<html>
  <head>
    <meta content="text/html; charset=ISO-8859-1"
      http-equiv="Content-Type">
  </head>
  <body text="#000000" bgcolor="#FFFFFF">
    <div class="moz-cite-prefix">Hi,<br>
      <br>
      Thanks for you reply.<br>
      <br>
      If I modify the configuration like this on the global_common : <br>
      <br>
      global {<br>
      &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; usage-count yes;<br>
      &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; # minor-count dialog-refresh disable-ip-verification<br>
      }<br>
      common {<br>
      &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; protocol C;<br>
      &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; handlers {<br>
      &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; # The following 3 handlers were disabled due to
      #576511.<br>
      &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; # Please check the DRBD manual and enable them, if
      they make sense in your setup.<br>
      &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; # pri-on-incon-degr
      "/usr/lib/drbd/notify-pri-on-incon-degr.sh;
      /usr/lib/drbd/notify-emergency-reboot.sh; echo b &gt;
      /proc/sysrq-trigger ; reboot -f";<br>
      &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; # pri-lost-after-sb
      "/usr/lib/drbd/notify-pri-lost-after-sb.sh;
      /usr/lib/drbd/notify-emergency-reboot.sh; echo b &gt;
      /proc/sysrq-trigger ; reboot -f";<br>
      &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; # local-io-error
      "/usr/lib/drbd/notify-io-error.sh;
      /usr/lib/drbd/notify-emergency-shutdown.sh; echo o &gt;
      /proc/sysrq-trigger ; halt -f";<br>
      &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; # fence-peer "/usr/lib/drbd/crm-fence-peer.sh";<br>
      &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; # split-brain "/usr/lib/drbd/notify-split-brain.sh
      root";<br>
      &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; # out-of-sync "/usr/lib/drbd/notify-out-of-sync.sh
      root";<br>
      &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; # before-resync-target
      "/usr/lib/drbd/snapshot-resync-target-lvm.sh -p 15 -- -c 16k";<br>
      &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; # after-resync-target
      /usr/lib/drbd/unsnapshot-resync-target-lvm.sh;<br>
      &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; }<br>
      &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; startup {<br>
      &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; # wfc-timeout degr-wfc-timeout
      outdated-wfc-timeout wait-after-sb<br>
      &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; }<br>
      &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; disk {<br>
      &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; # on-io-error fencing use-bmbv no-disk-barrier
      no-disk-flushes<br>
      &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; # no-disk-drain no-md-flushes max-bio-bvecs<br>
      &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; }<br>
      &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; net {<br>
      &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; ko-count 2<br>
      &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; timeout 50<br>
      &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; # sndbuf-size rcvbuf-size timeout connect-int
      ping-int ping-timeout max-buffers<br>
      &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; # max-epoch-size ko-count allow-two-primaries
      cram-hmac-alg shared-secret<br>
      &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; # after-sb-0pri after-sb-1pri after-sb-2pri
      data-integrity-alg no-tcp-cork<br>
      &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; }<br>
      <br>
      &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; syncer {<br>
      &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; # rate after al-extents use-rle cpu-mask
      verify-alg csums-alg<br>
      &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; }<br>
      }<br>
      <br>
      If I make this config one the secondary node, I can have a proper
      disconnection on the slave when we ave HW problems like on my
      previous post ?<br>
      <br>
      Thanks<br>
      <br>
      Matthieu Lejeune<br>
      <br>
      <br>
      <br>
      <br>
      Le 5/03/14 11:32, Philip Gaw a &eacute;crit&nbsp;:<br>
    </div>
    <blockquote cite="mid:5316FD47.3050101@darktech.org.uk" type="cite">
      <meta content="text/html; charset=ISO-8859-1"
        http-equiv="Content-Type">
      <blockquote cite="mid:5316F0DA.7050704@darktech.org.uk"
        type="cite">Hi Matthieu,<br>
        <br>
        <div class="moz-cite-prefix">On 05/03/2014 07:29, Matthieu
          Lejeune wrote:<br>
        </div>
        <blockquote cite="mid:5316D24E.8040201@exxoss.com" type="cite">
          <meta http-equiv="content-type" content="text/html;
            charset=ISO-8859-1">
          Hi all,<br>
          <br>
          I had a problem this night with a DRBD Primary/Slave.<br>
          <br>
          <br>
          The slave experienced a hardware issue (LSI controller
          freezed).<br>
          It seems the master hold I/O waiting for the slave to respond
          until timeout.<br>
          <br>
          <br>
          This caused all targets exported trough infiniband to be
          disconnected from the master.<br>
          <br>
          <br>
          So, practically, the master stop responding due to a failure
          on the slave.<br>
          <br>
          I had to hard reboot (power cycle) the slave because UDEV
          wasn't responding and did not allow normal reboot.<br>
          After slave reboot, drdb did reconnect. It was in status
          pri/sec uptodate/uptodate.<br>
          But the LSI controller immediatly timeout causing the same
          issue a second time.<br>
          <br>
          <br>
          How can we prevent issue on the slave to impact the master ?<br>
          <br>
        </blockquote>
        have a look at ko-count <br>
        <br>
        <dt style="color: rgb(0, 0, 0); font-family: sans-serif;
          font-size: 13px; font-style: normal; font-variant: normal;
          font-weight: normal; letter-spacing: normal; line-height:
          normal; orphans: auto; text-align: start; text-indent: 0px;
          text-transform: none; white-space: normal; widows: auto;
          word-spacing: 0px; -webkit-text-stroke-width: 0px;"><span
            class="term"><code class="option">ko-count<span
                class="Apple-converted-space">&nbsp;</span><em
                class="replaceable"><code>number</code></em></code></span></dt>
        <dd style="color: rgb(0, 0, 0); font-family: sans-serif;
          font-size: 13px; font-style: normal; font-variant: normal;
          font-weight: normal; letter-spacing: normal; line-height:
          normal; orphans: auto; text-align: start; text-indent: 0px;
          text-transform: none; white-space: normal; widows: auto;
          word-spacing: 0px; -webkit-text-stroke-width: 0px;">
          <p>In case the secondary node fails to complete a single write
            request for<span class="Apple-converted-space">&nbsp;</span><em
              class="replaceable"><code>count</code></em><span
              class="Apple-converted-space">&nbsp;</span>times the<span
              class="Apple-converted-space">&nbsp;</span><em
              class="replaceable"><code>timeout</code></em>, it is
            expelled from the cluster. (I.e. the primary node goes into<span
              class="Apple-converted-space">&nbsp;</span><code class="option">StandAlone</code><span
              class="Apple-converted-space">&nbsp;</span>mode.) The default
            value is 0, which disables this feature.</p>
        </dd>
        <dt><br>
        </dt>
        <a moz-do-not-send="true" class="moz-txt-link-freetext"
          href="http://www.drbd.org/users-guide/re-drbdconf.html">http://www.drbd.org/users-guide/re-drbdconf.html</a><br>
        <br>
        <blockquote cite="mid:5316D24E.8040201@exxoss.com" type="cite">
          <br>
          Thank you.<br>
          Matthieu Lejeune<br>
          <br>
          <br>
          drbd8-utils : &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 2:8.3.13-2&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
          amd64&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; RAID 1 over tcp/ip for Linux utilities<br>
          Debian : <br>
          root@ifprdstor8a:~/trunk# cat /proc/version <br>
          Linux version 3.2.0-4-amd64 (<a moz-do-not-send="true"
            class="moz-txt-link-abbreviated"
            href="mailto:debian-kernel@lists.debian.org">debian-kernel@lists.debian.org</a>)
          (gcc version 4.6.3 (Debian 4.6.3-14) ) #1 SMP Debian 3.2.51-1<br>
          root@ifprdstor8a:~/trunk# <br>
          <br>
          We are using the scst/srpt with the Trunk version of the 7
          January 2014<br>
          <br>
          I give you the config :<br>
          <b>drbd global : </b><b><br>
          </b><br>
          <a moz-do-not-send="true" class="moz-txt-link-abbreviated"
            href="mailto:root@ifprdstor8a:/etc/drbd.d#">root@ifprdstor8a:/etc/drbd.d#</a>
          cat global_common.conf <br>
          global {<br>
          &nbsp;&nbsp;&nbsp; usage-count yes;<br>
          &nbsp;&nbsp;&nbsp; # minor-count dialog-refresh disable-ip-verification<br>
          }<br>
          <br>
          common {<br>
          &nbsp;&nbsp;&nbsp; protocol C;<br>
          <br>
          &nbsp;&nbsp;&nbsp; handlers {<br>
          &nbsp;&nbsp;&nbsp; &nbsp;&nbsp;&nbsp; # The following 3 handlers were disabled due to
          #576511.<br>
          &nbsp;&nbsp;&nbsp; &nbsp;&nbsp;&nbsp; # Please check the DRBD manual and enable them, if
          they make sense in your setup.<br>
          &nbsp;&nbsp;&nbsp; &nbsp;&nbsp;&nbsp; # pri-on-incon-degr
          "/usr/lib/drbd/notify-pri-on-incon-degr.sh;
          /usr/lib/drbd/notify-emergency-reboot.sh; echo b &gt;
          /proc/sysrq-trigger ; reboot -f";<br>
          &nbsp;&nbsp;&nbsp; &nbsp;&nbsp;&nbsp; # pri-lost-after-sb
          "/usr/lib/drbd/notify-pri-lost-after-sb.sh;
          /usr/lib/drbd/notify-emergency-reboot.sh; echo b &gt;
          /proc/sysrq-trigger ; reboot -f";<br>
          &nbsp;&nbsp;&nbsp; &nbsp;&nbsp;&nbsp; # local-io-error "/usr/lib/drbd/notify-io-error.sh;
          /usr/lib/drbd/notify-emergency-shutdown.sh; echo o &gt;
          /proc/sysrq-trigger ; halt -f";<br>
          <br>
          &nbsp;&nbsp;&nbsp; &nbsp;&nbsp;&nbsp; # fence-peer "/usr/lib/drbd/crm-fence-peer.sh";<br>
          &nbsp;&nbsp;&nbsp; &nbsp;&nbsp;&nbsp; # split-brain "/usr/lib/drbd/notify-split-brain.sh
          root";<br>
          &nbsp;&nbsp;&nbsp; &nbsp;&nbsp;&nbsp; # out-of-sync "/usr/lib/drbd/notify-out-of-sync.sh
          root";<br>
          &nbsp;&nbsp;&nbsp; &nbsp;&nbsp;&nbsp; # before-resync-target
          "/usr/lib/drbd/snapshot-resync-target-lvm.sh -p 15 -- -c 16k";<br>
          &nbsp;&nbsp;&nbsp; &nbsp;&nbsp;&nbsp; # after-resync-target
          /usr/lib/drbd/unsnapshot-resync-target-lvm.sh;<br>
          &nbsp;&nbsp;&nbsp; }<br>
          <br>
          &nbsp;&nbsp;&nbsp; startup {<br>
          &nbsp;&nbsp;&nbsp; &nbsp;&nbsp;&nbsp; # wfc-timeout degr-wfc-timeout outdated-wfc-timeout
          wait-after-sb<br>
          &nbsp;&nbsp;&nbsp; }<br>
          <br>
          &nbsp;&nbsp;&nbsp; disk {<br>
          &nbsp;&nbsp;&nbsp; &nbsp;&nbsp;&nbsp; # on-io-error fencing use-bmbv no-disk-barrier
          no-disk-flushes<br>
          &nbsp;&nbsp;&nbsp; &nbsp;&nbsp;&nbsp; # no-disk-drain no-md-flushes max-bio-bvecs<br>
          &nbsp;&nbsp;&nbsp; }<br>
          <br>
          &nbsp;&nbsp;&nbsp; net {<br>
          &nbsp;&nbsp;&nbsp; &nbsp;&nbsp;&nbsp; # sndbuf-size rcvbuf-size timeout connect-int ping-int
          ping-timeout max-buffers<br>
          &nbsp;&nbsp;&nbsp; &nbsp;&nbsp;&nbsp; # max-epoch-size ko-count allow-two-primaries
          cram-hmac-alg shared-secret<br>
          &nbsp;&nbsp;&nbsp; &nbsp;&nbsp;&nbsp; # after-sb-0pri after-sb-1pri after-sb-2pri
          data-integrity-alg no-tcp-cork<br>
          &nbsp;&nbsp;&nbsp; }<br>
          <br>
          &nbsp;&nbsp;&nbsp; syncer {<br>
          &nbsp;&nbsp;&nbsp; &nbsp;&nbsp;&nbsp; # rate after al-extents use-rle cpu-mask verify-alg
          csums-alg<br>
          &nbsp;&nbsp;&nbsp; }<br>
          }<br>
          <br>
          <b>Ressource Configuration : </b><br>
          <br>
          <a moz-do-not-send="true" class="moz-txt-link-abbreviated"
            href="mailto:root@ifprdstor8a:/etc/drbd.d#">root@ifprdstor8a:/etc/drbd.d#</a>
          cat DSA801.res <br>
          resource DSA801 {<br>
          &nbsp; protocol C;<br>
          <br>
          &nbsp; startup {<br>
          &nbsp;&nbsp;&nbsp; wfc-timeout 0;<br>
          &nbsp; }<br>
          <br>
          &nbsp; disk {<br>
          &nbsp;&nbsp;&nbsp; on-io-error detach;<br>
          &nbsp; }<br>
          <br>
          &nbsp; syncer {<br>
          &nbsp;&nbsp;&nbsp; rate 400M;<br>
          &nbsp;&nbsp;&nbsp; verify-alg md5;<br>
          &nbsp; }<br>
          <br>
          &nbsp; on ifprdstor8a {<br>
          &nbsp;&nbsp;&nbsp; device&nbsp;&nbsp;&nbsp; /dev/drbd1;<br>
          &nbsp;&nbsp;&nbsp; disk&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; /dev/sda;<br>
          &nbsp;&nbsp;&nbsp; address&nbsp;&nbsp; 10.13.1.5:7788;<br>
          &nbsp;&nbsp;&nbsp; meta-disk internal;<br>
          &nbsp; }<br>
          <br>
          &nbsp; on ifprdstor8b {<br>
          &nbsp;&nbsp;&nbsp; device&nbsp;&nbsp;&nbsp; /dev/drbd1;<br>
          &nbsp;&nbsp;&nbsp; disk&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; /dev/sda;<br>
          &nbsp;&nbsp;&nbsp; address&nbsp;&nbsp; 10.13.1.6:7788;<br>
          &nbsp;&nbsp;&nbsp; meta-disk internal;<br>
          &nbsp; }<br>
          }<br>
          <br>
        </blockquote>
      </blockquote>
      <br>
      <br>
      <fieldset class="mimeAttachmentHeader"></fieldset>
      <br>
      <pre wrap="">_______________________________________________
drbd-user mailing list
<a class="moz-txt-link-abbreviated" href="mailto:drbd-user@lists.linbit.com">drbd-user@lists.linbit.com</a>
<a class="moz-txt-link-freetext" href="http://lists.linbit.com/mailman/listinfo/drbd-user">http://lists.linbit.com/mailman/listinfo/drbd-user</a>
</pre>
    </blockquote>
    <br>
    <div class="moz-signature"><br>
      &nbsp;
    </div>
  </body>
</html>