<div dir="ltr">I think it should be possible to have a sanity check like the following:<div>< 2.63 - sorry, that's not going to work</div><div>>=2.63, <2.67 - it kind of works but ipv6 is going to be messed up</div><div>>2.67 - we're all right</div><div><br></div><div>The runtime check on the dhcp agent is a startup check. Personally I think agents should run sanity checks at startup, but if there are concerns against that then let's separate runtime operations and sanity checks. In any case the logic for the dnsmasq version checks should not be duplicated - so if it's moved into sanity checks either the dhcp agent runs these checks at startup, or they would be not run at all by the agent.</div><div><br></div><div>Salvatore</div></div><div class="gmail_extra"><br><div class="gmail_quote">On 8 January 2015 at 13:17, Ihar Hrachyshka <span dir="ltr"><<a href="mailto:ihrachys@redhat.com" target="_blank">ihrachys@redhat.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
  
    
  
  <div bgcolor="#FFFFFF" text="#000000">
    I agree, that is one thing we should not check in runtime.<br>
    <br>
    In ideal world, it wouldn't even check version number, but
    capabilities. I's not clear though whether we can be any smarter
    than that (we would need to run dnsmasq and real dhcp clients to
    check actual capabilities).<span class="HOEnZb"><font color="#888888"><br>
    <br>
    /Ihar</font></span><div><div class="h5"><br>
    <br>
    <div>On 01/08/2015 12:55 PM, Miguel Ángel
      Ajo wrote:<br>
    </div>
    <blockquote type="cite">
      <div style="font-family:Helvetica;font-size:14px">
        <div>Now that I re-read the patch.</div>
        Shouldn't the version checking  need to be converted into a
        sanity check?<br>
      </div>
      <div>
        <div><br>
        </div>
        <div><span style="font-size:10pt">Miguel Ángel Ajo</span></div>
        <div><br>
        </div>
      </div>
      <p style="color:#a0a0a8">On Thursday, 8 de January de 2015 at
        12:51, Kevin Benton wrote:</p>
      <blockquote type="cite" style="border-left-style:solid;border-width:1px;margin-left:0px;padding-left:10px">
        <span>
          <div>
            <div>
              <div dir="ltr">Thanks for the insight.</div>
              <div><br>
                <div>On Thu, Jan 8, 2015 at 3:41 AM, Miguel Ángel Ajo <span dir="ltr"><<a href="mailto:majopela@redhat.com" target="_blank">majopela@redhat.com</a>></span>
                  wrote:<br>
                  <blockquote type="cite">
                    <div>
                      <div style="font-family:Helvetica;font-size:14px">Correct,
                        that’s the problem, what Kevin said should be
                        the ideal case, but distros have
                        <div>proven to fail satisfying this kind of
                          requirements earlier.</div>
                        <div><br>
                        </div>
                        <div>So at least a warning to the user may be
                          good to have IMHO.</div>
                      </div>
                      <span><font color="#888888">
                          <div>
                            <div><br>
                            </div>
                            <div><span style="font-size:10pt">Miguel
                                Ángel Ajo</span></div>
                            <div><br>
                            </div>
                          </div>
                        </font></span>
                      <div>
                        <div>
                          <p style="color:#a0a0a8">On Thursday, 8 de
                            January de 2015 at 12:36, Ihar Hrachyshka
                            wrote:</p>
                          <blockquote type="cite">
                            <div> <span>
                                <div>
                                  <div> The problem is probably due to
                                    the fact that some operators may run
                                    neutron from git and manage their
                                    dependencies in some other way; or
                                    distributions may suck sometimes, so
                                    packagers may miss the release note
                                    and fail to upgrade dnsmasq; or
                                    distributions may have their
                                    specific concerns on upgrading
                                    dnsmasq version, and would just
                                    backport the needed fix to their
                                    'claimed to 2.66' dnsmasq (common
                                    story in Red Hat world).<br>
                                    <br>
                                    <div>On 01/08/2015 05:25 AM, Kevin
                                      Benton wrote:<br>
                                    </div>
                                    <blockquote type="cite">
                                      <div>
                                        <div dir="ltr">If the new
                                          requirement is expressed in
                                          the neutron packages for the
                                          distro, wouldn't it be
                                          transparent to the operators?</div>
                                        <div><br>
                                          <div>On Wed, Jan 7, 2015 at
                                            6:57 AM, Kyle Mestery <span dir="ltr"><<a href="mailto:mestery@mestery.com" target="_blank">mestery@mestery.com</a>></span>
                                            wrote:<br>
                                            <blockquote type="cite">
                                              <div>
                                                <div dir="ltr">
                                                  <div>
                                                    <div><span>On Wed,
                                                        Jan 7, 2015 at
                                                        8:21 AM, Ihar
                                                        Hrachyshka <span dir="ltr"><<a href="mailto:ihrachys@redhat.com" target="_blank">ihrachys@redhat.com</a>></span>
                                                        wrote:<br>
                                                        <blockquote type="cite">
                                                          <div>Hi all,<br>
                                                          <br>
                                                          I've found out
                                                          that dnsmasq
                                                          < 2.67 does
                                                          not work
                                                          properly for
                                                          IPv6 clients
                                                          when it comes
                                                          to MAC address
                                                          matching (it
                                                          fails to
                                                          match, and so
                                                          clients get
                                                          'no addresses
                                                          available'
                                                          response).
                                                          I've requested
                                                          version bump
                                                          to 2.67 in: <a href="https://review.openstack.org/145482" target="_blank">https://review.openstack.org/145482</a><br>
                                                          <br>
                                                          </div>
                                                        </blockquote>
                                                      </span>
                                                      <div>Good catch,
                                                        thanks for
                                                        finding this
                                                        Ihar!<br>
                                                         <br>
                                                      </div>
                                                      <span>
                                                        <blockquote type="cite">
                                                          <div> Now,
                                                          since we've
                                                          already
                                                          released Juno
                                                          with IPv6 DHCP
                                                          stateful
                                                          support, and
                                                          DHCP agent
                                                          still has
                                                          minimal
                                                          version set to
                                                          2.63 there, we
                                                          have a dilemma
                                                          on how to
                                                          manage it from
                                                          stable
                                                          perspective.<br>
                                                          <br>
                                                          Obviously, we
                                                          should
                                                          communicate
                                                          the revealed
                                                          version
                                                          dependency to
                                                          deployers via
                                                          next release
                                                          notes.<br>
                                                          <br>
                                                          Should we also
                                                          backport the
                                                          minimal
                                                          version bump
                                                          to Juno? This
                                                          will result in
                                                          DHCP agent
                                                          failing to
                                                          start in case
                                                          packagers
                                                          don't bump
                                                          dnsmasq
                                                          version with
                                                          the next Juno
                                                          release. If we
                                                          don't bump the
                                                          version, we
                                                          may leave
                                                          deployers
                                                          uninformed
                                                          about the fact
                                                          that their
                                                          IPv6 stateful
                                                          instances
                                                          won't get any
                                                          IPv6 address
                                                          assigned.<br>
                                                          <br>
                                                          An alternative
                                                          is to add a
                                                          special check
                                                          just for Juno
                                                          that would
                                                          WARN
                                                          administrators
                                                          instead of
                                                          failing to
                                                          start DHCP
                                                          agent.<br>
                                                          <br>
                                                          Comments?<br>
                                                          <br>
                                                          </div>
                                                        </blockquote>
                                                      </span>
                                                      <div>Personally, I
                                                        think the WARN
                                                        may be the best
                                                        route to go.
                                                        Backporting a
                                                        change which
                                                        bumps the
                                                        required dnsmasq
                                                        version seems
                                                        like it may be
                                                        harder for
                                                        operators to
                                                        handle.<span><font color="#888888"><br>
                                                          <br>
                                                          Kyle<br>
                                                           <br>
                                                          </font></span></div>
                                                      <span>
                                                        <blockquote type="cite">
                                                          <div> /Ihar<br>
                                                          <br>
_______________________________________________<br>
                                                          OpenStack-dev
                                                          mailing list<br>
                                                          <a href="mailto:OpenStack-dev@lists.openstack.org" target="_blank">OpenStack-dev@lists.openstack.org</a><br>
                                                          <a 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>
                                                          </div>
                                                        </blockquote>
                                                      </span></div>
                                                    <br>
                                                  </div>
                                                </div>
                                                <br>
_______________________________________________<br>
                                                OpenStack-dev mailing
                                                list<br>
                                                <a href="mailto:OpenStack-dev@lists.openstack.org" target="_blank">OpenStack-dev@lists.openstack.org</a><br>
                                                <a 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>
                                              </div>
                                            </blockquote>
                                          </div>
                                          <br>
                                          <br clear="all">
                                          <div><br>
                                          </div>
                                          -- <br>
                                          <div>
                                            <div>Kevin Benton</div>
                                          </div>
                                        </div>
                                        <br>
                                        <fieldset></fieldset>
                                        <br>
                                        <pre>_______________________________________________
OpenStack-dev mailing list
<a href="mailto:OpenStack-dev@lists.openstack.org" target="_blank">OpenStack-dev@lists.openstack.org</a>
<a href="http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev" target="_blank">http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev</a>
</pre>
                                      </div>
                                    </blockquote>
                                    <br>
                                  </div>
                                  <div>
                                    <div>_______________________________________________</div>
                                    <div>OpenStack-dev mailing list</div>
                                    <div><a href="mailto:OpenStack-dev@lists.openstack.org" target="_blank">OpenStack-dev@lists.openstack.org</a></div>
                                    <div><a href="http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev" target="_blank">http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev</a></div>
                                  </div>
                                </div>
                              </span> </div>
                          </blockquote>
                          <div> <br>
                          </div>
                        </div>
                      </div>
                      <br>
                      _______________________________________________<br>
                      OpenStack-dev mailing list<br>
                      <a href="mailto:OpenStack-dev@lists.openstack.org" target="_blank">OpenStack-dev@lists.openstack.org</a><br>
                      <a 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>
                    </div>
                  </blockquote>
                </div>
                <br>
                <br clear="all">
                <div><br>
                </div>
                -- <br>
                <div>
                  <div>Kevin Benton</div>
                </div>
              </div>
            </div>
            <div>
              <div>_______________________________________________</div>
              <div>OpenStack-dev mailing list</div>
              <div><a href="mailto:OpenStack-dev@lists.openstack.org" target="_blank">OpenStack-dev@lists.openstack.org</a></div>
              <div><a href="http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev" target="_blank">http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev</a></div>
            </div>
          </div>
        </span> </blockquote>
      <div> <br>
      </div>
      <br>
      <fieldset></fieldset>
      <br>
      <pre>_______________________________________________
OpenStack-dev mailing list
<a href="mailto:OpenStack-dev@lists.openstack.org" target="_blank">OpenStack-dev@lists.openstack.org</a>
<a href="http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev" target="_blank">http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev</a>
</pre>
    </blockquote>
    <br>
  </div></div></div>

<br>_______________________________________________<br>
OpenStack-dev mailing list<br>
<a href="mailto:OpenStack-dev@lists.openstack.org">OpenStack-dev@lists.openstack.org</a><br>
<a 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>