Specifications

4
WWW.EXTREMENETWORKS.COM
MLXe Interface Modules
The MLXe provides a wide range of leading-edge
Ethernet modules for 100 GbE, 40 GbE, 10 GbE, 1 GbE
that enable organizations to use a single platform for many
applications. For more information about these modules,
see the tables below.
Versions Supported
Interface Module -DM -M -X -X2
24×1 GbE RJ45
24×1 GbE Fiber
4×10 GbE
8×10 GbE
24×10 GbE
20×10 GbE
20×1 GbE (upgradable to 10 GbE)
4×10 GbE IPsec
4×40 GbE
2×100 GbE CFP2
Interface Module Version Supported Route Table Scalability
-DM
Up to 256,000 IPv4 routes
Up to 64,000 IPv6 routes
-M
Up to 512,000 IPv4 routes
Up to 256,000 IPv6 routes
-X
Up to 1 million IPv4 routes
Up to 240,000 IPv6 routes
-X2
Up to 2.4 million IPv4 routes
Up to 1.8 million IPv6 routes
*Denotes full-slot module, all other interface modules are half-slot.
RIP/RIPng, OSPF/OSPFv3, IS-IS/IS-IS for IPv6, and BGP-4/BGP-MP
for IPv6
Secure Multi-VRF routing for supporting virtual routing applications
over non-MPLS backbones
VRRP and VRRP-E
Connecting IPv6 islands over IPv4 MPLS using IPv6 Provider Edge
(6PE) routers
6VPE enabling IPv6 multitenancy to the edge of the cloud
127-bit IPv6 interface addresses
Industry-leading scalability up to:*
10 million BGP routes
2.4 million IPv4 routes in hardware (FIB)
1.8 million IPv6 routes in hardware (FIB)
153,600 multicast routes
2,000 BGP peers per system
2,000 BGP/MPLS VPNs and up to 1 million VPN routes
48,000 VLLs per system
16,000 VPLS instances and up to 1 million VPLS MAC addresses and
64,000 RSVP-TE LSPs
4,094 VLANs and up to 2 million MAC addresses
Large-scale Equal Cost Multi-Path (ECMP); up to 32 paths for
unicast and multicast
Software-Defined Networking (SDN):
OpenFlow 1.3: QoS (for metering and enqueue), Group Table
(select and fast failover), QinQ (TAG type autorecognition), Active-
Standby Controller, IPv6, Transport Layer Security (TLS)
1.2 (controller interface)
OpenFlow in hybrid port mode with support for sFlow, IP, and
MPLS/VPLS (uplinks) with protected VLAN for additional flexibility
Up to 128,000 flows supported
True 12-tuple (Layer 2 and Layer 3) matching for a diverse set
of applications
Industry-leading performance for MPLS services, providing several
service choices:
IP over MPLS, IPv6 over MPLS (6PE), IPv6 Layer 3 VPNs (6VPE),
MPLS over GRE, Virtual Leased Line (VLL), Virtual Private LAN
Service (VPLS), BGP/MPLS VPN, Multi-VRF, routing over VPLS,
Max VPLS LSP Load Balance Scale for LER, RSVP TE Link Metric
for CSPF Computation, RSVP Auto-Bandwidth with Absolute
Threshold
Comprehensive MPLS signaling and path calculation algorithms for both
trac engineered and non-trac-engineered applications:
LDP, OSPF-TE , IS-IS-TE , RSVP-TE , CSPF, LDP over RSVP,
Point-to-Multipoint (P2MP) RSVP-TE LSP
MPLS FRR (detour, bypass) and hot standby paths for
trac protection
Superior high-availability design:
Redundant management modules
Redundant switch fabrics
Redundant power supplies and cooling system
Hitless Layer 2/3 failover with stateful OSPF IS-IS and BGP
Graceful Restart
Hitless (in-service) software upgrades with Graceful Restart
Advanced QoS:
Hierarchical Quality of Service (H-QoS): Supports up to four
levels of hierarchy— port, logical port (optional), customer
(optional), and service
Inbound and outbound two-rate threecolor trac policers
with accounting
Eight distinct priority levels
WRED support for congestion management and precedence
dropping (tunable via configuration)
Support for hybrid queue servicing disciplines: Mixed, Strict Priority,
and Weighted Fair Queuing
Route Table Scalability for MLXe
Interface Modules
Key Features
4-, 8-, 16-, and 32-slot systems for maximum deployment versatility
Up to 9.5 billion packets per second routing performance with non-blocking
15.36 Tbps data capacity
Ideal for demanding, high-density environments with up to:
64 100 GbE ports per system
128 40 GbE ports per system
768 10 GbE ports per system
1,536 1 GbE ports per system
Comprehensive, wire-speed, dual-stack IPv4/ IPv6 routing support based on
the Multi-Service IronWare OS:
High-performance, robust routing using Forwarding Information
Base (FIB) programming in hardware