[OpenAFS] AFS lag

Kim Kimball dhk@ccre.com
Tue, 26 May 2009 21:03:21 -0600


<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
  <meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type">
</head>
<body bgcolor="#ffffff" text="#000000">
Derrick Brashear wrote:
<blockquote cite="mid:2702163.1237432093423.JavaMail.root@m01"
 type="cite">
  <pre wrap="">On Wed, Mar 18, 2009 at 9:56 PM, Ken Hornstein <a class="moz-txt-link-rfc2396E" href="mailto:kenh@cmf.nrl.navy.mil">&lt;kenh@cmf.nrl.navy.mil&gt;</a> wrote:
  </pre>
  <blockquote type="cite">
    <blockquote type="cite">
      <pre wrap="">I'm no ubik engineer, but as far as I understand it, the protocol was not
designed for even numbers of participating servers. For best results, three
or five servers seem to be optimum.
      </pre>
    </blockquote>
  </blockquote>
</blockquote>
I hear this frequently, and don't see why it should be true.&nbsp; The tie
breaking mechanism during an election is simple.<br>
<br>
Kim<br>
<br>
<br>
<br>
<blockquote cite="mid:2702163.1237432093423.JavaMail.root@m01"
 type="cite">
  <blockquote type="cite">
    <pre wrap="">There is a lot of misinformation about Ubik out there; the voting
protocol is actually not complicated, it's just not documented well.
    </pre>
  </blockquote>
  <pre wrap=""><!---->
it's actually well-documented, if you find Kazar's paper on Quorum Completion.

  </pre>
  <blockquote type="cite">
    <pre wrap="">If your database servers are accessable via the Internet, we could take
a look at them via udebug. &nbsp;Really, there are only a few things that can
go wrong; of all of the pieces of AFS, I think Ubik is one of the most
bulletproof.
    </pre>
  </blockquote>
  <pre wrap=""><!---->
There are a couple (unlikely) open issues; See RT.


  </pre>
</blockquote>
<br>
</body>
</html>