<?xml version="1.0" encoding="utf-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
	<channel>
		<atom:link href="https://dev1galaxy.org/extern.php?action=feed&amp;tid=8066&amp;type=rss" rel="self" type="application/rss+xml" />
		<title><![CDATA[Dev1 Galaxy Forum / Refractasnapshot – Comparison of different Compression Algorithms]]></title>
		<link>https://dev1galaxy.org/viewtopic.php?id=8066</link>
		<description><![CDATA[The most recent posts in Refractasnapshot – Comparison of different Compression Algorithms.]]></description>
		<lastBuildDate>Sun, 28 Jun 2026 17:29:11 +0000</lastBuildDate>
		<generator>FluxBB</generator>
		<item>
			<title><![CDATA[Refractasnapshot – Comparison of different Compression Algorithms]]></title>
			<link>https://dev1galaxy.org/viewtopic.php?pid=64444#p64444</link>
			<description><![CDATA[<p>When I was creating an iso-image of my new Refracta based system, I have noticed, that zstd -22 and zstd -19 deliver the same results. Obviously the zstd ultra mode does not work properly within Refractasnapshot. This inspired me to do some testing of different compression options.<br />Here are my results:</p><p>Total size of the system on qemu hd:<br />/ 4,09 GB + /home/ 289 MB = 4,37 GB</p><p>xz:&#160; &#160; &#160; &#160; 1572 MB&#160; &#160; (-4.1% to zstd -19 | -15,4% to zstd -1)<br />xz -b 262144:&#160; &#160; 1540 MB&#160; &#160;&#160; &#160; (-2.0% to xz)<br />xz -b 524288:&#160; &#160; 1512 MB&#160; &#160;&#160; &#160; (-3.8% to xz)<br />xz -b 1048576:&#160; &#160; 1486 MB&#160; &#160;&#160; &#160; (-5.5% to xz)</p><p>xz -Xbcj x86&#160; &#160; 1548 MB&#160; &#160; (-1.5% to xz) </p><p>zstd -22:&#160; &#160; 1639 MB&#160; &#160; (&#039;zstd --ultra -22&#039; does not work)<br />zstd -19:&#160; &#160; 1639 MB&#160; &#160; (-11.8% to zstd -1)<br />zstd -15:&#160; &#160; 1652 MB&#160; &#160; (-11.1% to zstd -1)<br />zstd -8:&#160; &#160; 1727 MB&#160; &#160; (-7.0%&#160; to zstd -1)<br />zstd -3:&#160; &#160; 1800 MB&#160; &#160; (-3.1%&#160; to zstd -1)<br />zstd -1:&#160; &#160; 1858 MB&#160; &#160;&#160; &#160; (-17.2%&#160; to lz4)</p><p>lz4:&#160; &#160; &#160; &#160; 2245 MB&#160; &#160; (+20,8% to zstd -1)</p><p>Relative compression speed according to Alan Jude (2017):<br />[<a href="https://openzfs.org/w/images/b/b3/03-OpenZFS_2017_-_ZStandard_in_ZFS.pdf" rel="nofollow">https://openzfs.org/w/images/b/b3/03-Op … in_ZFS.pdf</a>]<br />zstd -19 = 1x&#160; &#160; | zstd --ultra -22 = 0.58X<br />zstd -15 = 2.3x&#160; | zstd -8 = 16.1x <br />zstd -3&#160; = 59.1x | zstd -1 = 101.5x<br />lz4 = 172.3x</p><p>According to Northon Torga (2024) xz and zstd -19 have a similar compression speed:<br />[<a href="https://ntorga.com/gzip-bzip2-xz-zstd-7z-brotli-or-lz4/" rel="nofollow">https://ntorga.com/gzip-bzip2-xz-zstd-7z-brotli-or-lz4/</a>]<br />zstd -19 = 1x&#160; |&#160; xz = 0.96x</p><p>Decompression speed of xz is about 5 times slower than zstd:<br />xz = ~0,18x&#160; |&#160; zstd = 1x&#160; |&#160; lz4 = ~ 3.2x</p><p>Depending on hardware and kind of files the actual results may differ.</p>]]></description>
			<author><![CDATA[dummy@example.com (Xaver)]]></author>
			<pubDate>Sun, 28 Jun 2026 17:29:11 +0000</pubDate>
			<guid>https://dev1galaxy.org/viewtopic.php?pid=64444#p64444</guid>
		</item>
	</channel>
</rss>
