<html>
  <head>
    <meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
  </head>
  <body>
    <p><br>
    </p>
    <div class="moz-cite-prefix">On 12/12/19 4:01 am, WK wrote:<br>
    </div>
    <blockquote type="cite"
      cite="mid:77d643fc-ad9e-d5d2-6815-65512609c2ea@bneit.com">
      <meta http-equiv="content-type" content="text/html; charset=UTF-8">
      <p>&lt;BUMP&gt; so I can get some sort of resolution on the issue
        (i.e. is it hardware, Gluster etc)<br>
      </p>
      <p>I guess what I really need to know is <br>
      </p>
      <p>1) Node 2 complains that it cant reach node 1 and node 3.  If
        this was an OS/Hardware networking issue and not internal to
        Gluster , then why didn't node1 and node3 have error message
        complaining about not reaching node2 <br>
      </p>
    </blockquote>
    <tt>[2019-12-05 22:00:43.739804] C
      [rpc-clnt-ping.c:155:rpc_clnt_ping_timer_expired]
      0-GL1image-client-2: server 10.255.1.1:49153 has not responded in
      the last 21 seconds, disconnecting.</tt><tt><br>
    </tt><tt> [2019-12-05 22:00:43.757095] C
      [rpc-clnt-ping.c:155:rpc_clnt_ping_timer_expired]
      0-GL1image-client-1: server 10.255.1.3:49152 has not responded in
      the last 21 seconds, disconnecting.</tt><tt><br>
    </tt><tt> [2019-12-05 22:00:43.757191] I [MSGID: 114018]
      [client.c:2323:client_rpc_notify] 0-GL1image-client-2:
      disconnected from GL1image-client-2. Client process will keep
      trying to connect to glusterd until brick's port is available</tt><tt><br>
    </tt><tt> [2019-12-05 22:00:43.757246] I [MSGID: 114018]
      [client.c:2323:client_rpc_notify] 0-GL1image-client-1:
      disconnected from GL1image-client-1. Client process will keep
      trying to connect to glusterd until brick's port is available</tt><tt><br>
    </tt>
    <p><tt> [2019-12-05 22:00:43.757266] W [MSGID: 108001]
        [afr-common.c:5608:afr_notify] 0-GL1image-replicate-0:
        Client-quorum is not met</tt></p>
    <p>This seems to indicate the mount on node 2 cannot reach 2 bricks.
      If quorum is not met, you will get ENOTCONN on the mount. Maybe
      check if the mount is still disconnected from the bricks (either
      statedump or looking at the .meta folder)?<br>
    </p>
    <blockquote type="cite"
      cite="mid:77d643fc-ad9e-d5d2-6815-65512609c2ea@bneit.com">
      <p> </p>
      <div class="moz-forward-container">2) how significant is it that
        the node was running 6.5 while node 1 and node 2 were running
        6.4</div>
    </blockquote>
    <p>Minor versions should be fine but it is always a good idea to
      have all nodes on the same version.</p>
    HTH,<br>
    Ravi<br>
    <blockquote type="cite"
      cite="mid:77d643fc-ad9e-d5d2-6815-65512609c2ea@bneit.com">
      <div class="moz-forward-container"><br>
      </div>
      <div class="moz-forward-container">-wk<br>
      </div>
      <div class="moz-forward-container"><br>
        -------- Forwarded Message --------
        <table class="moz-email-headers-table" cellspacing="0"
          cellpadding="0" border="0">
          <tbody>
            <tr>
              <th valign="BASELINE" nowrap="nowrap" align="RIGHT">Subject:
              </th>
              <td>VM freeze issue on simple gluster setup.</td>
            </tr>
            <tr>
              <th valign="BASELINE" nowrap="nowrap" align="RIGHT">Date:
              </th>
              <td>Thu, 5 Dec 2019 16:23:35 -0800</td>
            </tr>
            <tr>
              <th valign="BASELINE" nowrap="nowrap" align="RIGHT">From:
              </th>
              <td>WK <a class="moz-txt-link-rfc2396E"
                  href="mailto:wkmail@bneit.com" moz-do-not-send="true">&lt;wkmail@bneit.com&gt;</a></td>
            </tr>
            <tr>
              <th valign="BASELINE" nowrap="nowrap" align="RIGHT">To: </th>
              <td>Gluster Users <a class="moz-txt-link-rfc2396E"
                  href="mailto:gluster-users@gluster.org"
                  moz-do-not-send="true">&lt;gluster-users@gluster.org&gt;</a></td>
            </tr>
          </tbody>
        </table>
        <br>
        <br>
        I have a replica2+arbiter setup that is used for VMs.<br>
        <br>
        ip #.1 is the arb<br>
        <br>
        ip #.2 and #.3 are the kvm hosts.<br>
        <br>
        Two Volumes are involved and its gluster 6.5/Ubuntu 18.4/fuse
        The Gluster networking uses a  two ethernet card
        teamd/round-robin setup which *should* have stayed up if one of
        the ports had failed.<br>
        <br>
        I just had a number of VMs go Read-Only due to the below
        communication failure at 22:00 but only on kvm host  #2<br>
        <br>
        VMs on the same gluster volumes on kvm host 3 were unaffected.<br>
        <br>
        The logs on host #2 show the following:<br>
        <br>
        [2019-12-05 22:00:43.739804] C
        [rpc-clnt-ping.c:155:rpc_clnt_ping_timer_expired]
        0-GL1image-client-2: server 10.255.1.1:49153 has not responded
        in the last 21 seconds, disconnecting.<br>
        [2019-12-05 22:00:43.757095] C
        [rpc-clnt-ping.c:155:rpc_clnt_ping_timer_expired]
        0-GL1image-client-1: server 10.255.1.3:49152 has not responded
        in the last 21 seconds, disconnecting.<br>
        [2019-12-05 22:00:43.757191] I [MSGID: 114018]
        [client.c:2323:client_rpc_notify] 0-GL1image-client-2:
        disconnected from GL1image-client-2. Client process will keep
        trying to connect to glusterd until brick's port is available<br>
        [2019-12-05 22:00:43.757246] I [MSGID: 114018]
        [client.c:2323:client_rpc_notify] 0-GL1image-client-1:
        disconnected from GL1image-client-1. Client process will keep
        trying to connect to glusterd until brick's port is available<br>
        [2019-12-05 22:00:43.757266] W [MSGID: 108001]
        [afr-common.c:5608:afr_notify] 0-GL1image-replicate-0:
        Client-quorum is not met<br>
        [2019-12-05 22:00:43.790639] E
        [rpc-clnt.c:346:saved_frames_unwind] (--&gt;
/usr/lib/x86_64-linux-gnu/libglusterfs.so.0(_gf_log_callingfn+0x139)[0x7f030d045f59]
        (--&gt;
        /usr/lib/x86_64-linux-gnu/libgfrpc.so.0(+0xcbb0)[0x7f030cdf0bb0]
        (--&gt;
        /usr/lib/x86_64-linux-gnu/libgfrpc.so.0(+0xccce)[0x7f030cdf0cce]
        (--&gt;
/usr/lib/x86_64-linux-gnu/libgfrpc.so.0(rpc_clnt_connection_cleanup+0x95)[0x7f030cdf1c45]
        (--&gt;
        /usr/lib/x86_64-linux-gnu/libgfrpc.so.0(+0xe890)[0x7f030cdf2890]
        ))))) 0-GL1image-client-2: forced unwinding frame type(GlusterFS
        4.x v1) op(FXATTROP(34)) called at 2019-12-05 22:00:19.736456
        (xid=0x825bffb)<br>
        [2019-12-05 22:00:43.790655] W [MSGID: 114031]
        [client-rpc-fops_v2.c:1614:client4_0_fxattrop_cbk]
        0-GL1image-client-2: remote operation failed<br>
        [2019-12-05 22:00:43.790686] E
        [rpc-clnt.c:346:saved_frames_unwind] (--&gt;
/usr/lib/x86_64-linux-gnu/libglusterfs.so.0(_gf_log_callingfn+0x139)[0x7f030d045f59]
        (--&gt;
        /usr/lib/x86_64-linux-gnu/libgfrpc.so.0(+0xcbb0)[0x7f030cdf0bb0]
        (--&gt;
        /usr/lib/x86_64-linux-gnu/libgfrpc.so.0(+0xccce)[0x7f030cdf0cce]
        (--&gt;
/usr/lib/x86_64-linux-gnu/libgfrpc.so.0(rpc_clnt_connection_cleanup+0x95)[0x7f030cdf1c45]
        (--&gt;
        /usr/lib/x86_64-linux-gnu/libgfrpc.so.0(+0xe890)[0x7f030cdf2890]
        ))))) 0-GL1image-client-1: forced unwinding frame type(GlusterFS
        4.x v1) op(FXATTROP(34)) called at 2019-12-05 22:00:19.736428
        (xid=0x89fee01)<br>
        [2019-12-05 22:00:43.790703] W [MSGID: 114031]
        [client-rpc-fops_v2.c:1614:client4_0_fxattrop_cbk]
        0-GL1image-client-1: remote operation failed<br>
        [2019-12-05 22:00:43.790774] E [MSGID: 114031]
        [client-rpc-fops_v2.c:1393:client4_0_finodelk_cbk]
        0-GL1image-client-1: remote operation failed [Transport endpoint
        is not connected]<br>
        [2019-12-05 22:00:43.790777] E
        [rpc-clnt.c:346:saved_frames_unwind] (--&gt;
/usr/lib/x86_64-linux-gnu/libglusterfs.so.0(_gf_log_callingfn+0x139)[0x7f030d045f59]
        (--&gt;
        /usr/lib/x86_64-linux-gnu/libgfrpc.so.0(+0xcbb0)[0x7f030cdf0bb0]
        (--&gt;
        /usr/lib/x86_64-linux-gnu/libgfrpc.so.0(+0xccce)[0x7f030cdf0cce]
        (--&gt;
/usr/lib/x86_64-linux-gnu/libgfrpc.so.0(rpc_clnt_connection_cleanup+0x95)[0x7f030cdf1c45]
        (--&gt;
        /usr/lib/x86_64-linux-gnu/libgfrpc.so.0(+0xe890)[0x7f030cdf2890]
        ))))) 0-GL1image-client-2: forced unwinding frame type(GlusterFS
        4.x v1) op(FXATTROP(34)) called at 2019-12-05 22:00:19.736542
        (xid=0x825bffc)<br>
        [2019-12-05 22:00:43.790794] W [MSGID: 114029]
        [client-rpc-fops_v2.c:4873:client4_0_finodelk]
        0-GL1image-client-1: failed to send the fop<br>
        [2019-12-05 22:00:43.790806] W [MSGID: 114031]
        [client-rpc-fops_v2.c:1614:client4_0_fxattrop_cbk]
        0-GL1image-client-2: remote operation failed<br>
        [2019-12-05 22:00:43.790825] E [MSGID: 114031]
        [client-rpc-fops_v2.c:1393:client4_0_finodelk_cbk]
        0-GL1image-client-2: remote operation failed [Transport endpoint
        is not connected]<br>
        [2019-12-05 22:00:43.790842] W [MSGID: 114029]
        [client-rpc-fops_v2.c:4873:client4_0_finodelk]
        0-GL1image-client-2: failed to send the fop<br>
        <br>
        the fop/transport not connected errors just repeat for another
        50 lines or so until I hit 22:00:46 seconds at which point the
        Volumes appear to be fine (though the VMs were still read-only
        until I rebooted.<br>
        <br>
        [2019-12-05 22:00:46.987242] W
        [fuse-bridge.c:2827:fuse_readv_cbk] 0-glusterfs-fuse: 91701328:
        READ =&gt; -1 gfid=d883b7c4-97f5-4f12-9373-7987cfc7dee4
        fd=0x7f02f005b708 (Transport endpoint is not connected)<br>
        [2019-12-05 22:00:47.029947] W
        [fuse-bridge.c:2827:fuse_readv_cbk] 0-glusterfs-fuse: 91701329:
        READ =&gt; -1 gfid=d883b7c4-97f5-4f12-9373-7987cfc7dee4
        fd=0x7f02f005b708 (Transport endpoint is not connected)<br>
        [2019-12-05 22:00:49.901075] W
        [fuse-bridge.c:2827:fuse_readv_cbk] 0-glusterfs-fuse: 91701330:
        READ =&gt; -1 gfid=c342dba6-a2a2-49a8-be3f-cd320e90c956
        fd=0x7f02f002bee8 (Transport endpoint is not connected)<br>
        [2019-12-05 22:00:49.923525] W
        [fuse-bridge.c:2827:fuse_readv_cbk] 0-glusterfs-fuse: 91701331:
        READ =&gt; -1 gfid=c342dba6-a2a2-49a8-be3f-cd320e90c956
        fd=0x7f02f002bee8 (Transport endpoint is not connected)<br>
        [2019-12-05 22:00:49.970219] W
        [fuse-bridge.c:2827:fuse_readv_cbk] 0-glusterfs-fuse: 91701332:
        READ =&gt; -1 gfid=fcec6b7a-ad23-4449-aa09-107e113877a1
        fd=0x7f02f008dd58 (Transport endpoint is not connected)<br>
        [2019-12-05 22:00:50.023932] W
        [fuse-bridge.c:2827:fuse_readv_cbk] 0-glusterfs-fuse: 91701333:
        READ =&gt; -1 gfid=fcec6b7a-ad23-4449-aa09-107e113877a1
        fd=0x7f02f008dd58 (Transport endpoint is not connected)<br>
        [2019-12-05 22:00:54.807833] I
        [rpc-clnt.c:2028:rpc_clnt_reconfig] 0-GL1image-client-2:
        changing port to 49153 (from 0)<br>
        [2019-12-05 22:00:54.808043] I
        [rpc-clnt.c:2028:rpc_clnt_reconfig] 0-GL1image-client-1:
        changing port to 49152 (from 0)<br>
        [2019-12-05 22:00:46.115076] E [MSGID: 133014]
        [shard.c:1799:shard_common_stat_cbk] 0-GL1image-shard: stat
        failed: 7a5959d6-75fc-411d-8831-57a744776ed3 [Transport endpoint
        is not connected]<br>
        [2019-12-05 22:00:54.820394] I [MSGID: 114046]
        [client-handshake.c:1106:client_setvolume_cbk]
        0-GL1image-client-1: Connected to GL1image-client-1, attached to
        remote volume '/GLUSTER/GL1image'.<br>
        [2019-12-05 22:00:54.820447] I [MSGID: 114042]
        [client-handshake.c:930:client_post_handshake]
        0-GL1image-client-1: 10 fds open - Delaying child_up until they
        are re-opened<br>
        [2019-12-05 22:00:54.820549] I [MSGID: 114046]
        [client-handshake.c:1106:client_setvolume_cbk]
        0-GL1image-client-2: Connected to GL1image-client-2, attached to
        remote volume '/GLUSTER/GL1image'.<br>
        [2019-12-05 22:00:54.820568] I [MSGID: 114042]
        [client-handshake.c:930:client_post_handshake]
        0-GL1image-client-2: 10 fds open - Delaying child_up until they
        are re-opened<br>
        [2019-12-05 22:00:54.821381] I [MSGID: 114041]
        [client-handshake.c:318:client_child_up_reopen_done]
        0-GL1image-client-1: last fd open'd/lock-self-heal'd - notifying
        CHILD-UP<br>
        [2019-12-05 22:00:54.821406] I [MSGID: 108002]
        [afr-common.c:5602:afr_notify] 0-GL1image-replicate-0:
        Client-quorum is met<br>
        [2019-12-05 22:00:54.821446] I [MSGID: 114041]
        [client-handshake.c:318:client_child_up_reopen_done]
        0-GL1image-client-2: last fd open'd/lock-self-heal'd - notifying
        CHILD-UP<br>
        <br>
        What is odd is that the gluster logs on the #3 and #1 show
        absolutely ZERO gluster errors around that time nor do I show
        any Network/teamd errors on any of the  3 nodes (including the
        problem node #2)<br>
        <br>
        I've checked dmesg/syslog and every other log file on the box.<br>
        <br>
        According to a staff member, we had this same kvm host have the
        same problem about 3 weeks ago, it was written up as a fluke
        possible due to excess disk I/O, since we have been using
        gluster for years and rarely have seen issues, especially with
        very basic gluster usage.<br>
        <br>
        In this case those VMs weren't overly busy and now we have a
        repeat problem.<br>
        <br>
        So I am wondering where else I can look to diagnose the problem
        or should I abandon the hardware/setup?<br>
        <br>
        I assume its a networking issue and not on gluster, but I am
        confused why gluster nodes #1 and #3 didn't complain about not
        seeing #2? If the networking did drop out should they have
        noticed?<br>
        <br>
        There also doesn't appear to be any visible hard disk issues
        (smartd is running)<br>
        <br>
        Side Note: I have reset the tcp-timeout back to 42 seconds and
        will look at upgrading to 6.6. I also see that the ARB and the
        unaffected Gluster node were running Gluster 6.4 (I don't know
        why #2 is on 6.5 but I am checking on that as well, we turn off
        auto-upgrade)<br>
        <br>
        Maybe the mismatched versions are the culprit?<br>
        <br>
        Also, we have a large of these replica 2+1 gluster setups
        running gluster version from 5.x up and none of the others have
        had this issue<br>
        <br>
        Any advise would be appreciated.<br>
        <br>
        Sincerely,<br>
        <br>
        Wk<br>
        <br>
        <br>
        <br>
        <br>
        <br>
      </div>
      <br>
      <fieldset class="mimeAttachmentHeader"></fieldset>
      <pre class="moz-quote-pre" wrap="">________

Community Meeting Calendar:

APAC Schedule -
Every 2nd and 4th Tuesday at 11:30 AM IST
Bridge: <a class="moz-txt-link-freetext" href="https://bluejeans.com/441850968">https://bluejeans.com/441850968</a>

NA/EMEA Schedule -
Every 1st and 3rd Tuesday at 01:00 PM EDT
Bridge: <a class="moz-txt-link-freetext" href="https://bluejeans.com/441850968">https://bluejeans.com/441850968</a>

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="https://lists.gluster.org/mailman/listinfo/gluster-users">https://lists.gluster.org/mailman/listinfo/gluster-users</a>
</pre>
    </blockquote>
  </body>
</html>