<html>
  <head>
    <meta content="text/html; charset=windows-1252"
      http-equiv="Content-Type">
  </head>
  <body bgcolor="#FFFFFF" text="#000000">
    <br>
    <font face="Calibri">Instead of approaching it from the perspective
      of reporting on avail from a set or group of several entities, you
      could also approach it from the resulting data, which is 
      basically like a /summary of the avail type /percentages.  Or a
      /rollup to generate a /pie /chart.  I don't know, just some ideas,
      I'm OK with whatever you choose.<br>
      <br>
      <br>
    </font>
    <div class="moz-cite-prefix">On 9/15/2016 3:29 AM, Joel Takvorian
      wrote:<br>
    </div>
    <blockquote
cite="mid:CAJo5TFktaO=JN0__HwiP60rn00yStj_2NWEd1AadMxm7ym9gHQ@mail.gmail.com"
      type="cite">
      <div dir="ltr">According the Stephen we need to find another word
        than "aggregate" for the REST path, as it will be used for
        something else (and BTW, what is it for? Is there really no
        connection with multiple series aggregation?)
        <div><br>
        </div>
        <div>See <a moz-do-not-send="true"
            href="https://github.com/hawkular/hawkular-metrics/pull/597#r78810036">https://github.com/hawkular/hawkular-metrics/pull/597#r78810036</a><br>
          <div><br>
          </div>
          <div>So Jay proposed "<span
style="color:rgb(51,51,51);font-family:-apple-system,blinkmacsystemfont,&quot;segoe
              ui&quot;,roboto,helvetica,arial,sans-serif,&quot;apple
              color emoji&quot;,&quot;segoe ui emoji&quot;,&quot;segoe
              ui symbol&quot;;font-size:14px;line-height:21px">/set" or
              "/group"</span> earlier. I'm not very good at finding
            words but it's important especially for a REST API. If
            someone has a good idea please tell :)</div>
        </div>
        <div>I would also propose "/merged" or "/merge" (not sure about
          the "d")</div>
        <div><br>
        </div>
        <div>To sum up the idea here, is to take several availability
          series and merge them into a single one ; but the resulting
          series' data model is not exactly the same as an availability
          series since it contains a ratio map like {up: 0.9, down: 0.1}
          for each data point.</div>
        <div><br>
        </div>
        <div><br>
        </div>
      </div>
      <div class="gmail_extra"><br>
        <div class="gmail_quote">On Mon, Sep 5, 2016 at 8:51 AM, Joel
          Takvorian <span dir="ltr">&lt;<a moz-do-not-send="true"
              href="mailto:jtakvori@redhat.com" target="_blank">jtakvori@redhat.com</a>&gt;</span>
          wrote:<br>
          <blockquote class="gmail_quote" style="margin:0 0 0
            .8ex;border-left:1px #ccc solid;padding-left:1ex">
            <div dir="ltr">Yes, so there would be the (GET) <i>'/aggregate'</i> and
              the (POST) <i>'/aggregate/query'</i></div>
            <div class="HOEnZb">
              <div class="h5">
                <div class="gmail_extra"><br>
                  <div class="gmail_quote">On Fri, Sep 2, 2016 at 4:31
                    PM, Michael Burman <span dir="ltr">&lt;<a
                        moz-do-not-send="true"
                        href="mailto:miburman@redhat.com"
                        target="_blank">miburman@redhat.com</a>&gt;</span>
                    wrote:<br>
                    <blockquote class="gmail_quote" style="margin:0 0 0
                      .8ex;border-left:1px #ccc solid;padding-left:1ex">Hi,<br>
                      <br>
                      If you create some new endpoints, you must have
                      them in pure-REST-form in the first place. The
                      /query format should follow afterwards, if it's
                      necessary at all.<br>
                      <span><br>
                          - Micke<br>
                        <br>
                        ----- Original Message -----<br>
                      </span><span>From: "Joel Takvorian" &lt;<a
                          moz-do-not-send="true"
                          href="mailto:jtakvori@redhat.com"
                          target="_blank">jtakvori@redhat.com</a>&gt;<br>
                        To: "Discussions around Hawkular development"
                        &lt;<a moz-do-not-send="true"
                          href="mailto:hawkular-dev@lists.jboss.org"
                          target="_blank">hawkular-dev@lists.jboss.org</a>&gt;<br>
                      </span>
                      <div>
                        <div>Sent: Friday, September 2, 2016 4:16:46 PM<br>
                          Subject: Re: [Hawkular-dev] Availability
                          metrics: aggregate stats series<br>
                          <br>
                          Ok so i can use '/aggregate' for the time
                          being, but if someone has a better idea please
                          tell me.<br>
                          <br>
                          The input query would be basically same as the
                          existing query request on (POST) /raw/query :
                          list of ids, start, end, limit, order ; and
                          eventually 'tags' also<br>
                          <br>
                          Output would be a list of
                          DataPoint&lt;RatioMap&lt;Availabilit<wbr>yType&gt;&gt;<br>
                          RatioMap&lt;T&gt; being a map of &lt;T,
                          Double&gt; giving the ratio. Same data
                          structure can be used for aggregated string
                          metrics.<br>
                          <br>
                          Eventually the RatioMap object can contain an
                          integer " samples " that gives the number of
                          aggregated series on that point.<br>
                          <br>
                          On Fri, Sep 2, 2016 at 2:48 PM, Jay
                          Shaughnessy &lt; <a moz-do-not-send="true"
                            href="mailto:jshaughn@redhat.com"
                            target="_blank">jshaughn@redhat.com</a> &gt;
                          wrote:<br>
                          <br>
                          <br>
                          <br>
                          <br>
                          Joel, I'm not sure I was clear about the
                          legacy use of 'mixed'. MIXED was actually an
                          availability value, like UP or DOWN, but used
                          for a set of availability values that
                          contained a mix of UP, DOWN, etc. So, for a
                          group of resources if 7 were UP, 2 were DOWN
                          and 1 was UNKNOWN, that would have been mixed.<br>
                          <br>
                          I'm OK with your proposed endpoints but I
                          didn't want you to name it that way because of
                          history, because that's not how we used
                          'mixed'. If you preferred /aggregate, /set,
                          /group or whatever, go ahead and use that
                          instead.<br>
                          <br>
                          On 9/2/2016 7:58 AM, Joel Takvorian wrote:<br>
                          <br>
                          <br>
                          <br>
                          Also if I go on, it probably means create new
                          endpoints for mixed availability. For
                          instance:<br>
                          <br>
                          /availability/mixed<br>
                          <br>
                          Can you give an example of what you're
                          thinking for params and returned data?<br>
                          <br>
                          <br>
                          <br>
                          <br>
                          <br>
                          <br>
                          and in case we're interested in getting stats
                          of a mixed availability, something like:<br>
                          /availability/mixed/stats<br>
                          <br>
                          Parallel changes would be done for string data
                          type.<br>
                          <br>
                          On Fri, Sep 2, 2016 at 11:21 AM, Thomas Heute
                          &lt; <a moz-do-not-send="true"
                            href="mailto:theute@redhat.com"
                            target="_blank">theute@redhat.com</a> &gt;
                          wrote:<br>
                          <br>
                          <br>
                          <br>
                          That sounds like a pretty useful feature to
                          me.<br>
                          <br>
                          Is there any blocker or can Joel move forward
                          ?<br>
                          <br>
                          On Tue, Aug 30, 2016 at 6:14 PM, Jay
                          Shaughnessy &lt; <a moz-do-not-send="true"
                            href="mailto:jshaughn@redhat.com"
                            target="_blank">jshaughn@redhat.com</a> &gt;
                          wrote:<br>
                          <br>
                          <br>
                          <br>
                          <br>
                          Certainly you can't base overall application
                          health on some random aggregate avail. It's an
                          indicator, like so many other things, of where
                          problems could lie. I don't think there is
                          anything wrong with a percentage as a quick
                          indicator, from there you'd drill down as
                          needed. As Joel says, it depends also on what
                          you choose to aggregate. When your URL
                          response times are degraded, likely firing
                          alerts, you then want to understand why.
                          Aggregate avails could help answer the
                          questions. There's always examples of how to
                          misuse tools, a hammer can easily break your
                          thumb, doesn't mean hammers are bad.<br>
                          <br>
                          <br>
                          On 8/30/2016 11:45 AM, Joel Takvorian wrote:<br>
                          <br>
                          <br>
                          <br>
                          Just a precision because I'm not sure if I was
                          clear on that: the idea is to mix series based
                          on a list of ids, or tags. Not *everything*<br>
                          <br>
                          On Tue, Aug 30, 2016 at 5:39 PM, Joel
                          Takvorian &lt; <a moz-do-not-send="true"
                            href="mailto:jtakvori@redhat.com"
                            target="_blank">jtakvori@redhat.com</a> &gt;
                          wrote:<br>
                          <br>
                          <br>
                          <br>
                          I agree that you won't want to mix everything,
                          but you can still adopt some groupings that
                          are meaningful, for instance group all
                          front-end servers into a front-end
                          availability series, and all back-ends into
                          another series.<br>
                          <br>
                          Moreover, once you get all the availability as
                          ratio, it's easy to map to a binary
                          availability if it's what you're looking for.
                          The REST api will provide the data, then it's
                          up to you to display what is the most
                          relevant. I think ratio datapoints is an
                          easy-to-use, yet complete, information.<br>
                          <br>
                          Joel<br>
                          <br>
                          On Tue, Aug 30, 2016 at 5:16 PM, Michael
                          Burman &lt; <a moz-do-not-send="true"
                            href="mailto:miburman@redhat.com"
                            target="_blank">miburman@redhat.com</a> &gt;
                          wrote:<br>
                          <br>
                          <br>
                          Hi,<br>
                          <br>
                          So if I have 8 MySQLs, 4 primaries, 4
                          replicas. One primary is down and the replica
                          of that set is down as well. I request
                          Availability of my datastore and I get 80% UP.
                          If I had two replicas down instead, I would
                          get 80% UP. There's a huge difference in these
                          scenarios.<br>
                          <br>
                          I'm not a fan of percents for that simple
                          reason. Is my service up? Yes, it's 99% up,
                          only all the front-end servers are down.. ugh.<br>
                          <br>
                          - Micke<br>
                          <br>
                          ----- Original Message -----<br>
                          From: "John Sanda" &lt; <a
                            moz-do-not-send="true"
                            href="mailto:jsanda@redhat.com"
                            target="_blank">jsanda@redhat.com</a> &gt;<br>
                          To: "Discussions around Hawkular development"
                          &lt; <a moz-do-not-send="true"
                            href="mailto:hawkular-dev@lists.jboss.org"
                            target="_blank">hawkular-dev@lists.jboss.org</a>
                          &gt;<br>
                          Sent: Tuesday, August 30, 2016 4:11:07 PM<br>
                          Subject: Re: [Hawkular-dev] Availability
                          metrics: aggregate stats series<br>
                          <br>
                          I like the idea of aggregated availabilities,
                          but I don’t know that it can easily be
                          simplified to up/down. Let’s say we have 3
                          Cassandra nodes deployed with
                          replication_factor = 1. If one node goes down
                          we are at 66% availability.<br>
                          <br>
                          &gt; On Aug 29, 2016, at 3:24 AM, Joel
                          Takvorian &lt; <a moz-do-not-send="true"
                            href="mailto:jtakvori@redhat.com"
                            target="_blank">jtakvori@redhat.com</a> &gt;
                          wrote:<br>
                          &gt;<br>
                          &gt; Hello all,<br>
                          &gt;<br>
                          &gt; I'm still aiming to add some features to
                          the grafana plugin. I've started to integrate
                          availabilities, but now I'm facing a problem
                          when it comes to show aggregated
                          availabilities ; for example think about an
                          OpenShift pod that is scaled up to several
                          instances.<br>
                          &gt;<br>
                          &gt; Since availability is basically "up" or
                          "down" (or, to simplify with the other states
                          such as "unknown", say it's either "up" or
                          "non-up"), I propose to add this new feature:
                          availability stats with aggregation. The call
                          would be parameterized with an aggregation
                          method, which would be either "all of" or "any
                          of": with "all of" we consider that the
                          aggregated series is UP when all its parts are
                          UP.<br>
                          &gt;<br>
                          &gt; It would require a new endpoint since the
                          AvailabilityHandler currently only expose
                          stats queries with metric id as query
                          parameter - not suitable for multiple metrics.<br>
                          &gt;<br>
                          &gt; Any objection or remark for this feature?<br>
                          &gt;<br>
                          &gt; Joel<br>
                          &gt; ______________________________<wbr>_________________<br>
                          &gt; hawkular-dev mailing list<br>
                          &gt; <a moz-do-not-send="true"
                            href="mailto:hawkular-dev@lists.jboss.org"
                            target="_blank">hawkular-dev@lists.jboss.org</a><br>
                          &gt; <a moz-do-not-send="true"
                            href="https://lists.jboss.org/mailman/listinfo/hawkular-dev"
                            rel="noreferrer" target="_blank">https://lists.jboss.org/mailma<wbr>n/listinfo/hawkular-dev</a><br>
                          <br>
                          <br>
                          ______________________________<wbr>_________________<br>
                          hawkular-dev mailing list<br>
                          <a moz-do-not-send="true"
                            href="mailto:hawkular-dev@lists.jboss.org"
                            target="_blank">hawkular-dev@lists.jboss.org</a><br>
                          <a moz-do-not-send="true"
                            href="https://lists.jboss.org/mailman/listinfo/hawkular-dev"
                            rel="noreferrer" target="_blank">https://lists.jboss.org/mailma<wbr>n/listinfo/hawkular-dev</a><br>
                          <br>
                          ______________________________<wbr>_________________<br>
                          hawkular-dev mailing list<br>
                          <a moz-do-not-send="true"
                            href="mailto:hawkular-dev@lists.jboss.org"
                            target="_blank">hawkular-dev@lists.jboss.org</a><br>
                          <a moz-do-not-send="true"
                            href="https://lists.jboss.org/mailman/listinfo/hawkular-dev"
                            rel="noreferrer" target="_blank">https://lists.jboss.org/mailma<wbr>n/listinfo/hawkular-dev</a><br>
                          <br>
                          <br>
                          <br>
                          <br>
                          ______________________________<wbr>_________________<br>
                          hawkular-dev mailing list <a
                            moz-do-not-send="true"
                            href="mailto:hawkular-dev@lists.jboss.org"
                            target="_blank">hawkular-dev@lists.jboss.org</a>
                          <a moz-do-not-send="true"
                            href="https://lists.jboss.org/mailman/listinfo/hawkular-dev"
                            rel="noreferrer" target="_blank">https://lists.jboss.org/mailma<wbr>n/listinfo/hawkular-dev</a><br>
                          ______________________________<wbr>_________________
                          hawkular-dev mailing list <a
                            moz-do-not-send="true"
                            href="mailto:hawkular-dev@lists.jboss.org"
                            target="_blank">hawkular-dev@lists.jboss.org</a>
                          <a moz-do-not-send="true"
                            href="https://lists.jboss.org/mailman/listinfo/hawkular-dev"
                            rel="noreferrer" target="_blank">https://lists.jboss.org/mailma<wbr>n/listinfo/hawkular-dev</a><br>
                          ______________________________<wbr>_________________
                          hawkular-dev mailing list <a
                            moz-do-not-send="true"
                            href="mailto:hawkular-dev@lists.jboss.org"
                            target="_blank">hawkular-dev@lists.jboss.org</a>
                          <a moz-do-not-send="true"
                            href="https://lists.jboss.org/mailman/listinfo/hawkular-dev"
                            rel="noreferrer" target="_blank">https://lists.jboss.org/mailma<wbr>n/listinfo/hawkular-dev</a><br>
                          ______________________________<wbr>_________________<br>
                          hawkular-dev mailing list <a
                            moz-do-not-send="true"
                            href="mailto:hawkular-dev@lists.jboss.org"
                            target="_blank">hawkular-dev@lists.jboss.org</a>
                          <a moz-do-not-send="true"
                            href="https://lists.jboss.org/mailman/listinfo/hawkular-dev"
                            rel="noreferrer" target="_blank">https://lists.jboss.org/mailma<wbr>n/listinfo/hawkular-dev</a><br>
                          <br>
                          ______________________________<wbr>_________________<br>
                          hawkular-dev mailing list<br>
                          <a moz-do-not-send="true"
                            href="mailto:hawkular-dev@lists.jboss.org"
                            target="_blank">hawkular-dev@lists.jboss.org</a><br>
                          <a moz-do-not-send="true"
                            href="https://lists.jboss.org/mailman/listinfo/hawkular-dev"
                            rel="noreferrer" target="_blank">https://lists.jboss.org/mailma<wbr>n/listinfo/hawkular-dev</a><br>
                          <br>
                          <br>
                          <br>
                          ______________________________<wbr>_________________<br>
                          hawkular-dev mailing list<br>
                          <a moz-do-not-send="true"
                            href="mailto:hawkular-dev@lists.jboss.org"
                            target="_blank">hawkular-dev@lists.jboss.org</a><br>
                          <a moz-do-not-send="true"
                            href="https://lists.jboss.org/mailman/listinfo/hawkular-dev"
                            rel="noreferrer" target="_blank">https://lists.jboss.org/mailma<wbr>n/listinfo/hawkular-dev</a><br>
                          <br>
                          ______________________________<wbr>_________________<br>
                          hawkular-dev mailing list<br>
                          <a moz-do-not-send="true"
                            href="mailto:hawkular-dev@lists.jboss.org"
                            target="_blank">hawkular-dev@lists.jboss.org</a><br>
                          <a moz-do-not-send="true"
                            href="https://lists.jboss.org/mailman/listinfo/hawkular-dev"
                            rel="noreferrer" target="_blank">https://lists.jboss.org/mailma<wbr>n/listinfo/hawkular-dev</a><br>
                        </div>
                      </div>
                    </blockquote>
                  </div>
                  <br>
                </div>
              </div>
            </div>
          </blockquote>
        </div>
        <br>
      </div>
      <br>
      <fieldset class="mimeAttachmentHeader"></fieldset>
      <br>
      <pre wrap="">_______________________________________________
hawkular-dev mailing list
<a class="moz-txt-link-abbreviated" href="mailto:hawkular-dev@lists.jboss.org">hawkular-dev@lists.jboss.org</a>
<a class="moz-txt-link-freetext" href="https://lists.jboss.org/mailman/listinfo/hawkular-dev">https://lists.jboss.org/mailman/listinfo/hawkular-dev</a>
</pre>
    </blockquote>
    <br>
  </body>
</html>