<html>
  <head>
    <meta content="text/html; charset=utf-8" http-equiv="Content-Type">
  </head>
  <body bgcolor="#FFFFFF" text="#000000">
    <div class="moz-cite-prefix">Hi Ashish,<br>
      <br>
      It was an error on my side, nothing gluster related. The kernel
      version I was running had a bug that prevented the fuse module
      from loading and causing the brick replacement errors. After
      upgrading I can confirm that the process of replacing the brick
      works fine both with the brick to be replaced online or after
      killing the brick process.<br>
      <br>
      I have a question though. In a real (non virtual) server, if I
      setup the server in JBOD mode, and a drive fails, does gluster
      kill the brick pid? <br>
      <br>
      Regards,<br>
      Iñaki.<br>
      <br>
      On 07/08/2016 07:03 AM, Ashish Pandey wrote:<br>
    </div>
    <blockquote
      cite="mid:802626157.4056036.1467954208978.JavaMail.zimbra@redhat.com"
      type="cite">
      <div style="font-family: times new roman, new york, times, serif;
        font-size: 12pt; color: #000000">
        <div><br>
        </div>
        <div>Hi Iñaki<br>
          <div><br>
          </div>
          <div>The steps you are following don't have any issue.<br>
          </div>
          <div>I would like to have more information to debug this
            further.<br>
          </div>
          <div><br>
          </div>
          <div>1 - gluster v info &lt;volname&gt;<br>
          </div>
          <div>2 - gluster v status &lt;volname&gt; before and after
            running replace-brick<br>
          </div>
          3 - Brick logs (for this volume only) from
          /var/log/glusterfs/bricks/<br>
        </div>
        <div>4 - glusterd logs /var/log/glusterfs/  starts with
          "usr-local-etc-glusterfs-glusterd-&lt;volname&gt;"<br>
        </div>
        <div><br>
        </div>
        <div>Although it should not matter,  could you also try to
          replace a brick without killing that brick process?<br>
        </div>
        <div><br>
        </div>
        <div>Ashish<br>
        </div>
        <div><br>
        </div>
        <hr id="zwchr">
        <div
style="color:#000;font-weight:normal;font-style:normal;text-decoration:none;font-family:Helvetica,Arial,sans-serif;font-size:12pt;"
          data-mce-style="color: #000; font-weight: normal; font-style:
          normal; text-decoration: none; font-family:
          Helvetica,Arial,sans-serif; font-size: 12pt;"><b>From: </b>"itlinux_team"
          <a class="moz-txt-link-rfc2396E" href="mailto:itlinux@imppc.org">&lt;itlinux@imppc.org&gt;</a><br>
          <b>To: </b><a class="moz-txt-link-abbreviated" href="mailto:gluster-users@gluster.org">gluster-users@gluster.org</a><br>
          <b>Sent: </b>Wednesday, July 6, 2016 4:33:54 PM<br>
          <b>Subject: </b>[Gluster-users] replace brick in
          distributed-dispersed setup<br>
          <div><br>
          </div>
          Hi all,<br>
          <div><br>
          </div>
          I'm doing some testings with glusterfs in a virtualized
          environment <br>
          running a 3 x (8 + 4) distributed-dispersed volume simulating
          a 3 node <br>
          cluster with 12 drives per node configuration. The system
          versions are:<br>
          <div><br>
          </div>
          OS: Debian jessie kernel 3.16<br>
          <div><br>
          </div>
          Gluster:  3.8.0-2 installed from the gluster.org debian
          repository<br>
          <div><br>
          </div>
          I have tested the node failure scenario while some clients are
          running <br>
          some read/write operations and the setup works as expected.
          Now I'm <br>
          trying to test how to replace a faulty drive on this setup,
          however I'm <br>
          not able to replace a brick. To test it I have:<br>
          <div><br>
          </div>
          1: Find the pid of the brick I'd like to 'fail' and kill the
          process. <br>
          (tried removing the drive from the host but that would make
          the whole <br>
          guest unresponsive)<br>
          <div><br>
          </div>
          2: Attach a new virtual drive, format and mount it<br>
          <div><br>
          </div>
          3: Try the gluster volume replace-brick command<br>
          <div><br>
          </div>
          And I'm getting the following error:<br>
          <div><br>
          </div>
          gluster volume replace-brick vol_1
          glusterserver1:/ext/bricks/brick-1 <br>
          glusterserver1:/ext/bricks/brick-13 commit force<br>
          volume replace-brick: failed: Fuse unavailable<br>
            Replace-brick failed<br>
          <div><br>
          </div>
          I assume I'm doing something wrong but don't know what
          exactly. Looking <br>
          in the documentation I have not found information about brick
          <br>
          replacement in distributed-dispersed setups.<br>
          <div><br>
          </div>
          <br>
          Thanks!<br>
          <div><br>
          </div>
          Iñaki<br>
          <div><br>
          </div>
          <br>
          <div><br>
          </div>
          _______________________________________________<br>
          Gluster-users mailing list<br>
          <a class="moz-txt-link-abbreviated" href="mailto:Gluster-users@gluster.org">Gluster-users@gluster.org</a><br>
          <a class="moz-txt-link-freetext" href="http://www.gluster.org/mailman/listinfo/gluster-users">http://www.gluster.org/mailman/listinfo/gluster-users</a></div>
        <div><br>
        </div>
      </div>
    </blockquote>
    <br>
    <br>
  </body>
</html>