mars/index.html

177 lines
7.8 KiB
HTML

<!DOCTYPE html>
<html>
<head>
<meta charset='utf-8' />
<meta http-equiv="X-UA-Compatible" content="chrome=1" />
<meta name="description" content="Mars : Asynchronous Block-Level Storage Replication" />
<link rel="stylesheet" type="text/css" media="screen" href="stylesheets/stylesheet.css">
<title>Mars</title>
</head>
<body>
<!-- HEADER -->
<div id="header_wrap" class="outer">
<header class="inner">
<a id="forkme_banner" href="https://github.com/schoebel/mars">View on GitHub</a>
<h1 id="project_title">Mars</h1>
<h2 id="project_tagline">Asynchronous Block-Level Storage Replication</h2>
<section id="downloads">
<a class="zip_download_link" href="https://github.com/schoebel/mars/zipball/master">Download this project as a .zip file</a>
<a class="tar_download_link" href="https://github.com/schoebel/mars/tarball/master">Download this project as a tar.gz file</a>
</section>
</header>
</div>
<!-- MAIN CONTENT -->
<div id="main_content_wrap" class="outer">
<section id="main_content" class="inner">
<h1>
<a name="mars--multiversion-asynchronous-replication-system" class="anchor" href="#mars--multiversion-asynchronous-replication-system"><span class="octicon octicon-link"></span></a>MARS = Multiversion Asynchronous Replication System</h1>
<p><img src="https://github.com/schoebel/mars/blob/master/docu/images/earth-mars-transfer.jpg" alt="MARS Logo"></p>
<p>GPLed software AS IS, sponsored by 1&amp;1 Internet AG (<a href="http://www.1und1.de">www.1und1.de</a>). Contact: <a href="mailto:tst@1und1.de">tst@1und1.de</a></p>
<h2>
<a name="what-is-mars-light" class="anchor" href="#what-is-mars-light"><span class="octicon octicon-link"></span></a>What is MARS Light?</h2>
<p>Main features:</p>
<ul>
<li>Anytime Consistency</li>
<li>Arbitrary Distances</li>
<li>Tolerates Flaky Networks</li>
</ul><p>MARS Light is almost a drop-in replacement for DRBD (block-level storage replication). It runs as a Linux kernel module.</p>
<p>In contrast to plain DRBD, it works <em>asynchronously</em> and over
arbitrary distances. Our internal 1&amp;1 testing runs between datacenters
in the US and Europe. MARS uses very different technology under the
hood, similar to transaction logging of database systems.</p>
<p>Reliability: application and replication are completely decoupled.
Networking problems (e.g. packet loss, bottlenecks) have no
impact onto your application at the primary side.</p>
<p>Anytime Consistency: on a secondary node, its version of the underlying
disk device is always consistent in itself, but may be outdated
(represent a former state from the primary side). Thanks to
incremental replication of the transaction logfiles, usually the
lag-behind will be only a few seconds, or parts of a second.</p>
<p>Synchronous or near-synchronous operating modes are planned for
the future, but are expected to <em>reliably</em> work only over short
distances (less than 50km), due to fundamental properties
of distributed systems.</p>
<p>WARNING! Current stage is BETA. Don't put productive data on it!</p>
<h2>
<a name="documentation--manual" class="anchor" href="#documentation--manual"><span class="octicon octicon-link"></span></a>Documentation / Manual</h2>
<p>See <a href="https://github.com/schoebel/mars/blob/master/docu/mars-manual.pdf">https://github.com/schoebel/mars/blob/master/docu/mars-manual.pdf</a></p>
<h2>
<a name="concepts" class="anchor" href="#concepts"><span class="octicon octicon-link"></span></a>Concepts</h2>
<p>For a very short intro, see my LCA2013 presentation docu/MARS_LCA2013.pdf .</p>
<p>There is also an internal 2-years old concept paper which is so much outdated,
that I don't want to publish it. </p>
<p>The fundamental construction principle of the planned MARS Full
is called Instance Oriented Programming (IOP) and is described in
the following paper:</p>
<p><a href="http://athomux.net/papers/paper_inst2.pdf">http://athomux.net/papers/paper_inst2.pdf</a></p>
<h2>
<a name="history" class="anchor" href="#history"><span class="octicon octicon-link"></span></a>History</h2>
<p>As you can see in the git log, it evolved from a very experimental
concept study, starting in the Summer of 2010.
At this time, I was working on it in my spare time.</p>
<p>In Summer 2011, an "official" internal 1&amp;1 project started, which aimed
to deliver a proof of concept.</p>
<p>In February 2012, a pilot system was rolled out to an internal statistics
server, which collects statistics data from thousands of other servers,
and thus produces a very heavy random-access write load, formerly
replicated with DRBD (which led to performance problems due to massive
randomness). After switching to MARS, the performance was provably
better.
That server was selected because potential loss of statistics data
would be not be that critical as with other productive data, but
nevertheless it operates on productive data and loads.</p>
<p>After curing some small infancy problems, that server runs until today
without problems. It was upgraded to newer versions of MARS several
times (indicated by some of the git tags). Our sysadmins switched the
primary side a few times, without informing me, so I could
sleep better at night without knowing what they did ;)</p>
<p>In Summer 2012, the next "official" internal 1&amp;1 project started. Its goal
was to reach enterprise grade, and therefore to rollout MARS Light on
~15 productive servers, starting with less critical systems like ones
for test webspaces etc.</p>
<p>In December 2012 (shortly before Christmas), I got the official permission
from our CTO Henning Kettler to publish MARS under GPL on github.</p>
<p>Many thanks to him!</p>
<p>Before that point, I was bound to my working contract which kept internal
software as secret by default (when there was no explicit permission).</p>
<p>Now there is a chance to build up an opensource
community for MARS, partially outside of 1&amp;1.</p>
<p>Please contribute! I will be open.</p>
<p>I also try to respect the guidelines from Linus, but probably this
will need more work. I am already planning to invest some time into
community revision of the sourcecode, but there is not yet any schedule.</p>
<p>In May 2013, I got help by my new collegue Frank Liepold. He is working
on a fully automatic test suite which automates regression tests
(goal: rolling releases). That test suite is based on the internal
test suite of blkreplay and can be found in the test_suite/ subdirectory.</p>
<p>In November 2013, internal 1&amp;1 projects started for mass rollout to several thousands of servers.</p>
<p>Although the software continues to be labelled "beta" for the next future, it has reached enterprise grade due to our internal rating process.</p>
<h2>
<a name="future-plans" class="anchor" href="#future-plans"><span class="octicon octicon-link"></span></a>Future Plans</h2>
<p>Smaller Reworks: in Winter 2013/2014, some smaller changes to the symlink tree are planned, in order to make it more readable for humans and to prepare for future enhancements. They will only change the syntax, not the semantics. There will be an uprade plan, i.e. the old symlink tree remains usable; only newly created clusters will use the new structure.</p>
<p>In parallel, the software will be internally divided into three parts:</p>
<ol>
<li>Generic brick framework</li>
<li>AIO personality with XIO bricks</li>
<li>MARS Light application</li>
</ol><p>I hope this will make MARS more attractive for the mainline Linux kernel community.</p>
</section>
</div>
<!-- FOOTER -->
<div id="footer_wrap" class="outer">
<footer class="inner">
<p class="copyright">Mars maintained by <a href="https://github.com/schoebel">schoebel</a></p>
<p>Published with <a href="http://pages.github.com">GitHub Pages</a></p>
</footer>
</div>
</body>
</html>