Merge "Clean up of section_architecture_storage_focus.xml"

This commit is contained in:
Jenkins 2015-08-17 06:10:57 +00:00 committed by Gerrit Code Review
commit 967be3b56d

View File

@ -9,7 +9,7 @@
version="5.0" version="5.0"
xml:id="architecture-storage-hardware"> xml:id="architecture-storage-hardware">
<title>Architecture</title> <title>Architecture</title>
<para>There are three areas to consider when selecting storage hardware:</para> <para>Consider the following factors when selecting storage hardware:</para>
<itemizedlist> <itemizedlist>
<listitem> <listitem>
<para>Cost</para> <para>Cost</para>
@ -21,15 +21,15 @@
<para>Reliability</para> <para>Reliability</para>
</listitem> </listitem>
</itemizedlist> </itemizedlist>
<para>Storage-focused OpenStack clouds must reflect that the workloads are storage <para>Storage-focused OpenStack clouds must address I/O
intensive. These workloads are not compute intensive, nor are intensive workloads. These workloads are not CPU intensive,
they consistently network intensive. The network may be nor are they consistently network intensive. The network may be
heavily utilized to transfer storage, but they are not heavily utilized to transfer storage, but they are not
otherwise network intensive.</para> otherwise network intensive.</para>
<para>For a storage-focused OpenStack design architecture, the <para>The selection of storage hardware determines the overall
selection of storage hardware determines the overall performance and scalability of a storage-focused OpenStack design
performance and scalability of the design architecture. Several factors architecture.
impact the design process:</para> Several factors impact the design process, including:</para>
<variablelist> <variablelist>
<varlistentry> <varlistentry>
<term>Cost</term> <term>Cost</term>
@ -57,50 +57,35 @@
that performs well as it expands.</para> that performs well as it expands.</para>
</listitem> </listitem>
</varlistentry> </varlistentry>
<varlistentry>
<term>Expandability</term>
<listitem>
<para>Expandability is the overall ability of the solution
to grow. A storage solution that expands to 50 PB is
more expandable than a solution that only scales to 10 PB.</para>
<note>
<para>This meter is related to scalability.
</para>
</note>
</listitem>
</varlistentry>
</variablelist> </variablelist>
<para>Latency is a key consideration in a <para>Latency is a key consideration in a storage-focused OpenStack cloud.
storage-focused OpenStack cloud. Using solid-state disks Using solid-state disks (SSDs) to minimize latency and, to reduce CPU
(SSDs) to minimize latency for instance storage, and to reduce delays caused by waiting for the storage, increases performance. Use RAID
CPU delays caused by waiting for the storage, increases controller cards in compute hosts to improve the performance of the
performance. We recommend evaluating the underlying disk subsystem.</para>
gains from using RAID controller cards in compute hosts to <para>Depending on the storage architecture, you can adopt a scale-out
improve the performance of the underlying disk solution, or use a highly expandable and scalable centralized storage
subsystem.</para> array. If a centralized storage array is the right fit for
<para>The selection of storage architecture determines if a scale-out your requirements, then the array vendor determines the hardware
solution should be used or if a single, highly expandable and selection. It is possible to build a storage array using commodity
scalable centralized storage array would be a better choice. hardware with Open Source software, but requires people with expertise
If a centralized storage array is the right fit for the requirements to build such a system.</para>
then the array vendor determines the hardware selection. It is possible
to build a storage array using commodity hardware with Open Source
software, but requires people with expertise to build such a system.</para>
<para>On the other hand, a scale-out storage solution that <para>On the other hand, a scale-out storage solution that
uses direct-attached storage (DAS) in the servers may be an uses direct-attached storage (DAS) in the servers may be an
appropriate choice. This requires configuration of the server appropriate choice. This requires configuration of the server
hardware to support the storage solution.</para> hardware to support the storage solution.</para>
<para>Considerations affecting storage architecture (and corresponding <para>Considerations affecting storage architecture (and corresponding
storage hardware) of a Storage-focused OpenStack cloud:</para> storage hardware) of a Storage-focused OpenStack cloud include:</para>
<variablelist> <variablelist>
<varlistentry> <varlistentry>
<term>Connectivity</term> <term>Connectivity</term>
<listitem> <listitem>
<para>Based on the selected storage solution, ensure the <para>Based on the selected storage solution, ensure the
connectivity matches the storage solution requirements. connectivity matches the storage solution requirements.
If selecting centralized storage array, determine how the For example, if you select a centralized storage array,
hypervisors connect to the storage array. determine how the hypervisors connect to the storage
Connectivity can affect latency and thus performance. array. Connectivity can affect latency and thus
We recommended confirming that the network performance. We recommended confirming that the network
characteristics minimize latency to boost the characteristics minimize latency to boost the
overall performance of the design.</para> overall performance of the design.</para>
</listitem> </listitem>
@ -116,7 +101,7 @@
<term>Throughput</term> <term>Throughput</term>
<listitem> <listitem>
<para>Ensure that the storage solution <para>Ensure that the storage solution
throughput is optimized based on application throughput is optimized for your application
requirements.</para> requirements.</para>
</listitem> </listitem>
</varlistentry> </varlistentry>
@ -133,8 +118,8 @@
<section xml:id="compute-server-hardware-selection"> <section xml:id="compute-server-hardware-selection">
<title>Compute (server) hardware selection</title> <title>Compute (server) hardware selection</title>
<para>Evaluate Compute (server) hardware four <para>Four opposing factors determine the compute (server)
opposing dimensions:</para> hardware selection:</para>
<variablelist> <variablelist>
<varlistentry> <varlistentry>
<term>Server density</term> <term>Server density</term>
@ -161,7 +146,7 @@
<varlistentry> <varlistentry>
<term>Cost</term> <term>Cost</term>
<listitem> <listitem>
<para>The relative of the hardware weighted against <para>The relative cost of the hardware weighed against
the level of design effort needed to build the the level of design effort needed to build the
system.</para> system.</para>
</listitem> </listitem>
@ -213,8 +198,8 @@
(ODMs) or second-tier manufacturers.</para> (ODMs) or second-tier manufacturers.</para>
<warning> <warning>
<para>This may cause issues for organizations that have <para>This may cause issues for organizations that have
preferred vendor policies or concerns with support and hardware preferred vendor policies or concerns with support and
warranties of non-tier 1 vendors. hardware warranties of non-tier 1 vendors.
</para> </para>
</warning> </warning>
</listitem> </listitem>
@ -235,8 +220,8 @@
<listitem> <listitem>
<para>Rack-mounted servers <para>Rack-mounted servers
that support multiple independent servers in a single that support multiple independent servers in a single
2U or 3U enclosure, "sled servers", deliver increased density as 2U or 3U enclosure, "sled servers", deliver increased
compared to a typical 1U-2U rack-mounted servers.</para> density as compared to a typical 1U-2U rack-mounted servers.</para>
<para>For example, many sled servers offer four independent <para>For example, many sled servers offer four independent
dual-socket nodes in 2U for a total of 8 CPU sockets dual-socket nodes in 2U for a total of 8 CPU sockets
in 2U. However, the dual-socket limitation on in 2U. However, the dual-socket limitation on
@ -244,9 +229,9 @@
additional cost and configuration complexity.</para> additional cost and configuration complexity.</para>
</listitem> </listitem>
</itemizedlist> </itemizedlist>
<para>Other factors strongly influence server hardware <para>Other factors that influence server hardware
selection for a storage-focused OpenStack design selection for a storage-focused OpenStack design
architecture. The following is a list of these factors:</para> architecture include:</para>
<variablelist> <variablelist>
<varlistentry> <varlistentry>
<term>Instance density</term> <term>Instance density</term>
@ -262,7 +247,7 @@
<term>Host density</term> <term>Host density</term>
<listitem> <listitem>
<para>Another option to address the higher <para>Another option to address the higher
host count is to use a quad socket platform. Taking host count is to use a quad-socket platform. Taking
this approach decreases host density which also this approach decreases host density which also
increases rack count. This configuration affects the increases rack count. This configuration affects the
number of power connections and also impacts network number of power connections and also impacts network
@ -282,10 +267,10 @@
</varlistentry> </varlistentry>
</variablelist> </variablelist>
<para>Storage-focused OpenStack design architecture server <para>Storage-focused OpenStack design architecture server
hardware selection should focus on a "scale up" versus "scale hardware selection should focus on a "scale-up" versus
out" solution. The determination of which is the best "scale-out" solution. The determination of which is the best
solution, a smaller number of larger hosts or a larger number of solution (a smaller number of larger hosts or a larger number of
smaller hosts, depends on a combination of factors smaller hosts), depends on a combination of factors
including cost, power, cooling, physical rack and floor space, including cost, power, cooling, physical rack and floor space,
support-warranty, and manageability.</para> support-warranty, and manageability.</para>
</section> </section>
@ -366,8 +351,8 @@
<section xml:id="software-selection-arch-storage"> <section xml:id="software-selection-arch-storage">
<title>Software selection</title> <title>Software selection</title>
<para>Selecting software for a storage-focused <para>Factors that influence the software selection for a storage-focused
OpenStack architecture design includes three areas:</para> OpenStack architecture design include:</para>
<itemizedlist> <itemizedlist>
<listitem> <listitem>
<para>Operating system (OS) and hypervisor</para> <para>Operating system (OS) and hypervisor</para>
@ -385,19 +370,20 @@
<section xml:id="operating-system-and-hypervisor-arch-storage"> <section xml:id="operating-system-and-hypervisor-arch-storage">
<title>Operating system and hypervisor</title> <title>Operating system and hypervisor</title>
<para>Selecting the OS and hypervisor has a significant impact <para>Operating system (OS) and hypervisor have a significant impact
on the overall design and also affects server hardware on the overall design and also affect server hardware
selection. Ensure that the selected operating system and selection. Ensure the selected operating system and
hypervisor combination support the storage hardware and work hypervisor combination support the storage hardware and work
with the networking hardware selection and topology. with the networking hardware selection and topology.
For example, Link Aggregation Control Protocol (LACP) requires For example, Link Aggregation Control Protocol (LACP) requires
support from both the OS and hypervisor.</para> support from both the operating system and hypervisor.</para>
<para>OS and hypervisor selection affect the following areas:</para> <para>Operating system and hypervisor selection affect the following
areas:</para>
<variablelist> <variablelist>
<varlistentry> <varlistentry>
<term>Cost</term> <term>Cost</term>
<listitem> <listitem>
<para>Selection of a commercially supported <para>Selecting a commercially supported
hypervisor, such as Microsoft Hyper-V, results in hypervisor, such as Microsoft Hyper-V, results in
a different cost model than a a different cost model than a
community-supported open source hypervisor like community-supported open source hypervisor like
@ -434,7 +420,7 @@
<varlistentry> <varlistentry>
<term>Scale and performance</term> <term>Scale and performance</term>
<listitem> <listitem>
<para>Ensure that the selected OS <para>Ensure the selected OS
and hypervisor combination meet the appropriate scale and hypervisor combination meet the appropriate scale
and performance requirements needed for this storage and performance requirements needed for this storage
focused OpenStack cloud. The chosen architecture must focused OpenStack cloud. The chosen architecture must
@ -445,7 +431,7 @@
<varlistentry> <varlistentry>
<term>Security</term> <term>Security</term>
<listitem> <listitem>
<para>Ensure that the design can accommodate <para>Ensure the design can accommodate
the regular periodic installation of application the regular periodic installation of application
security patches while maintaining the required security patches while maintaining the required
workloads. The frequency of security patches for the workloads. The frequency of security patches for the
@ -457,19 +443,17 @@
<varlistentry> <varlistentry>
<term>Supported features</term> <term>Supported features</term>
<listitem> <listitem>
<para>Determine the required features of <para>Selecting the OS-hypervisor combination often determines
OpenStack. This often determines the the required features of OpenStack. Certain features are only
selection of the OS-hypervisor combination. Certain available with specific OSes or hypervisors. For example,
features are only available with specific OSes or if certain features are not available, you might need to modify
hypervisors. For example, if certain features are not the design to meet user requirements.</para>
available, you might need to modify the design to
meet user requirements.</para>
</listitem> </listitem>
</varlistentry> </varlistentry>
<varlistentry> <varlistentry>
<term>Interoperability</term> <term>Interoperability</term>
<listitem> <listitem>
<para>Any chosen OS/hypervisor combination should be chosen <para>The OS-hypervisor combination should be chosen
based on the interoperability with one another, and other based on the interoperability with one another, and other
OS-hyervisor combinations. Operational and troubleshooting OS-hyervisor combinations. Operational and troubleshooting
tools for one OS-hypervisor combination may differ tools for one OS-hypervisor combination may differ
@ -484,19 +468,19 @@
<section xml:id="openstack-components-arch-storage"> <section xml:id="openstack-components-arch-storage">
<title>OpenStack components</title> <title>OpenStack components</title>
<para>Which OpenStack components you choose can have a significant <para>The OpenStack components you choose can have a significant
impact on the overall design. While there are certain impact on the overall design. While there are certain
components that are always present, Compute and Image service, for components that are always present
example, there are other services that may not need to be (Compute and Image service, for example), there are other services
present. As an example, a certain design may not require that may not be required. As an example, a certain design
the Orchestration module. Omitting Orchestration would not typically have a may not require the Orchestration module. Omitting Orchestration would
significant impact on the overall design, however, if the not typically have a significant impact on the overall design,
architecture uses a replacement for OpenStack Object Storage for its however, if the architecture uses a replacement for OpenStack Object
storage component, this could potentially have significant Storage for its storage component, this could potentially have
impacts on the rest of the design.</para> significant impacts on the rest of the design.</para>
<para>A storage-focused design might require the ability to use <para>A storage-focused design might require the ability to use
Orchestration to launch instances with Block Storage volumes to perform Orchestration to launch instances with Block Storage volumes to
storage-intensive processing.</para> perform storage-intensive processing.</para>
<para>A storage-focused OpenStack design architecture typically uses the <para>A storage-focused OpenStack design architecture typically uses the
following components:</para> following components:</para>
<itemizedlist> <itemizedlist>
@ -507,11 +491,12 @@
<para>OpenStack dashboard (horizon)</para> <para>OpenStack dashboard (horizon)</para>
</listitem> </listitem>
<listitem> <listitem>
<para>OpenStack Compute (nova) (including the use of multiple hypervisor <para>OpenStack Compute (nova) (including the use of multiple
drivers)</para> hypervisor drivers)</para>
</listitem> </listitem>
<listitem> <listitem>
<para>OpenStack Object Storage (swift) (or another object storage solution)</para> <para>OpenStack Object Storage (swift) (or another object
storage solution)</para>
</listitem> </listitem>
<listitem> <listitem>
<para>OpenStack Block Storage (cinder)</para> <para>OpenStack Block Storage (cinder)</para>
@ -520,7 +505,8 @@
<para>OpenStack Image service (glance)</para> <para>OpenStack Image service (glance)</para>
</listitem> </listitem>
<listitem> <listitem>
<para>OpenStack Networking (neutron) or legacy networking (nova-network)</para> <para>OpenStack Networking (neutron) or legacy networking
(nova-network)</para>
</listitem> </listitem>
</itemizedlist> </itemizedlist>
<para>Excluding certain OpenStack components may limit or <para>Excluding certain OpenStack components may limit or
@ -547,7 +533,7 @@
services for instances. There are many additional networking services for instances. There are many additional networking
software packages that may be useful to manage the OpenStack software packages that may be useful to manage the OpenStack
components themselves. Some examples include HAProxy, components themselves. Some examples include HAProxy,
keepalived, and various routing daemons (like Quagga). The Keepalived, and various routing daemons (like Quagga). The
<citetitle>OpenStack High Availability Guide</citetitle> describes <citetitle>OpenStack High Availability Guide</citetitle> describes
some of these software packages, HAProxy in particular. See the <link some of these software packages, HAProxy in particular. See the <link
xlink:href="http://docs.openstack.org/high-availability-guide/content/ch-network.html">Network xlink:href="http://docs.openstack.org/high-availability-guide/content/ch-network.html">Network