<html>
  <head>
    <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
  </head>
  <body text="#000000" bgcolor="#FFFFFF">
    <div class="moz-cite-prefix">On 2018-05-01 01:42, Olivier Benghozi
      wrote:<br>
    </div>
    <blockquote type="cite"
      cite="mid:8E133F67-1C45-47C0-B892-099A31670154@wifirst.fr">
      <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
      Hi Roman,
      <div class=""><br class="">
      </div>
      <div class="">Brandon Leeberg in this ML also recently posted
        about the same issue (with the same prefix by the way),
        running SEOS-12.1.1.9 and 12.1.1.12p13. Nothing seems bad with
        this route.<br class="">
        <div><br class="">
        </div>
        <div>In fact, I found back a pcap capture (from december 2017)
          of a BGP session from one of my Juniper MX gears toward a
          BGP/Netflow collector, where I can see this route.</div>
        <div>And I can see after all that there's a difference between
          your version and what was transmitted by this MX</div>
        <div>For the AGGREGATOR attribute, the "partial" bit is at 0 in
          my capture (meaning tat the attribute is "complete", that is
          everything is OK), whereas in your case it is set at 1 (so the
          attribute begins with c0 instead of e0).</div>
        <div><br class="">
        </div>
        <div>In Brandon's case the "partial" bit was also at 1.</div>
        <div>So I suppose that this is what the SE code doesn't like.</div>
        <div><br class="">
        </div>
        <div>There's no serious reason for this flag to be set to 1 for
          this prefix (or it means that a BGP router transmitted this
          announcement without understanding what AGGREGATOR attribute
          was, which is ridiculous). That's probably a problem on the
          originator's side.</div>
        <div>But there's no reason for SEOS to consider this attribute
          as bad (and no reason to close the session since RFC7606, but
          SEOS is now a dead end).</div>
      </div>
    </blockquote>
    <br>
    This is also what comes from our analyse.<br>
    It looks like an implementation bug.<br>
    <br>
    The biggest issue that SEOS is in "end of maintenance state".
    However, they still release some newer patches (12.1.1.12p14)..<br>
    <br>
    I have sent this info to a person who worked as routing chief in
    Ericsson some time ago, maybe he has some actual contact... but no
    response for now...<br>
    <br>
    We are also facing lot of EPPA3 crashes and have no idea what to
    turn off...<br>
    <br>
    Regards,<br>
    Marcin<br>
    <br>
    <br>
    <br>
    <br>
    <blockquote type="cite"
      cite="mid:8E133F67-1C45-47C0-B892-099A31670154@wifirst.fr">
      <div class="">
        <div><br class="">
        </div>
        <div>However I guess that on Brandon case, the sessions was
          staying alive (juste error messages in the logs)...</div>
        <div>
          <div><br class="">
          </div>
          <div>Seems like a bug to me, I guess that only an Ericsson TAC
            engineer could help fix this SEOS BGP piece of code.</div>
          <div class=""><br class="">
          </div>
          <div class=""><br class="">
          </div>
          <div class="">Olivier</div>
          <div class=""><br class="">
          </div>
          <blockquote type="cite" class="">
            <div class="">On 30 apr. 2018 at 23:54, Соловьёв Роман
              Анатольевич <<a href="mailto:romanse@serdi.ru" class=""
                moz-do-not-send="true">romanse@serdi.ru</a>> wrote :</div>
            <br class="Apple-interchange-newline">
            <div class=""><br class="">
              <blockquote class="">
                <div dir="ltr" class="">
                  <div class="">
                    <div class="">
                      <div class="">Hi. Some issue is detected with SeOS
                        version SEOS-12.1.1.12p13-Release<br class="">
                      </div>
                      <div class="">The issue is about BGP protocol
                        handling. <br class="">
                      </div>
                      <div class="">The problem is, that SeOS close a
                        BGP session on receiving mailformed UPDATE
                        message from a peer. The peer is Juniper. <br
                          class="">
                      </div>
                      <div class=""><br class="">
                        On the peer side:<br class="">
                        <p class=""><font class="" size="4"><span
                              style="font-family:Calibri,sans-serif;color:rgb(31,73,125);"
                              class="" lang="EN-US">bgp_read_v4_message:11175:
                              NOTIFICATION received from 5.143.236.222
                              (External AS 48711): code 3 (Update
                              Message Error) subcode 4 (attribute flags
                              error), Data:  e0 <span class=""><span
                                  class="">07 08 00 03 02</span></span>
                              Apr 30 09:52:06 2018</span></font></p>
                      </div>
                      <div class=""><br class="">
                        <b class="">On SeOS side:</b><br class="">
                        <br class="">
                        bgp neighbor 5.143.236.221<br class="">
                        BGP neighbor: 5.143.236.221, remote AS: 12389,
                        external link<br class="">
                          Version: 4, router identifier: 178.34.128.3<br
                          class="">
                          State: Idle for 00:00:25<br class="">
                          Last read 00:00:25, last send 00:00:25<br
                          class="">
                          Hold time: configured 180, negotiated 0<br
                          class="">
                          Keepalive time: configured 30, negotiated 0<br
                          class="">
                          Local restart timer 120 sec, stale route
                        retain timer 180 sec<br class="">
                          Received restart timer 0 sec, flag 0x0<br
                          class="">
                          Number of hops external BGP neighbor may be
                        away: 1<br class="">
                          Minimum time between advertisement runs: 30
                        secs<br class="">
                          Source (local) IP address: 0.0.0.0<br class="">
                          Received messages: 0 (0 bytes), notifications:
                        0, in queue: 0<br class="">
                          Sent messages: 0 (0 bytes), notifications:
                        289, out queue: 0<br class="">
                          Last active open: 06:10:23, reason: Have not
                        registered with RIB<br class="">
                          Reset count: 289, last reset time: 00:00:25,
                        reset reason: N<b class="">otification sent
                          (update: attribute flags error)</b><br
                          class="">
                        <br class="">
                        show bgp neighbor 5.143.236.221 malform update <br
                          class="">
                        Apr 30 10:42:23 Malformed UPDATE msg (nbr
                        5.143.236.221, context 0x<span class=""><span
                            class="">40080002</span></span>, 80 bytes,
                        repeated 1512 times, reason: Invalid msg) - <br
                          class="">
                        ffff ffff ffff ffff ffff ffff ffff ffff <span
                          class=""><span class="">0050 0200 0000 3540
                            0101 0040 020</span></span>e <span class=""><span
                            class="">0203 0000 3065 0000</span></span>
                        0c<span class=""><span class="">97 0003 02</span></span>ed
                        <span class=""><span class="">4003 0405 8</span></span>fec
                        dd40 0600 e<span class=""><span class="">007
                            0800 0302</span></span> ed5b dc3f 01c0 <span
                          class=""><span class="">0808 3065 0006 3065
                            0007 185</span></span>b dc3f<br class="">
                      </div>
                      <br class="">
                    </div>
                    Lets parse this data.<br class="">
                    ffff ffff ffff ffff ffff ffff ffff ffff - the init
                    marker<br class="">
                  </div>
                  0050 - totak message length - 80 bytes<br class="">
                  <div class=""><br class="">
                    <b class="">02</b> -  UPDATE <br class="">
                    <div class=""><b class="">0000</b> Length of
                      Withdrawn Routes <br class="">
                      <b class="">0035</b> Total size of attributes (<b
                        class="">53 bytes</b>)<br class="">
                    </div>
                    <div class=""><br class="">
                    </div>
                    <div class="">Attributes:<br class="">
                    </div>
                    <div class=""><b class=""><span class=""><span
                            class="">40 01 01 00</span></span></b><br
                        class="">
                      ORIGIN (IGP) <br class="">
                    </div>
                    <div class=""><br class="">
                    </div>
                    <div class=""><b class="">40 02 0e <span class=""><span
                            class="">02 03 0000 3065 0000 0</span></span>c<span
                          class=""><span class="">97 0003 02</span></span>ed</b><br
                        class="">
                    </div>
                    <div class="">40-flags<br class="">
                    </div>
                    <div class="">02 -  AS_PATH<br class="">
                    </div>
                    <div class="">0e - length  - 14 <b class="">bytes<br
                          class="">
                      </b></div>
                    <div class="">02 - segment type AS_SEQUENCE <br
                        class="">
                      03  - 3 AS length<br class="">
                      <span class=""><span class="">0000 3065 0000 0</span></span>c<span
                        class=""><span class="">97 0003 02</span></span>ed
                      -  ASN itself (12389,3223,197357)<br class="">
                    </div>
                    <div class=""><br class="">
                      <b class=""><span class=""><span class="">40 03 04
                            05 8</span></span>f ec dd<br class="">
                      </b>NEXT_HOP<b class=""> </b>5.143.236.221<b
                        class=""><br class="">
                      </b></div>
                    <div class=""><b class=""><br class="">
                      </b></div>
                    <div class=""><b class="">40 06 00 <br class="">
                      </b>an empty ATOMIC_AGGREGATE attribute<br
                        class="">
                    </div>
                    <div class=""><br class="">
                      <b class="">e<span class=""><span class="">0 07 08
                            0003 02</span></span>ed   5b dc 3f 01 </b><br
                        class="">
                      AGGREGATOR AS 197357 IP 93.220.63.1<br class="">
                    </div>
                    <div class=""><br class="">
                    </div>
                    <div class=""><b class="">c<span class=""><span
                            class="">0 08 08 3065 0006 3065 0007</span></span>
                        <br class="">
                      </b></div>
                    <div class="">COMMUNITY 12389:6 12389:7<b class=""><br
                          class="">
                      </b></div>
                    <div class=""><br class="">
                      <b class="">18 5b dc 3f <br class="">
                      </b></div>
                    Prefixes<b class="">  </b><a
                      href="http://91.220.63.0/24" target="_blank"
                      class="" moz-do-not-send="true">91.220.63.0/24</a><b
                      class=""><br class="">
                    </b><br class="">
                  </div>
                  <div class="">According the notification message  SeOS
                    threats the AGGREGATOR attribute flags as mailfomed:<br
                      class="">
                    <b class="">e<span class=""><span class="">0 07 08
                          0003 02</span></span>ed   5b dc 3f 01 </b></div>
                  <div class="">I don't see anything wrong with it. <br
                      class="">
                    IMHO the AGGRETATOR attribute is composed with all
                    RFC requirements<br class="">
                  </div>
                  <div class=""><br class="">
                    Can somebody explain me such unexpected behavior?<br
                      class="">
                  </div>
                  <div class=""><br class="">
                  </div>
                </div>
              </blockquote>
            </div>
          </blockquote>
        </div>
        <br class="">
      </div>
      <br>
      <fieldset class="mimeAttachmentHeader"></fieldset>
      <br>
      <pre wrap="">_______________________________________________
redback-nsp mailing list
<a class="moz-txt-link-abbreviated" href="mailto:redback-nsp@puck.nether.net">redback-nsp@puck.nether.net</a>
<a class="moz-txt-link-freetext" href="https://puck.nether.net/mailman/listinfo/redback-nsp">https://puck.nether.net/mailman/listinfo/redback-nsp</a>
</pre>
    </blockquote>
    <p><br>
    </p>
    <div class="moz-signature">-- <br>
      <p style="font-family: Helvetica, Arial, sans-serif; font-size:
        12px; line-height: 14px; color: #999999;"><span id="name-input"
          class="txt" style="font-weight: bold;">Marcin Kuczera</span> <span
          id="title-sep">/</span> <span id="title-input" class="txt"
          style="color: #999;">Wiceprezes Zarządu / CTO<span><br>
            <span id="mobile-input" class="txt" style="color: #999;">+48
              32 440 80 71<span id="email-sep" class="txt">/</span> <a
                id="email-input" class="link email" style="color:
                #51930b;" href="mailto:marcin.kuczera@leon.pl">marcin.kuczera@leon.pl</a></span></span></span></p>
      <p style="font-family: Helvetica, Arial, sans-serif; font-size:
        12px; line-height: 14px;"><span id="company-input" class="txt"
          style="font-weight: bold; color: #999;">Leon Sp. z o.o.</span>
        <span id="office-sep" class="txt" style="color: #999;"> </span>
        <span id="address-sep"><br>
        </span> <span id="address-input" class="txt" style="color:
          #999;">ul. Kilińskiego 33d, 44-200 Rybnik </span><br>
        <a id="website-input" class="link" style="color: #51930b;"
          href="http://www.leon.pl/">http://www.leon.pl/</a></p>
      <p id="disclaimer-input" class="txt" style="font-family:
        Helvetica, Arial, sans-serif; color: #999999; font-size: 12px;
        line-height: 14px;">INTERNET | TELEWIZJA | TELEFON</p>
      <p style="font-family: Helvetica, Arial, sans-serif; font-size:
        12px; line-height: 14px;"><span id="office-sep" class="txt"
          style="color: #999;">KRS 0000223101 Sąd Rejonowy w Gliwicach<br>
          <span id="mobile-input" class="txt" style="color: #999;"></span><span
            id="mobile-input" class="txt" style="color: #999;"> Kapitał
            zakładowy 576.700 zł<br>
            <span id="mobile-input" class="txt" style="color: #999;"></span><span
              id="mobile-input" class="txt" style="color: #999;"> NIP:
              6332068698</span></span></span></p>
    </div>
  </body>
</html>