nbdkit-xz-plugin man page

nbdkit-xz-plugin — nbdkit xz plugin

Synopsis

 nbdkit xz [file=]FILENAME.xz

Description

"nbdkit-xz-plugin" is a file serving plugin for nbdkit(1).

It serves the named "FILENAME.xz" over NBD, uncompressing it on the fly.  The plugin only supports read-only connections.

Getting Best Random Access Performance from XZ

xz(1) files are split into streams and blocks.  Most xz files contain only one stream which contains one or more blocks.  You can find out how many streams and blocks are in an xz file by doing:

 $ xz --list winxp.img.xz
 Strms  Blocks   Compressed Uncompressed  Ratio  Check   Filename
     1       1  2,100.0 MiB  6,144.0 MiB  0.342  CRC64   winxp.img.xz
    ↑↑↑     ↑↑↑
  streams  blocks

xz files are seekable on block boundaries only.  Seeking is done by seeking directly to the lower block boundary, then uncompressing data until the precise byte is reached.  As uncompressing xz data is slow, to get best random access performance, you must prepare your xz files with many small blocks.

Use the --block-size option with a small-ish block size.  For example this is the same image compressed with a 16 MB block size:

 $ xz --best --block-size=16777216 winxp.img
             ─────────────────────
 $ xz --list winxp.img.xz
 Strms  Blocks   Compressed Uncompressed  Ratio  Check   Filename
     1     384  2,120.1 MiB  6,144.0 MiB  0.345  CRC64   winxp.img.xz
           ↑↑↑
          blocks

This file can be accessed randomly.  At most 16 MB will have to be uncompressed to seek to any byte.

As you would expect, xz cannot compress as efficiently when using a smaller block size.  The space penalty in the above example is < 1% of the compressed file size.

Parameters

file=FILENAME.xz

Serve the file named "FILENAME.xz".

This parameter is required.

In nbdkit ≥ 1.8, "file=" may be omitted.  To ensure that the filename does not end up being parsed accidentally as "key=value", prefix relative paths with "./" (absolute paths do not need modification).

maxblock=SIZE

The maximum block size that the plugin will read.  The plugin will refuse to read xz files that contain any block larger than this size.

See the discussion above about creating xz files with small block sizes in order to reduce memory usage and increase performance.

This parameter is optional.  If not specified it defaults to 512M.

maxdepth=N

Maximum number of blocks stored in the LRU block cache.

This parameter is optional.  If not specified it defaults to 8.

The plugin may allocate up to maximum block size in file * maxdepth bytes of memory per connection.

See Also

nbdkit(1), nbdkit-plugin(3), xz(1).

Authors

Richard W.M. Jones

License

Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met:

THIS SOFTWARE IS PROVIDED BY RED HAT AND CONTRIBUTORS ''AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL RED HAT OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

Referenced By

nbdkit(1), nbdkit-cow-filter(1), nbdkit-gzip-plugin(1), nbdkit-plugin(3), virt-builder(1).

2018-09-18 nbdkit-1.7.3 NBDKIT