentions a separate storage/replication network and also references the swift documentation for larger deployments. 6) The Ubuntu and RHEL/CentOS/Fedora packages include ancient upstream/example configuration files that contain deprecated/defunct options and lack options that enable useful features. Packages take a while to fix, so the instructions now pull the upstream configuration files. 7) Removed the steps to create disk labels and partitions on the storage nodes because a variety of methods exist and users should choose their own method. 8) Changed rsync to run as a standard service rather than use the xinetd wrapper on RHEL/CentOS/Fedora. 9) Separated account, container, and object ring creation steps into separate sections to improve clarity and reduce typos. 10) Reduced partition power from 18 to 10 during ring creation because the example architecture shouldn't scale to the higher value using the example architecture. 11) Changed storage node service management steps on Ubuntu to use 'swift-init' rather than manually starting a large number of services. 12) Modified file names and IDs to increase consistency with other content in the guide. I recommend backporting to Juno as this chapter wasn't functional in time for release. Also, I'm planning to improve the diagrams in a future patch. Change-Id: I26710efe16e8cb6ed1a20ebc70b3f6d5962536d1 Partial-Bug: #1389382 Implements: blueprint installation-guide-improvements backport: juno
2014-11-06 10:51:07 -06:00
</note>
<step>
<para>To create the Identity service credentials, complete these
steps:</para>
<substeps>
<step>
<para>Create the <literal>swift</literal> user:</para>
<screen><prompt>$</prompt> <userinput>openstack user create --password-prompt swift</userinput>
<computeroutput>User Password:
Repeat User Password:
Update swift content I updated the swift content in the installation guide as follows: 1) Consolidated Identity service configuration into proxy service configuration section. 2) Changed initial example architecture to install the proxy service on the controller node instead of a separate node. However, I used wording that supports separate and/or multiple proxy services. 3) The steps to configure memcached break installations with horizon because the latter references 127.0.0.1 instead of the management network interface IP address. After consolidating the proxy service to a single instance on the controller node, it now references 127.0.0.1. 4) Changed initial example architecture to install two storage nodes rather than five storage nodes to reduce the barrier to entry for users that want to try swift. However, I used wording that supports additional storage nodes and also references the swift documentation for larger deployments. 5) Changed the initial example architecture to use only the management network instead of a separate storage/replication network because the original instructions didn't actually use the latter. Fully implementing a separate storage/replication network requires a considerably more complex configuration and eliminating it reduces the barrier to entry for users that want to try swift. However, I used wording that mentions a separate storage/replication network and also references the swift documentation for larger deployments. 6) The Ubuntu and RHEL/CentOS/Fedora packages include ancient upstream/example configuration files that contain deprecated/defunct options and lack options that enable useful features. Packages take a while to fix, so the instructions now pull the upstream configuration files. 7) Removed the steps to create disk labels and partitions on the storage nodes because a variety of methods exist and users should choose their own method. 8) Changed rsync to run as a standard service rather than use the xinetd wrapper on RHEL/CentOS/Fedora. 9) Separated account, container, and object ring creation steps into separate sections to improve clarity and reduce typos. 10) Reduced partition power from 18 to 10 during ring creation because the example architecture shouldn't scale to the higher value using the example architecture. 11) Changed storage node service management steps on Ubuntu to use 'swift-init' rather than manually starting a large number of services. 12) Modified file names and IDs to increase consistency with other content in the guide. I recommend backporting to Juno as this chapter wasn't functional in time for release. Also, I'm planning to improve the diagrams in a future patch. Change-Id: I26710efe16e8cb6ed1a20ebc70b3f6d5962536d1 Partial-Bug: #1389382 Implements: blueprint installation-guide-improvements backport: juno
2014-11-06 10:51:07 -06:00
+----------+----------------------------------+
| Field | Value |
+----------+----------------------------------+
| email | None |
| enabled | True |
| id | d535e5cbd2b74ac7bfb97db9cced3ed6 |
| name | swift |
| username | swift |
Update swift content I updated the swift content in the installation guide as follows: 1) Consolidated Identity service configuration into proxy service configuration section. 2) Changed initial example architecture to install the proxy service on the controller node instead of a separate node. However, I used wording that supports separate and/or multiple proxy services. 3) The steps to configure memcached break installations with horizon because the latter references 127.0.0.1 instead of the management network interface IP address. After consolidating the proxy service to a single instance on the controller node, it now references 127.0.0.1. 4) Changed initial example architecture to install two storage nodes rather than five storage nodes to reduce the barrier to entry for users that want to try swift. However, I used wording that supports additional storage nodes and also references the swift documentation for larger deployments. 5) Changed the initial example architecture to use only the management network instead of a separate storage/replication network because the original instructions didn't actually use the latter. Fully implementing a separate storage/replication network requires a considerably more complex configuration and eliminating it reduces the barrier to entry for users that want to try swift. However, I used wording that mentions a separate storage/replication network and also references the swift documentation for larger deployments. 6) The Ubuntu and RHEL/CentOS/Fedora packages include ancient upstream/example configuration files that contain deprecated/defunct options and lack options that enable useful features. Packages take a while to fix, so the instructions now pull the upstream configuration files. 7) Removed the steps to create disk labels and partitions on the storage nodes because a variety of methods exist and users should choose their own method. 8) Changed rsync to run as a standard service rather than use the xinetd wrapper on RHEL/CentOS/Fedora. 9) Separated account, container, and object ring creation steps into separate sections to improve clarity and reduce typos. 10) Reduced partition power from 18 to 10 during ring creation because the example architecture shouldn't scale to the higher value using the example architecture. 11) Changed storage node service management steps on Ubuntu to use 'swift-init' rather than manually starting a large number of services. 12) Modified file names and IDs to increase consistency with other content in the guide. I recommend backporting to Juno as this chapter wasn't functional in time for release. Also, I'm planning to improve the diagrams in a future patch. Change-Id: I26710efe16e8cb6ed1a20ebc70b3f6d5962536d1 Partial-Bug: #1389382 Implements: blueprint installation-guide-improvements backport: juno
2014-11-06 10:51:07 -06:00
+----------+----------------------------------+</computeroutput></screen>
</step>
<step>
<para>Add the <literal>admin</literal> role to the
<literal>swift</literal> user:</para>
<screen><prompt>$</prompt> <userinput>openstack role add --project service --user swift admin</userinput>
<computeroutput>+-------+----------------------------------+
| Field | Value |
+-------+----------------------------------+
| id | cd2cb9a39e874ea69e5d4b896eb16128 |
| name | admin |
+-------+----------------------------------+</computeroutput></screen>
Update swift content I updated the swift content in the installation guide as follows: 1) Consolidated Identity service configuration into proxy service configuration section. 2) Changed initial example architecture to install the proxy service on the controller node instead of a separate node. However, I used wording that supports separate and/or multiple proxy services. 3) The steps to configure memcached break installations with horizon because the latter references 127.0.0.1 instead of the management network interface IP address. After consolidating the proxy service to a single instance on the controller node, it now references 127.0.0.1. 4) Changed initial example architecture to install two storage nodes rather than five storage nodes to reduce the barrier to entry for users that want to try swift. However, I used wording that supports additional storage nodes and also references the swift documentation for larger deployments. 5) Changed the initial example architecture to use only the management network instead of a separate storage/replication network because the original instructions didn't actually use the latter. Fully implementing a separate storage/replication network requires a considerably more complex configuration and eliminating it reduces the barrier to entry for users that want to try swift. However, I used wording that mentions a separate storage/replication network and also references the swift documentation for larger deployments. 6) The Ubuntu and RHEL/CentOS/Fedora packages include ancient upstream/example configuration files that contain deprecated/defunct options and lack options that enable useful features. Packages take a while to fix, so the instructions now pull the upstream configuration files. 7) Removed the steps to create disk labels and partitions on the storage nodes because a variety of methods exist and users should choose their own method. 8) Changed rsync to run as a standard service rather than use the xinetd wrapper on RHEL/CentOS/Fedora. 9) Separated account, container, and object ring creation steps into separate sections to improve clarity and reduce typos. 10) Reduced partition power from 18 to 10 during ring creation because the example architecture shouldn't scale to the higher value using the example architecture. 11) Changed storage node service management steps on Ubuntu to use 'swift-init' rather than manually starting a large number of services. 12) Modified file names and IDs to increase consistency with other content in the guide. I recommend backporting to Juno as this chapter wasn't functional in time for release. Also, I'm planning to improve the diagrams in a future patch. Change-Id: I26710efe16e8cb6ed1a20ebc70b3f6d5962536d1 Partial-Bug: #1389382 Implements: blueprint installation-guide-improvements backport: juno
2014-11-06 10:51:07 -06:00
</step>
<step>
<para>Create the <literal>swift</literal> service entity:</para>
<screen><prompt>$</prompt> <userinput>openstack service create --type object-store \
--description "OpenStack Object Storage" swift</userinput>
Update swift content I updated the swift content in the installation guide as follows: 1) Consolidated Identity service configuration into proxy service configuration section. 2) Changed initial example architecture to install the proxy service on the controller node instead of a separate node. However, I used wording that supports separate and/or multiple proxy services. 3) The steps to configure memcached break installations with horizon because the latter references 127.0.0.1 instead of the management network interface IP address. After consolidating the proxy service to a single instance on the controller node, it now references 127.0.0.1. 4) Changed initial example architecture to install two storage nodes rather than five storage nodes to reduce the barrier to entry for users that want to try swift. However, I used wording that supports additional storage nodes and also references the swift documentation for larger deployments. 5) Changed the initial example architecture to use only the management network instead of a separate storage/replication network because the original instructions didn't actually use the latter. Fully implementing a separate storage/replication network requires a considerably more complex configuration and eliminating it reduces the barrier to entry for users that want to try swift. However, I used wording that mentions a separate storage/replication network and also references the swift documentation for larger deployments. 6) The Ubuntu and RHEL/CentOS/Fedora packages include ancient upstream/example configuration files that contain deprecated/defunct options and lack options that enable useful features. Packages take a while to fix, so the instructions now pull the upstream configuration files. 7) Removed the steps to create disk labels and partitions on the storage nodes because a variety of methods exist and users should choose their own method. 8) Changed rsync to run as a standard service rather than use the xinetd wrapper on RHEL/CentOS/Fedora. 9) Separated account, container, and object ring creation steps into separate sections to improve clarity and reduce typos. 10) Reduced partition power from 18 to 10 during ring creation because the example architecture shouldn't scale to the higher value using the example architecture. 11) Changed storage node service management steps on Ubuntu to use 'swift-init' rather than manually starting a large number of services. 12) Modified file names and IDs to increase consistency with other content in the guide. I recommend backporting to Juno as this chapter wasn't functional in time for release. Also, I'm planning to improve the diagrams in a future patch. Change-Id: I26710efe16e8cb6ed1a20ebc70b3f6d5962536d1 Partial-Bug: #1389382 Implements: blueprint installation-guide-improvements backport: juno
2014-11-06 10:51:07 -06:00
<computeroutput>+-------------+----------------------------------+
| Field | Value |
Update swift content I updated the swift content in the installation guide as follows: 1) Consolidated Identity service configuration into proxy service configuration section. 2) Changed initial example architecture to install the proxy service on the controller node instead of a separate node. However, I used wording that supports separate and/or multiple proxy services. 3) The steps to configure memcached break installations with horizon because the latter references 127.0.0.1 instead of the management network interface IP address. After consolidating the proxy service to a single instance on the controller node, it now references 127.0.0.1. 4) Changed initial example architecture to install two storage nodes rather than five storage nodes to reduce the barrier to entry for users that want to try swift. However, I used wording that supports additional storage nodes and also references the swift documentation for larger deployments. 5) Changed the initial example architecture to use only the management network instead of a separate storage/replication network because the original instructions didn't actually use the latter. Fully implementing a separate storage/replication network requires a considerably more complex configuration and eliminating it reduces the barrier to entry for users that want to try swift. However, I used wording that mentions a separate storage/replication network and also references the swift documentation for larger deployments. 6) The Ubuntu and RHEL/CentOS/Fedora packages include ancient upstream/example configuration files that contain deprecated/defunct options and lack options that enable useful features. Packages take a while to fix, so the instructions now pull the upstream configuration files. 7) Removed the steps to create disk labels and partitions on the storage nodes because a variety of methods exist and users should choose their own method. 8) Changed rsync to run as a standard service rather than use the xinetd wrapper on RHEL/CentOS/Fedora. 9) Separated account, container, and object ring creation steps into separate sections to improve clarity and reduce typos. 10) Reduced partition power from 18 to 10 during ring creation because the example architecture shouldn't scale to the higher value using the example architecture. 11) Changed storage node service management steps on Ubuntu to use 'swift-init' rather than manually starting a large number of services. 12) Modified file names and IDs to increase consistency with other content in the guide. I recommend backporting to Juno as this chapter wasn't functional in time for release. Also, I'm planning to improve the diagrams in a future patch. Change-Id: I26710efe16e8cb6ed1a20ebc70b3f6d5962536d1 Partial-Bug: #1389382 Implements: blueprint installation-guide-improvements backport: juno
2014-11-06 10:51:07 -06:00
+-------------+----------------------------------+
| description | OpenStack Object Storage |
| enabled | True |
| id | 75ef509da2c340499d454ae96a2c5c34 |
| name | swift |
| type | object-store |
Update swift content I updated the swift content in the installation guide as follows: 1) Consolidated Identity service configuration into proxy service configuration section. 2) Changed initial example architecture to install the proxy service on the controller node instead of a separate node. However, I used wording that supports separate and/or multiple proxy services. 3) The steps to configure memcached break installations with horizon because the latter references 127.0.0.1 instead of the management network interface IP address. After consolidating the proxy service to a single instance on the controller node, it now references 127.0.0.1. 4) Changed initial example architecture to install two storage nodes rather than five storage nodes to reduce the barrier to entry for users that want to try swift. However, I used wording that supports additional storage nodes and also references the swift documentation for larger deployments. 5) Changed the initial example architecture to use only the management network instead of a separate storage/replication network because the original instructions didn't actually use the latter. Fully implementing a separate storage/replication network requires a considerably more complex configuration and eliminating it reduces the barrier to entry for users that want to try swift. However, I used wording that mentions a separate storage/replication network and also references the swift documentation for larger deployments. 6) The Ubuntu and RHEL/CentOS/Fedora packages include ancient upstream/example configuration files that contain deprecated/defunct options and lack options that enable useful features. Packages take a while to fix, so the instructions now pull the upstream configuration files. 7) Removed the steps to create disk labels and partitions on the storage nodes because a variety of methods exist and users should choose their own method. 8) Changed rsync to run as a standard service rather than use the xinetd wrapper on RHEL/CentOS/Fedora. 9) Separated account, container, and object ring creation steps into separate sections to improve clarity and reduce typos. 10) Reduced partition power from 18 to 10 during ring creation because the example architecture shouldn't scale to the higher value using the example architecture. 11) Changed storage node service management steps on Ubuntu to use 'swift-init' rather than manually starting a large number of services. 12) Modified file names and IDs to increase consistency with other content in the guide. I recommend backporting to Juno as this chapter wasn't functional in time for release. Also, I'm planning to improve the diagrams in a future patch. Change-Id: I26710efe16e8cb6ed1a20ebc70b3f6d5962536d1 Partial-Bug: #1389382 Implements: blueprint installation-guide-improvements backport: juno
2014-11-06 10:51:07 -06:00
+-------------+----------------------------------+</computeroutput></screen>
</step>
</substeps>
</step>
<step>
<para>Create the Object Storage service API endpoint:</para>
<screen><prompt>$</prompt> <userinput>openstack endpoint create \
Update swift content I updated the swift content in the installation guide as follows: 1) Consolidated Identity service configuration into proxy service configuration section. 2) Changed initial example architecture to install the proxy service on the controller node instead of a separate node. However, I used wording that supports separate and/or multiple proxy services. 3) The steps to configure memcached break installations with horizon because the latter references 127.0.0.1 instead of the management network interface IP address. After consolidating the proxy service to a single instance on the controller node, it now references 127.0.0.1. 4) Changed initial example architecture to install two storage nodes rather than five storage nodes to reduce the barrier to entry for users that want to try swift. However, I used wording that supports additional storage nodes and also references the swift documentation for larger deployments. 5) Changed the initial example architecture to use only the management network instead of a separate storage/replication network because the original instructions didn't actually use the latter. Fully implementing a separate storage/replication network requires a considerably more complex configuration and eliminating it reduces the barrier to entry for users that want to try swift. However, I used wording that mentions a separate storage/replication network and also references the swift documentation for larger deployments. 6) The Ubuntu and RHEL/CentOS/Fedora packages include ancient upstream/example configuration files that contain deprecated/defunct options and lack options that enable useful features. Packages take a while to fix, so the instructions now pull the upstream configuration files. 7) Removed the steps to create disk labels and partitions on the storage nodes because a variety of methods exist and users should choose their own method. 8) Changed rsync to run as a standard service rather than use the xinetd wrapper on RHEL/CentOS/Fedora. 9) Separated account, container, and object ring creation steps into separate sections to improve clarity and reduce typos. 10) Reduced partition power from 18 to 10 during ring creation because the example architecture shouldn't scale to the higher value using the example architecture. 11) Changed storage node service management steps on Ubuntu to use 'swift-init' rather than manually starting a large number of services. 12) Modified file names and IDs to increase consistency with other content in the guide. I recommend backporting to Juno as this chapter wasn't functional in time for release. Also, I'm planning to improve the diagrams in a future patch. Change-Id: I26710efe16e8cb6ed1a20ebc70b3f6d5962536d1 Partial-Bug: #1389382 Implements: blueprint installation-guide-improvements backport: juno
2014-11-06 10:51:07 -06:00
--publicurl 'http://<replaceable>controller</replaceable>:8080/v1/AUTH_%(tenant_id)s' \
--internalurl 'http://<replaceable>controller</replaceable>:8080/v1/AUTH_%(tenant_id)s' \
--adminurl http://<replaceable>controller</replaceable>:8080 \
--region RegionOne \
object-store</userinput>
<computeroutput>+--------------+----------------------------------------------+
| Field | Value |
+--------------+----------------------------------------------+
| adminurl | http://controller:8080/ |
| id | af534fb8b7ff40a6acf725437c586ebe |
| internalurl | http://controller:8080/v1/AUTH_%(tenant_id)s |
| publicurl | http://controller:8080/v1/AUTH_%(tenant_id)s |
| region | RegionOne |
| service_id | 75ef509da2c340499d454ae96a2c5c34 |
| service_name | swift |
| service_type | object-store |
+--------------+----------------------------------------------+</computeroutput></screen>
Update swift content I updated the swift content in the installation guide as follows: 1) Consolidated Identity service configuration into proxy service configuration section. 2) Changed initial example architecture to install the proxy service on the controller node instead of a separate node. However, I used wording that supports separate and/or multiple proxy services. 3) The steps to configure memcached break installations with horizon because the latter references 127.0.0.1 instead of the management network interface IP address. After consolidating the proxy service to a single instance on the controller node, it now references 127.0.0.1. 4) Changed initial example architecture to install two storage nodes rather than five storage nodes to reduce the barrier to entry for users that want to try swift. However, I used wording that supports additional storage nodes and also references the swift documentation for larger deployments. 5) Changed the initial example architecture to use only the management network instead of a separate storage/replication network because the original instructions didn't actually use the latter. Fully implementing a separate storage/replication network requires a considerably more complex configuration and eliminating it reduces the barrier to entry for users that want to try swift. However, I used wording that mentions a separate storage/replication network and also references the swift documentation for larger deployments. 6) The Ubuntu and RHEL/CentOS/Fedora packages include ancient upstream/example configuration files that contain deprecated/defunct options and lack options that enable useful features. Packages take a while to fix, so the instructions now pull the upstream configuration files. 7) Removed the steps to create disk labels and partitions on the storage nodes because a variety of methods exist and users should choose their own method. 8) Changed rsync to run as a standard service rather than use the xinetd wrapper on RHEL/CentOS/Fedora. 9) Separated account, container, and object ring creation steps into separate sections to improve clarity and reduce typos. 10) Reduced partition power from 18 to 10 during ring creation because the example architecture shouldn't scale to the higher value using the example architecture. 11) Changed storage node service management steps on Ubuntu to use 'swift-init' rather than manually starting a large number of services. 12) Modified file names and IDs to increase consistency with other content in the guide. I recommend backporting to Juno as this chapter wasn't functional in time for release. Also, I'm planning to improve the diagrams in a future patch. Change-Id: I26710efe16e8cb6ed1a20ebc70b3f6d5962536d1 Partial-Bug: #1389382 Implements: blueprint installation-guide-improvements backport: juno
2014-11-06 10:51:07 -06:00
</step>
</procedure>
<procedure>
<title>To install and configure the controller node components</title>
<note>
<para>Default configuration files vary by distribution. You might need
to add these sections and options rather than modifying existing
sections and options. Also, an ellipsis (...) in the configuration
snippets indicates potential default configuration options that you
should retain.</para>
</note>
Update swift content I updated the swift content in the installation guide as follows: 1) Consolidated Identity service configuration into proxy service configuration section. 2) Changed initial example architecture to install the proxy service on the controller node instead of a separate node. However, I used wording that supports separate and/or multiple proxy services. 3) The steps to configure memcached break installations with horizon because the latter references 127.0.0.1 instead of the management network interface IP address. After consolidating the proxy service to a single instance on the controller node, it now references 127.0.0.1. 4) Changed initial example architecture to install two storage nodes rather than five storage nodes to reduce the barrier to entry for users that want to try swift. However, I used wording that supports additional storage nodes and also references the swift documentation for larger deployments. 5) Changed the initial example architecture to use only the management network instead of a separate storage/replication network because the original instructions didn't actually use the latter. Fully implementing a separate storage/replication network requires a considerably more complex configuration and eliminating it reduces the barrier to entry for users that want to try swift. However, I used wording that mentions a separate storage/replication network and also references the swift documentation for larger deployments. 6) The Ubuntu and RHEL/CentOS/Fedora packages include ancient upstream/example configuration files that contain deprecated/defunct options and lack options that enable useful features. Packages take a while to fix, so the instructions now pull the upstream configuration files. 7) Removed the steps to create disk labels and partitions on the storage nodes because a variety of methods exist and users should choose their own method. 8) Changed rsync to run as a standard service rather than use the xinetd wrapper on RHEL/CentOS/Fedora. 9) Separated account, container, and object ring creation steps into separate sections to improve clarity and reduce typos. 10) Reduced partition power from 18 to 10 during ring creation because the example architecture shouldn't scale to the higher value using the example architecture. 11) Changed storage node service management steps on Ubuntu to use 'swift-init' rather than manually starting a large number of services. 12) Modified file names and IDs to increase consistency with other content in the guide. I recommend backporting to Juno as this chapter wasn't functional in time for release. Also, I'm planning to improve the diagrams in a future patch. Change-Id: I26710efe16e8cb6ed1a20ebc70b3f6d5962536d1 Partial-Bug: #1389382 Implements: blueprint installation-guide-improvements backport: juno
2014-11-06 10:51:07 -06:00
<step>
<para>Install the packages:</para>
<note>
<para>Complete OpenStack environments already include some of these
packages.</para>
</note>
<screen os="ubuntu;debian"><prompt>#</prompt> <userinput>apt-get install swift swift-proxy python-swiftclient python-keystoneclient \
python-keystonemiddleware memcached</userinput></screen>
<screen os="rhel;centos;fedora"><prompt>#</prompt> <userinput>yum install openstack-swift-proxy python-swiftclient python-keystone-auth-token \
python-keystonemiddleware memcached</userinput></screen>
<screen os="sles;opensuse"><prompt>#</prompt> <userinput>zypper install openstack-swift-proxy python-swiftclient python-keystoneclient \
python-keystonemiddleware python-xml memcached</userinput></screen>
Update swift content I updated the swift content in the installation guide as follows: 1) Consolidated Identity service configuration into proxy service configuration section. 2) Changed initial example architecture to install the proxy service on the controller node instead of a separate node. However, I used wording that supports separate and/or multiple proxy services. 3) The steps to configure memcached break installations with horizon because the latter references 127.0.0.1 instead of the management network interface IP address. After consolidating the proxy service to a single instance on the controller node, it now references 127.0.0.1. 4) Changed initial example architecture to install two storage nodes rather than five storage nodes to reduce the barrier to entry for users that want to try swift. However, I used wording that supports additional storage nodes and also references the swift documentation for larger deployments. 5) Changed the initial example architecture to use only the management network instead of a separate storage/replication network because the original instructions didn't actually use the latter. Fully implementing a separate storage/replication network requires a considerably more complex configuration and eliminating it reduces the barrier to entry for users that want to try swift. However, I used wording that mentions a separate storage/replication network and also references the swift documentation for larger deployments. 6) The Ubuntu and RHEL/CentOS/Fedora packages include ancient upstream/example configuration files that contain deprecated/defunct options and lack options that enable useful features. Packages take a while to fix, so the instructions now pull the upstream configuration files. 7) Removed the steps to create disk labels and partitions on the storage nodes because a variety of methods exist and users should choose their own method. 8) Changed rsync to run as a standard service rather than use the xinetd wrapper on RHEL/CentOS/Fedora. 9) Separated account, container, and object ring creation steps into separate sections to improve clarity and reduce typos. 10) Reduced partition power from 18 to 10 during ring creation because the example architecture shouldn't scale to the higher value using the example architecture. 11) Changed storage node service management steps on Ubuntu to use 'swift-init' rather than manually starting a large number of services. 12) Modified file names and IDs to increase consistency with other content in the guide. I recommend backporting to Juno as this chapter wasn't functional in time for release. Also, I'm planning to improve the diagrams in a future patch. Change-Id: I26710efe16e8cb6ed1a20ebc70b3f6d5962536d1 Partial-Bug: #1389382 Implements: blueprint installation-guide-improvements backport: juno
2014-11-06 10:51:07 -06:00
</step>
<step os="ubuntu;debian">
<para>Create the <literal>/etc/swift</literal> directory.</para>
</step>
<step os="ubuntu;debian;rhel;centos;fedora">
<para>Obtain the proxy service configuration file from the Object
Storage source repository:</para>
<screen><prompt>#</prompt> <userinput>curl -o /etc/swift/proxy-server.conf \
https://git.openstack.org/cgit/openstack/swift/plain/etc/proxy-server.conf-sample?h=stable/kilo</userinput></screen>
Update swift content I updated the swift content in the installation guide as follows: 1) Consolidated Identity service configuration into proxy service configuration section. 2) Changed initial example architecture to install the proxy service on the controller node instead of a separate node. However, I used wording that supports separate and/or multiple proxy services. 3) The steps to configure memcached break installations with horizon because the latter references 127.0.0.1 instead of the management network interface IP address. After consolidating the proxy service to a single instance on the controller node, it now references 127.0.0.1. 4) Changed initial example architecture to install two storage nodes rather than five storage nodes to reduce the barrier to entry for users that want to try swift. However, I used wording that supports additional storage nodes and also references the swift documentation for larger deployments. 5) Changed the initial example architecture to use only the management network instead of a separate storage/replication network because the original instructions didn't actually use the latter. Fully implementing a separate storage/replication network requires a considerably more complex configuration and eliminating it reduces the barrier to entry for users that want to try swift. However, I used wording that mentions a separate storage/replication network and also references the swift documentation for larger deployments. 6) The Ubuntu and RHEL/CentOS/Fedora packages include ancient upstream/example configuration files that contain deprecated/defunct options and lack options that enable useful features. Packages take a while to fix, so the instructions now pull the upstream configuration files. 7) Removed the steps to create disk labels and partitions on the storage nodes because a variety of methods exist and users should choose their own method. 8) Changed rsync to run as a standard service rather than use the xinetd wrapper on RHEL/CentOS/Fedora. 9) Separated account, container, and object ring creation steps into separate sections to improve clarity and reduce typos. 10) Reduced partition power from 18 to 10 during ring creation because the example architecture shouldn't scale to the higher value using the example architecture. 11) Changed storage node service management steps on Ubuntu to use 'swift-init' rather than manually starting a large number of services. 12) Modified file names and IDs to increase consistency with other content in the guide. I recommend backporting to Juno as this chapter wasn't functional in time for release. Also, I'm planning to improve the diagrams in a future patch. Change-Id: I26710efe16e8cb6ed1a20ebc70b3f6d5962536d1 Partial-Bug: #1389382 Implements: blueprint installation-guide-improvements backport: juno
2014-11-06 10:51:07 -06:00
</step>
<step>
<para>Edit the <filename>/etc/swift/proxy-server.conf</filename>
file and complete the following actions:</para>
<substeps>
<step>
<para>In the <literal>[DEFAULT]</literal> section, configure
the bind port, user, and configuration directory:</para>
<programlisting language="ini">[DEFAULT]
...
bind_port = 8080
user = swift
swift_dir = /etc/swift</programlisting>
</step>
<step>
<para>In the <literal>[pipeline:main]</literal> section, enable
Update swift content I updated the swift content in the installation guide as follows: 1) Consolidated Identity service configuration into proxy service configuration section. 2) Changed initial example architecture to install the proxy service on the controller node instead of a separate node. However, I used wording that supports separate and/or multiple proxy services. 3) The steps to configure memcached break installations with horizon because the latter references 127.0.0.1 instead of the management network interface IP address. After consolidating the proxy service to a single instance on the controller node, it now references 127.0.0.1. 4) Changed initial example architecture to install two storage nodes rather than five storage nodes to reduce the barrier to entry for users that want to try swift. However, I used wording that supports additional storage nodes and also references the swift documentation for larger deployments. 5) Changed the initial example architecture to use only the management network instead of a separate storage/replication network because the original instructions didn't actually use the latter. Fully implementing a separate storage/replication network requires a considerably more complex configuration and eliminating it reduces the barrier to entry for users that want to try swift. However, I used wording that mentions a separate storage/replication network and also references the swift documentation for larger deployments. 6) The Ubuntu and RHEL/CentOS/Fedora packages include ancient upstream/example configuration files that contain deprecated/defunct options and lack options that enable useful features. Packages take a while to fix, so the instructions now pull the upstream configuration files. 7) Removed the steps to create disk labels and partitions on the storage nodes because a variety of methods exist and users should choose their own method. 8) Changed rsync to run as a standard service rather than use the xinetd wrapper on RHEL/CentOS/Fedora. 9) Separated account, container, and object ring creation steps into separate sections to improve clarity and reduce typos. 10) Reduced partition power from 18 to 10 during ring creation because the example architecture shouldn't scale to the higher value using the example architecture. 11) Changed storage node service management steps on Ubuntu to use 'swift-init' rather than manually starting a large number of services. 12) Modified file names and IDs to increase consistency with other content in the guide. I recommend backporting to Juno as this chapter wasn't functional in time for release. Also, I'm planning to improve the diagrams in a future patch. Change-Id: I26710efe16e8cb6ed1a20ebc70b3f6d5962536d1 Partial-Bug: #1389382 Implements: blueprint installation-guide-improvements backport: juno
2014-11-06 10:51:07 -06:00
the appropriate modules:</para>
<programlisting language="ini">[pipeline:main]
pipeline = catch_errors gatekeeper healthcheck proxy-logging cache container_sync bulk ratelimit authtoken keystoneauth container-quotas account-quotas slo dlo proxy-logging proxy-server</programlisting>
Update swift content I updated the swift content in the installation guide as follows: 1) Consolidated Identity service configuration into proxy service configuration section. 2) Changed initial example architecture to install the proxy service on the controller node instead of a separate node. However, I used wording that supports separate and/or multiple proxy services. 3) The steps to configure memcached break installations with horizon because the latter references 127.0.0.1 instead of the management network interface IP address. After consolidating the proxy service to a single instance on the controller node, it now references 127.0.0.1. 4) Changed initial example architecture to install two storage nodes rather than five storage nodes to reduce the barrier to entry for users that want to try swift. However, I used wording that supports additional storage nodes and also references the swift documentation for larger deployments. 5) Changed the initial example architecture to use only the management network instead of a separate storage/replication network because the original instructions didn't actually use the latter. Fully implementing a separate storage/replication network requires a considerably more complex configuration and eliminating it reduces the barrier to entry for users that want to try swift. However, I used wording that mentions a separate storage/replication network and also references the swift documentation for larger deployments. 6) The Ubuntu and RHEL/CentOS/Fedora packages include ancient upstream/example configuration files that contain deprecated/defunct options and lack options that enable useful features. Packages take a while to fix, so the instructions now pull the upstream configuration files. 7) Removed the steps to create disk labels and partitions on the storage nodes because a variety of methods exist and users should choose their own method. 8) Changed rsync to run as a standard service rather than use the xinetd wrapper on RHEL/CentOS/Fedora. 9) Separated account, container, and object ring creation steps into separate sections to improve clarity and reduce typos. 10) Reduced partition power from 18 to 10 during ring creation because the example architecture shouldn't scale to the higher value using the example architecture. 11) Changed storage node service management steps on Ubuntu to use 'swift-init' rather than manually starting a large number of services. 12) Modified file names and IDs to increase consistency with other content in the guide. I recommend backporting to Juno as this chapter wasn't functional in time for release. Also, I'm planning to improve the diagrams in a future patch. Change-Id: I26710efe16e8cb6ed1a20ebc70b3f6d5962536d1 Partial-Bug: #1389382 Implements: blueprint installation-guide-improvements backport: juno
2014-11-06 10:51:07 -06:00
<note>
<para>For more information on other modules that enable
additional features, see the
<link xlink:href="http://docs.openstack.org/developer/swift/deployment_guide.html"
>Deployment Guide</link>.</para>
</note>
</step>
<step>
<para>In the <literal>[app:proxy-server]</literal> section, enable
automatic account creation:</para>
Update swift content I updated the swift content in the installation guide as follows: 1) Consolidated Identity service configuration into proxy service configuration section. 2) Changed initial example architecture to install the proxy service on the controller node instead of a separate node. However, I used wording that supports separate and/or multiple proxy services. 3) The steps to configure memcached break installations with horizon because the latter references 127.0.0.1 instead of the management network interface IP address. After consolidating the proxy service to a single instance on the controller node, it now references 127.0.0.1. 4) Changed initial example architecture to install two storage nodes rather than five storage nodes to reduce the barrier to entry for users that want to try swift. However, I used wording that supports additional storage nodes and also references the swift documentation for larger deployments. 5) Changed the initial example architecture to use only the management network instead of a separate storage/replication network because the original instructions didn't actually use the latter. Fully implementing a separate storage/replication network requires a considerably more complex configuration and eliminating it reduces the barrier to entry for users that want to try swift. However, I used wording that mentions a separate storage/replication network and also references the swift documentation for larger deployments. 6) The Ubuntu and RHEL/CentOS/Fedora packages include ancient upstream/example configuration files that contain deprecated/defunct options and lack options that enable useful features. Packages take a while to fix, so the instructions now pull the upstream configuration files. 7) Removed the steps to create disk labels and partitions on the storage nodes because a variety of methods exist and users should choose their own method. 8) Changed rsync to run as a standard service rather than use the xinetd wrapper on RHEL/CentOS/Fedora. 9) Separated account, container, and object ring creation steps into separate sections to improve clarity and reduce typos. 10) Reduced partition power from 18 to 10 during ring creation because the example architecture shouldn't scale to the higher value using the example architecture. 11) Changed storage node service management steps on Ubuntu to use 'swift-init' rather than manually starting a large number of services. 12) Modified file names and IDs to increase consistency with other content in the guide. I recommend backporting to Juno as this chapter wasn't functional in time for release. Also, I'm planning to improve the diagrams in a future patch. Change-Id: I26710efe16e8cb6ed1a20ebc70b3f6d5962536d1 Partial-Bug: #1389382 Implements: blueprint installation-guide-improvements backport: juno
2014-11-06 10:51:07 -06:00
<programlisting language="ini">[app:proxy-server]
...
account_autocreate = true</programlisting>
</step>
<step>
<para>In the <literal>[filter:keystoneauth]</literal> section,
configure the operator roles:</para>
<programlisting language="ini">[filter:keystoneauth]
use = egg:swift#keystoneauth
...
operator_roles = admin,_member_</programlisting>
</step>
<step>
<para>In the <literal>[filter:authtoken]</literal> section,
configure Identity service access:</para>
<programlisting language="ini">[filter:authtoken]
paste.filter_factory = keystonemiddleware.auth_token:filter_factory
...
auth_uri = http://<replaceable>controller</replaceable>:5000
auth_url = http://<replaceable>controller</replaceable>:35357
auth_plugin = password
project_domain_id = default
user_domain_id = default
project_name = service
username = swift
password = <replaceable>SWIFT_PASS</replaceable>
Update swift content I updated the swift content in the installation guide as follows: 1) Consolidated Identity service configuration into proxy service configuration section. 2) Changed initial example architecture to install the proxy service on the controller node instead of a separate node. However, I used wording that supports separate and/or multiple proxy services. 3) The steps to configure memcached break installations with horizon because the latter references 127.0.0.1 instead of the management network interface IP address. After consolidating the proxy service to a single instance on the controller node, it now references 127.0.0.1. 4) Changed initial example architecture to install two storage nodes rather than five storage nodes to reduce the barrier to entry for users that want to try swift. However, I used wording that supports additional storage nodes and also references the swift documentation for larger deployments. 5) Changed the initial example architecture to use only the management network instead of a separate storage/replication network because the original instructions didn't actually use the latter. Fully implementing a separate storage/replication network requires a considerably more complex configuration and eliminating it reduces the barrier to entry for users that want to try swift. However, I used wording that mentions a separate storage/replication network and also references the swift documentation for larger deployments. 6) The Ubuntu and RHEL/CentOS/Fedora packages include ancient upstream/example configuration files that contain deprecated/defunct options and lack options that enable useful features. Packages take a while to fix, so the instructions now pull the upstream configuration files. 7) Removed the steps to create disk labels and partitions on the storage nodes because a variety of methods exist and users should choose their own method. 8) Changed rsync to run as a standard service rather than use the xinetd wrapper on RHEL/CentOS/Fedora. 9) Separated account, container, and object ring creation steps into separate sections to improve clarity and reduce typos. 10) Reduced partition power from 18 to 10 during ring creation because the example architecture shouldn't scale to the higher value using the example architecture. 11) Changed storage node service management steps on Ubuntu to use 'swift-init' rather than manually starting a large number of services. 12) Modified file names and IDs to increase consistency with other content in the guide. I recommend backporting to Juno as this chapter wasn't functional in time for release. Also, I'm planning to improve the diagrams in a future patch. Change-Id: I26710efe16e8cb6ed1a20ebc70b3f6d5962536d1 Partial-Bug: #1389382 Implements: blueprint installation-guide-improvements backport: juno
2014-11-06 10:51:07 -06:00
delay_auth_decision = true</programlisting>
<para>Replace <replaceable>SWIFT_PASS</replaceable> with the
password you chose for the <literal>swift</literal> user in the
Identity service.</para>
<note>
<para>Comment out or remove any other options in the
<literal>[filter:authtoken]</literal> section.</para>
Update swift content I updated the swift content in the installation guide as follows: 1) Consolidated Identity service configuration into proxy service configuration section. 2) Changed initial example architecture to install the proxy service on the controller node instead of a separate node. However, I used wording that supports separate and/or multiple proxy services. 3) The steps to configure memcached break installations with horizon because the latter references 127.0.0.1 instead of the management network interface IP address. After consolidating the proxy service to a single instance on the controller node, it now references 127.0.0.1. 4) Changed initial example architecture to install two storage nodes rather than five storage nodes to reduce the barrier to entry for users that want to try swift. However, I used wording that supports additional storage nodes and also references the swift documentation for larger deployments. 5) Changed the initial example architecture to use only the management network instead of a separate storage/replication network because the original instructions didn't actually use the latter. Fully implementing a separate storage/replication network requires a considerably more complex configuration and eliminating it reduces the barrier to entry for users that want to try swift. However, I used wording that mentions a separate storage/replication network and also references the swift documentation for larger deployments. 6) The Ubuntu and RHEL/CentOS/Fedora packages include ancient upstream/example configuration files that contain deprecated/defunct options and lack options that enable useful features. Packages take a while to fix, so the instructions now pull the upstream configuration files. 7) Removed the steps to create disk labels and partitions on the storage nodes because a variety of methods exist and users should choose their own method. 8) Changed rsync to run as a standard service rather than use the xinetd wrapper on RHEL/CentOS/Fedora. 9) Separated account, container, and object ring creation steps into separate sections to improve clarity and reduce typos. 10) Reduced partition power from 18 to 10 during ring creation because the example architecture shouldn't scale to the higher value using the example architecture. 11) Changed storage node service management steps on Ubuntu to use 'swift-init' rather than manually starting a large number of services. 12) Modified file names and IDs to increase consistency with other content in the guide. I recommend backporting to Juno as this chapter wasn't functional in time for release. Also, I'm planning to improve the diagrams in a future patch. Change-Id: I26710efe16e8cb6ed1a20ebc70b3f6d5962536d1 Partial-Bug: #1389382 Implements: blueprint installation-guide-improvements backport: juno
2014-11-06 10:51:07 -06:00
</note>
</step>
<step>
<para>In the <literal>[filter:cache]</literal> section, configure
the <application>memcached</application> location:</para>
<programlisting language="ini">[filter:cache]
...
memcache_servers = 127.0.0.1:11211</programlisting>
</step>
</substeps>
</step>
</procedure>
</section>