<html>
  <head>
    <meta content="text/html; charset=utf-8" http-equiv="Content-Type">
  </head>
  <body bgcolor="#FFFFFF" text="#000000">
    <p><br>
    </p>
    <br>
    <div class="moz-cite-prefix">On 09/05/17 19:18, hvjunk wrote:<br>
    </div>
    <blockquote
      cite="mid:E9BD1C5E-D891-4608-9D34-69DEAC4A57BA@gmail.com"
      type="cite">
      <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
      <br class="">
      <div>
        <blockquote type="cite" class="">
          <div class="">On 03 May 2017, at 07:49 , Jiffin Tony Thottan
            &lt;<a moz-do-not-send="true"
              href="mailto:jthottan@redhat.com" class="">jthottan@redhat.com</a>&gt;
            wrote:</div>
          <br class="Apple-interchange-newline">
          <div class="">
            <div class=""><br class="">
              <br class="">
              On 02/05/17 15:27, hvjunk wrote:<br class="">
              <blockquote type="cite" class="">Good day,<br class="">
                <br class="">
                I’m busy setting up/testing NFS-HA with GlusterFS
                storage across VMs running Debian 8. GlusterFS volume to
                be "replica 3 arbiter 1"<br class="">
                <br class="">
                In the NFS-ganesha information I’ve gleamed thus far, it
                mentions the "gluster volume set all
                cluster.enable-shared-storage enable”.<br class="">
                <br class="">
                My first question is this: is that shared volume that
                gets created/setup, suppose to be resilient across
                reboots?<br class="">
                 It appears to not be the case in my test setup thus
                far, that that mount doesn’t get recreated/remounted
                after a reboot.<br class="">
              </blockquote>
              <br class="">
              Following is the script which creates shared storage and
              mount it in the node, plus an entry will be added to
              /etc/fstab<br class="">
              <a moz-do-not-send="true"
href="https://github.com/gluster/glusterfs/blob/master/extras/hook-scripts/set/post/S32gluster_enable_shared_storage.sh"
                class="">https://github.com/gluster/glusterfs/blob/master/extras/hook-scripts/set/post/S32gluster_enable_shared_storage.sh</a><br
                class="">
              <br class="">
              But there is a possibility such that, if glusterd(I hope u
              have enabled enabled glusterd service) is not started
              before<br class="">
              systemd tries mount the shared storage then it will fail.<br
                class="">
            </div>
          </div>
        </blockquote>
        <div><br class="">
        </div>
      </div>
    </blockquote>
    <br>
    <br>
    Thanks for the systemd helper script<br>
    --<br>
    Jiffin<br>
    <br>
    <blockquote
      cite="mid:E9BD1C5E-D891-4608-9D34-69DEAC4A57BA@gmail.com"
      type="cite">
      <div>Thank Jiffin,</div>
      <div><br class="">
      </div>
      <div> I since found that (1) you need to wait a bit  for the
        cluster to “settle” with that script having executed, before you
        reboot the cluster (As you might see in my bitbucket ansible
        scripts in <a moz-do-not-send="true"
          href="https://bitbucket.org/dismyne/gluster-ansibles/src"
          class="">https://bitbucket.org/dismyne/gluster-ansibles/src</a> ) …
        something to add in the manuals perhaps to warn people to wait
        for that script to finish before rebooting node/vm/server(s)?</div>
      <div><br class="">
      </div>
      <div> (2) the default configuration, can’t bootstrap the
        /gluster_shared_storage volume/directory reliably from a clean
        shutdown-reboot of the whole cluster!!!</div>
      <div><br class="">
      </div>
      <div>The problem: SystemD and it’s wanting to have the control
        over /etc/fstab and the mounting, and and and…. (and I’ll not
        empty my mind about L.P. based on his remarks in: <a
          moz-do-not-send="true"
href="https://github.com/systemd/systemd/issues/4468#issuecomment-255711912"
          class="">https://github.com/systemd/systemd/issues/4468#issuecomment-255711912</a> after
        my struggling with this issue)</div>
      <div><br class="">
      </div>
      <div><br class="">
      </div>
      <div>To have a reliably bootstrapped (from all nodes down booting
        up) I'm using the following SystemD service and helper script(s)
        to have the gluster cluster node mount their local mounts (like
        /gluster_shared_storage) reliably:</div>
      <div><br class="">
      </div>
      <div><a moz-do-not-send="true"
href="https://bitbucket.org/dismyne/gluster-ansibles/src/24b62dcc858364ee3744d351993de0e8e35c2680/ansible/files/glusterfsmounts.service-centos?at=master"
          class="">https://bitbucket.org/dismyne/gluster-ansibles/src/24b62dcc858364ee3744d351993de0e8e35c2680/ansible/files/glusterfsmounts.service-centos?at=master</a></div>
      <br>
      <fieldset class="mimeAttachmentHeader"></fieldset>
      <br>
      <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
      <div><a moz-do-not-send="true"
href="https://bitbucket.org/dismyne/gluster-ansibles/src/24b62dcc858364ee3744d351993de0e8e35c2680/ansible/files/test-mounts.sh?at=master"
          class="">https://bitbucket.org/dismyne/gluster-ansibles/src/24b62dcc858364ee3744d351993de0e8e35c2680/ansible/files/test-mounts.sh?at=master</a></div>
      <br>
      <fieldset class="mimeAttachmentHeader"></fieldset>
      <br>
      <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
      <div><br class="">
      </div>
      <div><br class="">
      </div>
      <div><br class="">
        <blockquote type="cite" class="">
          <div class="">
            <div class=""><br class="">
              --<br class="">
              Jiffin<br class="">
              <blockquote type="cite" class=""><br class="">
                If the mount is not resilient, ie. not recreated/mounted
                by glusterfs and neither added to the /etc/fstab by
                glusterfs, why the initial auto mount by glusterfs and
                not afterwards with a reboot?<br class="">
                <br class="">
                The biggest “issue” I have found with glusterfs is the
                interaction with SystemD and mounts that fails and don’t
                get properly retried later (Will email separately on
                that issue) during bootstrapping of the cluster, and
                that is why I need to confirm the reasoning/etc. on this
                initial auto-mounting, but then the need to manually add
                it into the /etc/fstab<br class="">
                <br class="">
                Thank you<br class="">
                Hendrik<br class="">
                <br class="">
                _______________________________________________<br
                  class="">
                Gluster-users mailing list<br class="">
                <a moz-do-not-send="true"
                  href="mailto:Gluster-users@gluster.org" class="">Gluster-users@gluster.org</a><br
                  class="">
                <a class="moz-txt-link-freetext" href="http://lists.gluster.org/mailman/listinfo/gluster-users">http://lists.gluster.org/mailman/listinfo/gluster-users</a><br
                  class="">
              </blockquote>
              <br class="">
            </div>
          </div>
        </blockquote>
      </div>
      <br class="">
    </blockquote>
    <br>
  </body>
</html>