<?xml version="1.0" encoding="UTF-8"?>
<rss xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:taxo="http://purl.org/rss/1.0/modules/taxonomy/" version="2.0">
  <channel>
    <title>topic Re: ISIS SPBM MTU calculation in ExtremeSwitching (VSP/Fabric Engine)</title>
    <link>https://community.extremenetworks.com/t5/extremeswitching-vsp-fabric/isis-spbm-mtu-calculation/m-p/74151#M1042</link>
    <description>&lt;P&gt;Hi Fijs&lt;/P&gt;&lt;P&gt;I would say the remaining 50 bytes are for Fabric Extend (VXLAN) encapsualtion. If you than encaspsulate this into IPsec, there will be another 150 Bytes added, so you will reach a max. MTU of 1671 Bytes.&lt;/P&gt;&lt;P&gt;Regards&lt;/P&gt;&lt;P&gt;Thomas&lt;/P&gt;</description>
    <pubDate>Tue, 18 Aug 2020 19:18:16 GMT</pubDate>
    <dc:creator>Thomas_Gfeller</dc:creator>
    <dc:date>2020-08-18T19:18:16Z</dc:date>
    <item>
      <title>ISIS SPBM MTU calculation</title>
      <link>https://community.extremenetworks.com/t5/extremeswitching-vsp-fabric/isis-spbm-mtu-calculation/m-p/74150#M1041</link>
      <description>&lt;P&gt;Hi,&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;The docs / guides specify that in order to form an ISIS adjacency, an MTU of at least 1594 bytes is required.&lt;/P&gt;&lt;P&gt;Why 1594? What is missing in the below calculation?&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;&lt;STRONG&gt;Original Ethernet Frame:&lt;/STRONG&gt;&lt;BR /&gt;Destination MAC address = 6 bytes&lt;/P&gt;&lt;P&gt;Source MAC address = 6 bytes&lt;/P&gt;&lt;P&gt;802.1Q header = 4 bytes (0x8100 + 802.1p field (3 bits) + CFI (1 bit) + VLAN-id (12 bits)&lt;/P&gt;&lt;P&gt;Ether Type (or length) = 2 bytes (e.g. 0x0800 for IP, 0x0806 for ARP)&lt;/P&gt;&lt;P&gt;default maximum payload of Ethernet Frame = 1500 bytes&lt;/P&gt;&lt;P&gt;Cyclic Redundancy Check (CRC) = 4 bytes&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;This Ethernet frame of 1522 bytes is then encapsulated with 22 extra bytes:&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;&lt;STRONG&gt;802.1ah Encapsulation:&lt;/STRONG&gt;&lt;BR /&gt;Backbone Destination MAC address = 6 bytes&lt;/P&gt;&lt;P&gt;Backbone Source MAC address = 6 bytes&lt;/P&gt;&lt;P&gt;Backbone Ether Type = 2 bytes (0x88A8)&lt;/P&gt;&lt;P&gt;Backbone VLAN-id = 2 bytes&lt;/P&gt;&lt;P&gt;Service Type = 2 bytes (0x88E7)&lt;/P&gt;&lt;P&gt;Service Flags = 1 byte&lt;/P&gt;&lt;P&gt;Service Identifier = 3 bytes&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;This gives a frame size of &lt;STRONG&gt;1544 &lt;/STRONG&gt;bytes.&lt;/P&gt;</description>
      <pubDate>Tue, 18 Aug 2020 18:37:59 GMT</pubDate>
      <guid>https://community.extremenetworks.com/t5/extremeswitching-vsp-fabric/isis-spbm-mtu-calculation/m-p/74150#M1041</guid>
      <dc:creator>Fijs</dc:creator>
      <dc:date>2020-08-18T18:37:59Z</dc:date>
    </item>
    <item>
      <title>Re: ISIS SPBM MTU calculation</title>
      <link>https://community.extremenetworks.com/t5/extremeswitching-vsp-fabric/isis-spbm-mtu-calculation/m-p/74151#M1042</link>
      <description>&lt;P&gt;Hi Fijs&lt;/P&gt;&lt;P&gt;I would say the remaining 50 bytes are for Fabric Extend (VXLAN) encapsualtion. If you than encaspsulate this into IPsec, there will be another 150 Bytes added, so you will reach a max. MTU of 1671 Bytes.&lt;/P&gt;&lt;P&gt;Regards&lt;/P&gt;&lt;P&gt;Thomas&lt;/P&gt;</description>
      <pubDate>Tue, 18 Aug 2020 19:18:16 GMT</pubDate>
      <guid>https://community.extremenetworks.com/t5/extremeswitching-vsp-fabric/isis-spbm-mtu-calculation/m-p/74151#M1042</guid>
      <dc:creator>Thomas_Gfeller</dc:creator>
      <dc:date>2020-08-18T19:18:16Z</dc:date>
    </item>
    <item>
      <title>Re: ISIS SPBM MTU calculation</title>
      <link>https://community.extremenetworks.com/t5/extremeswitching-vsp-fabric/isis-spbm-mtu-calculation/m-p/74152#M1043</link>
      <description>&lt;P&gt;Thanks Thomas, this makes sense!&lt;/P&gt;</description>
      <pubDate>Tue, 18 Aug 2020 22:32:46 GMT</pubDate>
      <guid>https://community.extremenetworks.com/t5/extremeswitching-vsp-fabric/isis-spbm-mtu-calculation/m-p/74152#M1043</guid>
      <dc:creator>Fijs</dc:creator>
      <dc:date>2020-08-18T22:32:46Z</dc:date>
    </item>
    <item>
      <title>Re: ISIS SPBM MTU calculation</title>
      <link>https://community.extremenetworks.com/t5/extremeswitching-vsp-fabric/isis-spbm-mtu-calculation/m-p/74153#M1044</link>
      <description>&lt;P&gt;These are my notes..&lt;/P&gt;&lt;P&gt;SPB MTUs&lt;BR /&gt;========&lt;BR /&gt;MAC-in-MAC adds 22 bytes to packets; so Ethernet maximum frame size can grow worst case from 1522 (assuming worst case q-tagged packet) to 1544&lt;BR /&gt;&amp;nbsp;&amp;nbsp; &amp;nbsp;64-&amp;gt;86&amp;nbsp;&amp;nbsp; &amp;nbsp;to&amp;nbsp;&amp;nbsp; &amp;nbsp;1522-&amp;gt;1544&lt;/P&gt;&lt;P&gt;Fabric Extend VXLAN encap adds 14 (18 if WAN interface q-tagged) bytes for MAC&amp;amp;etype + 20 bytes (IP header) + 8 bytes (UDP header) + 8 bytes (VXLAN header) = + 50 (54 if WAN interface q-tagged) bytes&lt;BR /&gt;&amp;nbsp;&amp;nbsp; &amp;nbsp;86-&amp;gt;136&amp;nbsp;&amp;nbsp; &amp;nbsp;to&amp;nbsp;&amp;nbsp; &amp;nbsp;1544-&amp;gt;1594&amp;nbsp;&amp;nbsp; &amp;nbsp;on untagged connection to WAN&lt;BR /&gt;&amp;nbsp;&amp;nbsp; &amp;nbsp;86-&amp;gt;140&amp;nbsp;&amp;nbsp; &amp;nbsp;to&amp;nbsp;&amp;nbsp; &amp;nbsp;1544-&amp;gt;1598&amp;nbsp;&amp;nbsp; &amp;nbsp;on q-tagged connection to WAN&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;</description>
      <pubDate>Thu, 20 Aug 2020 19:31:13 GMT</pubDate>
      <guid>https://community.extremenetworks.com/t5/extremeswitching-vsp-fabric/isis-spbm-mtu-calculation/m-p/74153#M1044</guid>
      <dc:creator>Ludovico_Steven</dc:creator>
      <dc:date>2020-08-20T19:31:13Z</dc:date>
    </item>
    <item>
      <title>Re: ISIS SPBM MTU calculation</title>
      <link>https://community.extremenetworks.com/t5/extremeswitching-vsp-fabric/isis-spbm-mtu-calculation/m-p/74154#M1045</link>
      <description>&lt;P&gt;Fijs,&lt;/P&gt;&lt;P&gt;It is depending of what/where do you tag/untag the packets.&lt;/P&gt;&lt;P&gt;Here the output of an excel when I was on a project to validate the MTU’s and throughput of the Fabric:&lt;/P&gt;&lt;FIGURE&gt;&lt;span class="lia-inline-image-display-wrapper" image-alt="6b6e1b6cf2f247d3bdc9fdd32ae9e9c1_d26facaf-2d3b-4f97-8ca2-e4b2e615d856.png"&gt;&lt;img src="https://community.extremenetworks.com/t5/image/serverpage/image-id/5714i81A591D9CAF34411/image-size/large?v=v2&amp;amp;px=999" role="button" title="6b6e1b6cf2f247d3bdc9fdd32ae9e9c1_d26facaf-2d3b-4f97-8ca2-e4b2e615d856.png" alt="6b6e1b6cf2f247d3bdc9fdd32ae9e9c1_d26facaf-2d3b-4f97-8ca2-e4b2e615d856.png" /&gt;&lt;/span&gt;&lt;/FIGURE&gt;&lt;FIGURE&gt;&lt;span class="lia-inline-image-display-wrapper" image-alt="6b6e1b6cf2f247d3bdc9fdd32ae9e9c1_3f2b2641-a3f3-47b6-9bd6-408a08984116.png"&gt;&lt;img src="https://community.extremenetworks.com/t5/image/serverpage/image-id/3976i0E2D3CA9F422324B/image-size/large?v=v2&amp;amp;px=999" role="button" title="6b6e1b6cf2f247d3bdc9fdd32ae9e9c1_3f2b2641-a3f3-47b6-9bd6-408a08984116.png" alt="6b6e1b6cf2f247d3bdc9fdd32ae9e9c1_3f2b2641-a3f3-47b6-9bd6-408a08984116.png" /&gt;&lt;/span&gt;&lt;/FIGURE&gt;&lt;P&gt;Mig&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;</description>
      <pubDate>Thu, 27 Aug 2020 04:23:07 GMT</pubDate>
      <guid>https://community.extremenetworks.com/t5/extremeswitching-vsp-fabric/isis-spbm-mtu-calculation/m-p/74154#M1045</guid>
      <dc:creator>Miguel-Angel_RO</dc:creator>
      <dc:date>2020-08-27T04:23:07Z</dc:date>
    </item>
    <item>
      <title>Re: ISIS SPBM MTU calculation</title>
      <link>https://community.extremenetworks.com/t5/extremeswitching-vsp-fabric/isis-spbm-mtu-calculation/m-p/121326#M3176</link>
      <description>&lt;P&gt;Hi,&lt;/P&gt;&lt;P&gt;The 1594-byte requirement comes from the IS-IS SPBM control packets, not just the Ethernet frame you calculated. Your math only covers the outer Ethernet + 802.1ah headers around a normal 1500-byte payload, which indeed lands around 1544 bytes. However, IS-IS uses CLNS (Connectionless Network Service) packets carried directly over Layer-2, and those packets include additional headers like the LLC header (802.2) and the IS-IS PDU fields themselves. When those protocol headers are added, the effective frame size needed on the wire increases beyond the simple Ethernet encapsulation you listed. Get it here:&amp;nbsp;&lt;A href="https://calculateurheure.fr/" target="_blank"&gt;https://calculateurheure.fr/&lt;/A&gt;&lt;/P&gt;&lt;P&gt;In SPBM deployments vendors typically recommend 1594 bytes as a safe MTU so IS-IS LSPs and other control PDUs can be exchanged without fragmentation. The extra ~50 bytes account for the LLC header, CLNS/IS-IS protocol overhead, and implementation safety margin used by many platforms. So your 1544-byte calculation is correct for the encapsulation portion, but it doesn’t include the IS-IS control protocol overhead, which is why the documented minimum MTU is higher.&lt;/P&gt;&lt;P&gt;Regards&lt;/P&gt;</description>
      <pubDate>Wed, 11 Mar 2026 10:28:39 GMT</pubDate>
      <guid>https://community.extremenetworks.com/t5/extremeswitching-vsp-fabric/isis-spbm-mtu-calculation/m-p/121326#M3176</guid>
      <dc:creator>jerrygen</dc:creator>
      <dc:date>2026-03-11T10:28:39Z</dc:date>
    </item>
  </channel>
</rss>

