<html>
  <head>
    <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
  </head>
  <body text="#000000" bgcolor="#FFFFFF">
    <br>
    <div class="moz-cite-prefix">On 09/04/2017 07:35 PM, Atin Mukherjee
      wrote:<br>
    </div>
    <blockquote type="cite"
cite="mid:CAGNCGH3f6S5uCs4Gti+Uc6QgBvxvZPuNS5invu_FbneCVpHjbg@mail.gmail.com">
      <div dir="ltr">
        <div>Ravi/Karthick,<br>
          <br>
        </div>
        If one of the self heal process is down, will the statstics
        heal-count command work?<br>
      </div>
    </blockquote>
    <br>
    No it doesn't seem to: glusterd stage-op phase fails because shd was
    down on that node and we error out.<br>
    FWIW, the error message <tt>"Gathering crawl statistics on volume
      GROUP-WORK has been unsuccessful on bricks that are down. Please
      check if all brick processes are running."</tt> is incorrect and
    once <a class="moz-txt-link-freetext" href="https://review.gluster.org/#/c/15724/">https://review.gluster.org/#/c/15724/</a> gets merged, you will get
    the correct error message like so:<br>
    <i><br>
    </i><tt>root@vm2 glusterfs]# gluster v heal testvol statistics </tt><tt><br>
    </tt><tt>Gathering crawl statistics on volume testvol has been
      unsuccessful:</tt><tt><br>
    </tt><tt> Staging failed on vm1. Error: Self-heal daemon is not
      running. Check self-heal daemon log file.</tt><i><br>
    </i><br>
    <br>
    -Ravi<br>
    <blockquote type="cite"
cite="mid:CAGNCGH3f6S5uCs4Gti+Uc6QgBvxvZPuNS5invu_FbneCVpHjbg@mail.gmail.com">
      <div class="gmail_extra"><br>
        <div class="gmail_quote">On Mon, Sep 4, 2017 at 7:24 PM,
          lejeczek <span dir="ltr">&lt;<a
              href="mailto:peljasz@yahoo.co.uk" target="_blank"
              moz-do-not-send="true">peljasz@yahoo.co.uk</a>&gt;</span>
          wrote:<br>
          <blockquote class="gmail_quote" style="margin:0 0 0
            .8ex;border-left:1px #ccc solid;padding-left:1ex">1) one
            peer, out of four, got separated from the network, from the
            rest of the cluster.<br>
            2) that unavailable(while it was unavailable) peer got
            detached with "gluster peer detach" command which succeeded,
            so now cluster comprise of three peers<br>
            3) Self-heal daemon (for some reason) does not start(with an
            attempt to restart glusted) on the peer which probed that
            fourth peer.<br>
            4) fourth unavailable peer is still up &amp; running but is
            inaccessible to other peers for network is disconnected,
            segmented. That peer's gluster status show peer is still in
            the cluster.<br>
            5) So, fourth peer's gluster(nor other processes) stack did
            not fail nor crushed, just network got, is disconnected.<br>
            6) peer status show ok &amp; connected for current three
            peers.<br>
            <br>
            This is third time when it happens to me, very same way:
            each time net-disjointed peer was brought back online then
            statistics &amp; details worked again.<br>
            <br>
            can you not reproduce it?<br>
            <br>
            $ gluster vol info QEMU-VMs<br>
            <br>
            Volume Name: QEMU-VMs<br>
            Type: Replicate<br>
            Volume ID: 8709782a-daa5-4434-a816-c4e0ae<wbr>f8fef2<br>
            Status: Started<br>
            Snapshot Count: 0<br>
            Number of Bricks: 1 x 3 = 3<br>
            Transport-type: tcp<br>
            Bricks:<br>
            Brick1: 10.5.6.32:/__.aLocalStorages/0<wbr>/0-GLUSTERs/0GLUSTER-QEMU-VMs<br>
            Brick2: 10.5.6.49:/__.aLocalStorages/0<wbr>/0-GLUSTERs/0GLUSTER-QEMU-VMs<br>
            Brick3: 10.5.6.100:/__.aLocalStorages/<wbr>0/0-GLUSTERs/0GLUSTER-QEMU-VMs<br>
            Options Reconfigured:<br>
            transport.address-family: inet<br>
            nfs.disable: on<br>
            storage.owner-gid: 107<br>
            storage.owner-uid: 107<br>
            performance.readdir-ahead: on<br>
            geo-replication.indexing: on<br>
            geo-replication.ignore-pid-che<wbr>ck: on<br>
            changelog.changelog: on<br>
            <br>
            $ gluster vol status QEMU-VMs<br>
            Status of volume: QEMU-VMs<br>
            Gluster process                       <wbr>      TCP Port 
            RDMA Port Online  Pid<br>
            ------------------------------<wbr>------------------------------<wbr>------------------<br>
            Brick 10.5.6.32:/__.aLocalStorages/0<wbr>/0-GLUS<br>
            TERs/0GLUSTER-QEMU-VMs        <wbr>              49156     0
            Y       9302<br>
            Brick 10.5.6.49:/__.aLocalStorages/0<wbr>/0-GLUS<br>
            TERs/0GLUSTER-QEMU-VMs        <wbr>              49156     0
            Y       7610<br>
            Brick 10.5.6.100:/__.aLocalStorages/<wbr>0/0-GLU<br>
            STERs/0GLUSTER-QEMU-VMs       <wbr>              49156     0
            Y       11013<br>
            Self-heal Daemon on localhost               N/A       N/A
            Y       3069276<br>
            Self-heal Daemon on 10.5.6.32               N/A       N/A
            Y       3315870<br>
            Self-heal Daemon on 10.5.6.49               N/A       N/A
            N       N/A  &lt;--- HERE<br>
            Self-heal Daemon on 10.5.6.17               N/A       N/A
            Y       5163<br>
            <br>
            Task Status of Volume QEMU-VMs<br>
            ------------------------------<wbr>------------------------------<wbr>------------------<br>
            There are no active volume tasks<br>
            <br>
            $ gluster vol heal QEMU-VMs statistics heal-count<br>
            Gathering count of entries to be healed on volume QEMU-VMs
            has been unsuccessful on bricks that are down. Please check
            if all brick processes are running.<span class=""><br>
              <br>
              <br>
              <br>
              On 04/09/17 11:47, Atin Mukherjee wrote:<br>
            </span>
            <blockquote class="gmail_quote" style="margin:0 0 0
              .8ex;border-left:1px #ccc solid;padding-left:1ex"><span
                class="">
                Please provide the output of gluster volume info,
                gluster volume status and gluster peer status.<br>
                <br>
              </span>
              <div>
                <div class="h5">
                  On Mon, Sep 4, 2017 at 4:07 PM, lejeczek &lt;<a
                    href="mailto:peljasz@yahoo.co.uk" target="_blank"
                    moz-do-not-send="true">peljasz@yahoo.co.uk</a>
                  &lt;mailto:<a href="mailto:peljasz@yahoo.co.uk"
                    target="_blank" moz-do-not-send="true">peljasz@yahoo.co.uk</a>&gt;&gt;
                  wrote:<br>
                  <br>
                      hi all<br>
                  <br>
                      this:<br>
                      $ vol heal $_vol info<br>
                      outputs ok and exit code is 0<br>
                      But if I want to see statistics:<br>
                      $ gluster vol heal $_vol statistics<br>
                      Gathering crawl statistics on volume GROUP-WORK
                  has<br>
                      been unsuccessful on bricks that are down. Please<br>
                      check if all brick processes are running.<br>
                  <br>
                      I suspect - gluster inability to cope with a
                  situation<br>
                      where one peer(which is not even a brick for a
                  single<br>
                      vol on the cluster!) is inaccessible to the rest
                  of<br>
                      cluster.<br>
                      I have not played with any other variations of
                  this<br>
                      case, eg. more than one peer goes down, etc.<br>
                      But I hope someone could try to replicate this
                  simple<br>
                      test case.<br>
                  <br>
                      Cluster and vols, when something like this
                  happens,<br>
                      seem accessible and as such "all" works, except
                  when<br>
                      you want more details.<br>
                      This also fails:<br>
                      $ gluster vol status $_vol detail<br>
                      Error : Request timed out<br>
                  <br>
                      My gluster(3.10.5-1.el7.x86_64) exhibits these<br>
                      symptoms every time one(at least) peers goes out
                  of<br>
                      the rest reach.<br>
                  <br>
                      maybe @devel can comment?<br>
                  <br>
                      many thanks, L.<br>
                      ______________________________<wbr>_________________<br>
                      Gluster-users mailing list<br>
                      <a href="mailto:Gluster-users@gluster.org"
                    target="_blank" moz-do-not-send="true">Gluster-users@gluster.org</a><br>
                </div>
              </div>
                  &lt;mailto:<a href="mailto:Gluster-users@gluster.org"
                target="_blank" moz-do-not-send="true">Gluster-users@gluster.<wbr>org</a>&gt;<br>
                  <a
                href="http://lists.gluster.org/mailman/listinfo/gluster-users"
                rel="noreferrer" target="_blank" moz-do-not-send="true">http://lists.gluster.org/mailm<wbr>an/listinfo/gluster-users</a><br>
                  &lt;<a
                href="http://lists.gluster.org/mailman/listinfo/gluster-users"
                rel="noreferrer" target="_blank" moz-do-not-send="true">http://lists.gluster.org/mail<wbr>man/listinfo/gluster-users</a>&gt;<br>
              <br>
              <br>
            </blockquote>
            <br>
          </blockquote>
        </div>
        <br>
      </div>
      <br>
      <fieldset class="mimeAttachmentHeader"></fieldset>
      <br>
      <pre wrap="">_______________________________________________
Gluster-users mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Gluster-users@gluster.org">Gluster-users@gluster.org</a>
<a class="moz-txt-link-freetext" href="http://lists.gluster.org/mailman/listinfo/gluster-users">http://lists.gluster.org/mailman/listinfo/gluster-users</a></pre>
    </blockquote>
    <br>
  </body>
</html>