154 lines
9.1 KiB
XML
154 lines
9.1 KiB
XML
<section xml:id="GPFS-driver" 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">
|
|
<title>IBM GPFS Volume Driver</title>
|
|
<para>The General Parallel File System (GPFS) is a cluster file system that provides concurrent
|
|
access to file systems from multiple nodes. The storage provided by these nodes can be
|
|
direct attached, network attached, SAN attached or a combination of these methods. GPFS
|
|
provides many features beyond common data access including data replication, policy based
|
|
storage management, and space efficient file snapshot and clone operations.</para>
|
|
<section xml:id="GPFS-driver-background">
|
|
<title>How the GPFS Driver Works</title>
|
|
<para>This driver enables the use of GPFS in a similar fashion as the NFS driver. With the
|
|
GPFS driver, instances do not actually access a storage device at the block level.
|
|
Instead, volume backing files are created in a GPFS file system and mapped to instances,
|
|
which emulate a block device.</para>
|
|
<para>
|
|
<note>
|
|
<para>GPFS software must be installed and running on nodes where Cinder volume and
|
|
Nova compute services are running in the OpenStack environment. A GPFS file
|
|
system must also be created and mounted on these nodes before starting the
|
|
<literal>cinder-volume</literal> service. The details of these GPFS specific
|
|
steps are covered in GPFS Administration documentation.</para>
|
|
</note>
|
|
</para>
|
|
<para>Optionally, Glance can be configured to store images on a GPFS file system. When
|
|
Cinder volumes are created from Glance images, if both image and volume data reside in
|
|
the same GPFS file system, the data from image files is moved efficiently to Cinder
|
|
volumes using copy on write optimization strategy.</para>
|
|
</section>
|
|
<section xml:id="GPFS-driver-options">
|
|
<title>Enabling the GPFS Driver</title>
|
|
<para>To use Cinder with the GPFS driver, first set the <literal>volume_driver</literal> in
|
|
<filename>cinder.conf</filename>:</para>
|
|
<programlisting>volume_driver = cinder.volume.drivers.gpfs.GPFSDriver</programlisting>
|
|
<para>The following table contains the configuration options supported by the GPFS
|
|
driver.</para>
|
|
<xi:include href="../../../common/tables/cinder-gpfs_volume.xml"/>
|
|
<note>
|
|
<para>The flag <literal>gpfs_images_share_mode</literal> is only valid if the Image service
|
|
is configured to use GPFS with <literal>gpfs_images_dir</literal> flag. Also note,
|
|
when the value of this flag is <literal>copy_on_write</literal>, the paths
|
|
specified by the flags <literal>gpfs_mount_point_base</literal> and
|
|
<literal>gpfs_images_dir</literal> must both reside in the same GPFS file system
|
|
and in the same GPFS fileset.
|
|
</para>
|
|
</note>
|
|
</section>
|
|
<section xml:id="GPFS-volume-options">
|
|
<title>Volume Creation Options</title>
|
|
<para>It is possible to specify additional volume configuration options on a per-volume
|
|
basis by specifiying volume metadata. The volume is created using the specified options.
|
|
Changing the metadata after the volume is created has no effect. The following table
|
|
lists the volume creation options supported by the GPFS volume driver.</para>
|
|
<table frame="all">
|
|
<title>List of Volume Creation Options for GPFS Volume Driver:</title>
|
|
<tgroup cols="2">
|
|
<colspec colname="c1" colnum="1" colwidth="1*"/>
|
|
<colspec colname="c2" colnum="2" colwidth="1.74*"/>
|
|
<thead>
|
|
<row>
|
|
<entry>Metadata Item Name</entry>
|
|
<entry>Description</entry>
|
|
</row>
|
|
</thead>
|
|
<tbody>
|
|
<row>
|
|
<entry><literal>fstype</literal></entry>
|
|
<entry>The driver will create a file system or swap area on the new volume.
|
|
If <literal>fstype=swap</literal> is specified, the mkswap command is
|
|
used to create a swap area. Otherwise the mkfs command is passed the
|
|
specified type, for example ext3, ext4, etc.</entry>
|
|
</row>
|
|
<row>
|
|
<entry><literal>fslabel</literal></entry>
|
|
<entry>The driver will set the file system label for the file system
|
|
specified by fstype option. This value is only used if fstype is
|
|
specified.</entry>
|
|
</row>
|
|
<row>
|
|
<entry><literal>data_pool_name</literal></entry>
|
|
<entry>
|
|
<para>The driver will assign the volume file to the specified GPFS
|
|
storage pool. Note that the GPFS storage pool must already be
|
|
created.</para>
|
|
</entry>
|
|
</row>
|
|
<row>
|
|
<entry><literal>replicas</literal></entry>
|
|
<entry>
|
|
<para>Specify how many copies of the volume file to create. Valid values
|
|
are 1, 2, and, for GPFS V3.5.0.7 and later, 3. This value cannot be
|
|
greater than the value of the MaxDataReplicas attribute of the file
|
|
system.</para>
|
|
</entry>
|
|
</row>
|
|
<row>
|
|
<entry><literal>dio</literal></entry>
|
|
<entry>
|
|
<para>Enable or disable the Direct I/O caching policy for the volume
|
|
file. Valid values are "yes" and "no".</para>
|
|
</entry>
|
|
</row>
|
|
<row>
|
|
<entry><literal>write_affinity_depth</literal></entry>
|
|
<entry>
|
|
<para>Specify the allocation policy to be used for the volume file. Note
|
|
that this option only works if "allow-write-affinity" is set for the
|
|
GPFS data pool.</para>
|
|
</entry>
|
|
</row>
|
|
<row>
|
|
<entry><literal>block_group_factor</literal></entry>
|
|
<entry>
|
|
<para>Specify how many blocks are laid out sequentially in the volume
|
|
file to behave like a single large block. This option only works if
|
|
"allow-write-affinity" is set for the GPFS data pool.</para>
|
|
</entry>
|
|
</row>
|
|
<row>
|
|
<entry><literal>write_affinity_failure_group</literal></entry>
|
|
<entry>
|
|
<para>Specify the range of nodes (in GPFS shared nothing architecture)
|
|
where replicas of blocks in the volume file are to be written. See
|
|
GPFS Administration and Programming Reference guide for more details
|
|
on this option.</para>
|
|
</entry>
|
|
</row>
|
|
</tbody>
|
|
</tgroup>
|
|
</table>
|
|
<simplesect>
|
|
<title>Example Using Volume Creation Options</title>
|
|
<para>This example shows the creation of a 50GB volume with an ext4 filesystem labeled
|
|
<literal>newfs</literal>and direct IO enabled:</para>
|
|
<screen><prompt>$</prompt><userinput>cinder create --metadata fstype=ext4 fslabel=newfs dio=yes --display-name volume_1 50</userinput> </screen>
|
|
</simplesect>
|
|
</section>
|
|
<section xml:id="GPFS-operational-notes">
|
|
<title>Operational Notes for GPFS Driver</title>
|
|
<simplesect>
|
|
<title>Snapshots and Clones</title>
|
|
<para>Volume snapshots are implemented using the GPFS file clone feature. Whenever a new
|
|
snapshot is created, the snapshot file is efficiently created as a read-only clone
|
|
parent of the volume, and the volume file uses copy on write optimization strategy
|
|
to minimize data movement.</para>
|
|
<para>Similarly when a new volume is created from a snapshot or from an existing volume,
|
|
the same approach is taken. The same approach is also used when a new volume is
|
|
created from a Glance image, if the source image is in raw format, and
|
|
<literal>gpfs_images_share_mode</literal> is set to
|
|
<literal>copy_on_write</literal>.</para>
|
|
</simplesect>
|
|
</section>
|
|
</section>
|