<div dir="auto">In most cases, cloud is just someone else's computer...</div><div class="gmail_extra"><br><div class="gmail_quote">On 2 Feb 2017 7:14 PM, "Chad Kelly" <<a href="mailto:chad@cpkws.com.au">chad@cpkws.com.au</a>> wrote:<br type="attribution"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
  
    
  
  <div bgcolor="#FFFFFF" text="#000000">
    <div class="m_-699787050440374874moz-cite-prefix">On 2/2/2017 6:50 PM, Mark Smith wrote:<br>
    </div>
    <blockquote type="cite">
      <div dir="auto">
        <div><br>
          <div class="gmail_extra"><br>
            <div class="gmail_quote">On 2 Feb. 2017 4:30 pm, "Chad
              Kelly" <<a href="mailto:chad@cpkws.com.au" target="_blank">chad@cpkws.com.au</a>>
              wrote:<br type="attribution">
              <blockquote class="m_-699787050440374874quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
                <div class="m_-699787050440374874quoted-text">On 2/2/2017 3:19 PM, <a href="mailto:ausnog-request@lists.ausnog.net" target="_blank">ausnog-request@lists.ausnog.ne<wbr>t</a>
                  wrote:<br>
                  <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
                    Of course when people say we have 2 core data
                    centers, this should imply no<br>
                    data center is allowed to run over 50% capacity.
                    It's odd/strange that 3<br>
                    active core data centers should sound so unorthodox,
                    yet this is the only<br>
                    way to assure you can run your DCs at 65% and handle
                    a DC going black. Begs<br>
                    the question why 4 active core DCs isn't standard
                    architecture for core<br>
                    national infrastructure (which would assure high
                    availability under 75%<br>
                    load), and 2x efficient in idle infrastructure.<br>
                  </blockquote>
                  <br>
                </div>
                I like your idea in theory.<br>
              </blockquote>
            </div>
          </div>
        </div>
        <div dir="auto"><br>
        </div>
        <div dir="auto">It's not theory. At one of the ISP's I've worked
          for we scaled out BRASes this way. As you add units of
          capacity, the required redundancy capacity required to cover a
          single unit failure reduces across all the other units. It
          works when you can divide your problem up into smaller
          sub-problems and distribute them across a pool.</div>
        <div dir="auto"><br>
        </div>
        <div dir="auto">The argument sometimes used against it is that
          it is more devices to manage. True, however that is tractable
          by using config templates, automation and device management
          platforms ("software defined networks"). The problems of
          managing many devices is not a new one if you've spent any
          time managing fleets of desktop PCs.</div>
        <div dir="auto"><br>
        </div>
        <div dir="auto">
          <div class="gmail_extra">
            <div class="gmail_quote">
              <blockquote class="m_-699787050440374874quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><br>
                <br>
                <br>
                <br>
                <br>
                <br>
                But building data centres costs money and a significant
                amount of it.<br>
              </blockquote>
            </div>
          </div>
        </div>
        <div dir="auto"><br>
        </div>
        <div dir="auto"><br>
        </div>
        <div dir="auto"><br>
        </div>
        <div dir="auto"><span style="font-family:sans-serif">You get
            what you pay for. If you need high availability, you need to
            be prepared to pay the price if it. If you can't afford the
            price, then it is likely your availability requirements are
            greater than they really need to be. Put a dollar cost
            against the consequence of a failure, and you might find you
            really do need to pay the price of the HA you want.</span><br>
        </div>
        <div dir="auto"><span style="font-family:sans-serif"><br>
          </span></div>
        <div dir="auto"><span style="font-family:sans-serif">If you
            can't afford to build DCs, you rent space in other people's
            to meet your availability goals.</span></div>
        <div dir="auto"><br>
        </div>
        <div dir="auto"><br>
        </div>
        <div dir="auto">
          <div class="gmail_extra">
            <div class="gmail_quote">
              <blockquote class="m_-699787050440374874quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><br>
                I remember when the Warrnambool exchange fire occurred,
                a discussion was had around fire suppression and the
                lack of it in a critical exchange for regional Victoria.<br>
                <br>
                Begs the question did they have appropriate levels of
                fire suppression equipment installed?<br>
                <br>
                No good having multiple lots of equipment if its not
                being protected from fire properly.<br>
              </blockquote>
            </div>
          </div>
        </div>
        <div dir="auto"><br>
        </div>
        <div dir="auto"><br>
        </div>
        <div dir="auto">A better architecture is one where a facility
          fire has a far smaller impact.</div>
        <div dir="auto"><br>
        </div>
        <div dir="auto">Your unit of expansion is your potential unit of
          failure. Larger units of expansion, larger consequences of
          failure.</div>
        <div dir="auto"><br>
        </div>
        <div dir="auto">Regards,</div>
        <div dir="auto">Mark.</div>
        <div dir="auto"><br>
        </div>
        <div dir="auto"><br>
        </div>
        <div dir="auto"><br>
        </div>
        <div dir="auto">
          <div class="gmail_extra">
            <div class="gmail_quote">
              <blockquote class="m_-699787050440374874quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
                <br>
                <snip></blockquote>
            </div>
          </div>
        </div>
      </div>
    </blockquote>
    <div dir="auto">
      <div dir="auto">
        <div class="gmail_extra">
          <div class="gmail_quote">
            <blockquote class="m_-699787050440374874quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">I get
              where you are coming from. <br>
            </blockquote>
          </div>
        </div>
      </div>
    </div>
    <p>A group of us were discussing the true meaning of cloud in terms
      of web hosting the other day, I basically said that if the server
      isn't setup with load balancing across multiple DC's that it isn't
      really proper cloud hosting. It needs to be setup with high
      availability. <br>
    </p>
    <p>A lot of providers use the term cloud when its not. <br>
    </p>
    <p><br>
    </p>
    <pre class="m_-699787050440374874moz-signature" cols="72">-- 
Chad Kelly 
Manager 
CPK Web Services 
web <a class="m_-699787050440374874moz-txt-link-abbreviated" href="http://www.cpkws.com.au" target="_blank">www.cpkws.com.au</a>
phone 03 9013 4853</pre>
  </div>

<br>______________________________<wbr>_________________<br>
AusNOG mailing list<br>
<a href="mailto:AusNOG@lists.ausnog.net">AusNOG@lists.ausnog.net</a><br>
<a href="http://lists.ausnog.net/mailman/listinfo/ausnog" rel="noreferrer" target="_blank">http://lists.ausnog.net/<wbr>mailman/listinfo/ausnog</a><br>
<br></blockquote></div></div>