<html>
  <head>
    <meta content="text/html; charset=windows-1252"
      http-equiv="Content-Type">
  </head>
  <body bgcolor="#FFFFFF" text="#000000">
    <div class="moz-cite-prefix">On 10/28/2014 02:50 PM, John Griffith
      wrote:<br>
    </div>
    <blockquote
cite="mid:CA+qL3LVJs2BRUFp=J++NLULXRSRH67MQL9LZcA5psTYPFo6cNg@mail.gmail.com"
      type="cite">
      <meta http-equiv="Content-Type" content="text/html;
        charset=windows-1252">
      <div dir="ltr"><br>
        <div class="gmail_extra"><br>
          <div class="gmail_quote">On Tue, Oct 28, 2014 at 12:37 PM, Dan
            Genin <span dir="ltr"><<a moz-do-not-send="true"
                href="mailto:daniel.genin@jhuapl.edu" target="_blank">daniel.genin@jhuapl.edu</a>></span>
            wrote:<br>
            <blockquote class="gmail_quote" style="margin:0px 0px 0px
0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex">Great,
              thank you, Duncan. I will then proceed with the shared
              volume group.<span class=""><font color="#888888"><br>
                  <br>
                  Dan</font></span>
              <div class="">
                <div class="h5"><br>
                  <br>
                  On 10/28/2014 02:06 PM, Duncan Thomas wrote:<br>
                  <blockquote class="gmail_quote" style="margin:0px 0px
                    0px
0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex">Cinder
                    volumes are always (unless you go change the
                    default) in the<br>
                    form: volume-<uuid>, and since the string
                    'volume-' is never a valid<br>
                    uuid, then I think we can work around nova volumes
                    fine when we come<br>
                    to write our tests.<br>
                    <br>
                    Sorry for the repeated circling on this, but I think
                    I'm now happy.<br>
                    <br>
                    Thanks<br>
                    <br>
                    <br>
                    <br>
                    On 28 October 2014 17:53, Dan Genin <<a
                      moz-do-not-send="true"
                      href="mailto:daniel.genin@jhuapl.edu"
                      target="_blank">daniel.genin@jhuapl.edu</a>>
                    wrote:<br>
                    <blockquote class="gmail_quote" style="margin:0px
                      0px 0px
0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex">On
                      10/28/2014 11:56 AM, Dean Troyer wrote:<br>
                      <br>
                      On Tue, Oct 28, 2014 at 9:27 AM, Dan Genin <<a
                        moz-do-not-send="true"
                        href="mailto:daniel.genin@jhuapl.edu"
                        target="_blank">daniel.genin@jhuapl.edu</a>>
                      wrote:<br>
                      <blockquote class="gmail_quote" style="margin:0px
                        0px 0px
0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex">So
                        this brings us back to the original proposal of
                        having separate backing<br>
                        files for Cinder and Nova which Dean thought
                        might take too much space.<br>
                      </blockquote>
                      <br>
                      Between Cinder, Nova and Swift (and Ceph, etc)
                      everybody wants some loopback<br>
                      disk images.  DevStack's Swift and Ceph
                      configurations assume loopback<br>
                      devices and do no sharing.<br>
                      <br>
                      <blockquote class="gmail_quote" style="margin:0px
                        0px 0px
0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex">Duncan,
                        could you please elaborate on the pain a single
                        volume group is<br>
                        likely to cause for Cinder? Is it a show
                        stopper?<br>
                      </blockquote>
                      <br>
                      Back in the day, DevStack was built to configure
                      Cinder (and Nova Volume<br>
                      before that) to use a specific existing volume
                      group (VOLUME_GROUP_NAME) or<br>
                      create a loopback file if necessary.  With the
                      help of VOLUME_NAME_PREFIX<br>
                      and volume_name_template DevStack knew which
                      logical volumes belong to<br>
                      Cinder and could Do The Right Thing.<br>
                      <br>
                      With three loopback files being created, all
                      wanting larger and larger<br>
                      defaults, adding a fourth becomes Just One More
                      Thing.  If Nova's use of LVM<br>
                      is similar enough to Cinder's (uses deterministic
                      naming for the LVs) I'm<br>
                      betting we could make it work.<br>
                      <br>
                      dt<br>
                      <br>
                      Nova's disk names are of the form
                      <instance-uuid>_<disk_type>. So<br>
                      deterministic but, unfortunately, not necessarily
                      predictable. It sounds<br>
                      like Duncan is saying that Cinder needs a fixed
                      prefix for testing its<br>
                      functionality. I will be honest, I am not
                      optimistic about convincing Nova<br>
                      to change their disk naming scheme for the sake of
                      LVM testing. Far more<br>
                      important changes have lingered for months and
                      sometimes longer.<br>
                      <br>
                      It sounds like you are concerned about two issues
                      with regard to the<br>
                      separate volume groups approach: 1) potential loop
                      device shortage and 2)<br>
                      growing space demand. The second issue, it seems
                      to me, will arise no matter<br>
                      which of the two solutions we choose. More space
                      will be required for<br>
                      testing Nova's LVM functionality one way or
                      another, although, using a<br>
                      shared volume group would permit a more efficient
                      use of the available<br>
                      space. The first issue is, indeed, a direct
                      consequence of the choice to use<br>
                      distinct volume groups. However, the number of
                      available loop devices can be<br>
                      increased by passing the appropriate boot
                      parameter to the kernel, which can<br>
                      be easy or hard depending on how the test VMs are
                      spun up.<br>
                      <br>
                      I am not saying that we should necessarily go the
                      way of separate volume<br>
                      groups but, assuming for the moment that changing
                      Nova's disk naming scheme<br>
                      is not an option, we need to figure out what will
                      bring the least amount of<br>
                      pain forcing Cinder tests to work around Nova
                      volumes or create separate<br>
                      volume groups.<br>
                      <br>
                      Let me know what you think.<br>
                      Dan<br>
                      <br>
                      <br>
                      --<br>
                      <br>
                      Dean Troyer<br>
                      <a moz-do-not-send="true"
                        href="mailto:dtroyer@gmail.com" target="_blank">dtroyer@gmail.com</a><br>
                      <br>
                      <br>
                      _______________________________________________<br>
                      OpenStack-dev mailing list<br>
                      <a moz-do-not-send="true"
                        href="mailto:OpenStack-dev@lists.openstack.org"
                        target="_blank">OpenStack-dev@lists.openstack.org</a><br>
                      <a moz-do-not-send="true"
                        href="http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev"
                        target="_blank">http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev</a><br>
                      <br>
                      <br>
                      <br>
                      _______________________________________________<br>
                      OpenStack-dev mailing list<br>
                      <a moz-do-not-send="true"
                        href="mailto:OpenStack-dev@lists.openstack.org"
                        target="_blank">OpenStack-dev@lists.openstack.org</a><br>
                      <a moz-do-not-send="true"
                        href="http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev"
                        target="_blank">http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev</a><br>
                      <br>
                    </blockquote>
                    <br>
                    <br>
                  </blockquote>
                  <br>
                  <br>
                </div>
              </div>
              <br>
              _______________________________________________<br>
              OpenStack-dev mailing list<br>
              <a moz-do-not-send="true"
                href="mailto:OpenStack-dev@lists.openstack.org">OpenStack-dev@lists.openstack.org</a><br>
              <a moz-do-not-send="true"
                href="http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev"
                target="_blank">http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev</a><br>
              <br>
            </blockquote>
          </div>
          <br>
        </div>
        <div class="gmail_extra">Noticed my response never posted....
          think there's something up with my mail client, so if you get
          this a few more times forgive me :)</div>
        <div class="gmail_extra"><br>
        </div>
        <div class="gmail_extra">But....</div>
        <div class="gmail_extra">
          <pre style="margin-top:0px;margin-bottom:0px;padding:0px;font-family:inherit;font-size:13px;font-stretch:inherit;line-height:16.25px;white-space:pre-wrap;word-wrap:break-word;color:rgb(31,31,31)">The idea of sharing a VG
between Nova and Cinder is only relevant in an all in one deployment
anyway, it's a specific edge case for testing.  It certainly (IMHO) does
not warrant any changes in Nova and Cinder.  Also keep in mind that at
some point (I think we're already there) we need to consider whether our
default gating and setup can continue to be done on a single node
anyway.
</pre>
        </div>
      </div>
    </blockquote>
    Agree.<br>
    <blockquote
cite="mid:CA+qL3LVJs2BRUFp=J++NLULXRSRH67MQL9LZcA5psTYPFo6cNg@mail.gmail.com"
      type="cite">
      <div dir="ltr">
        <div class="gmail_extra">
          <pre style="margin-top:0px;margin-bottom:0px;padding:0px;font-family:inherit;font-size:13px;font-stretch:inherit;line-height:16.25px;white-space:pre-wrap;word-wrap:break-word;color:rgb(31,31,31)">
The answer to this seems relatively simple to me, Dean pointed out just
add a loopback device specifically for Nova LVM testing and move on.</pre>
        </div>
      </div>
    </blockquote>
    Dean is actually for using a shared VG to conserve space and loop
    devices.<br>
    <br>
    Dan<br>
    <blockquote
cite="mid:CA+qL3LVJs2BRUFp=J++NLULXRSRH67MQL9LZcA5psTYPFo6cNg@mail.gmail.com"
      type="cite">
      <br>
      <fieldset class="mimeAttachmentHeader"></fieldset>
      <br>
      <pre wrap="">_______________________________________________
OpenStack-dev mailing list
<a class="moz-txt-link-abbreviated" href="mailto:OpenStack-dev@lists.openstack.org">OpenStack-dev@lists.openstack.org</a>
<a class="moz-txt-link-freetext" href="http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev">http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev</a>
</pre>
    </blockquote>
    <br>
  </body>
</html>