Juniper SP Products Update Ivan Lysogor 4 th

MARKET & CHALLENGES € Margin pressure Long Lead Times Operational Complexity Successful business requires

FULL PORTFOLIO COVERAGE ACX PTX North. Star Controller MX MX ACX MOBILE ACX HOME

MX Product Update 4 Copyright © 2014 Juniper Networks, Inc.

MX Portfolio Overview 9600 Gbps One TRIO Architecture One UNIVERSAL EDGE 4800 Gbps 2860

MPC 6 E Overview MPC 6 E Description 480 (520) Gbps modular line card

Routing system upgrade scenario Highlights Driver package installed With Continuity Support May upgrade to

MX Data Center Gateway Internet DC Gateway EVPN • • 10 13. 2 R

EVPN Advantages Link Efficiency Convergence L 3 and L 2 All Active forwarding with

PTX Update 12 Copyright © 2014 Juniper Networks, Inc.

PTX Series Routers PTX 1000: Distributed Converged Supercore Power 2. 88 Tb Distributed Core

PTX Series Routers PTX Product Family of Routers: Technical Specifications PTX 1 K PTX

v. MX introduction 15 Copyright © 2014 Juniper Networks, Inc.

Physical vs. Virtual Each option has its own strength, and it is created with

VMX overview • Efficient separation of control and data-plane Data packets are switched within

v. MX Use Cases 18 Copyright © 2014 Juniper Networks, Inc.

General consideration of v. MX deployment • v. MX behaves the same way as

Agility example: Bring up a new service in a POP 1. Install a new

Proof of concept lab validation or SW certification • CAPEX or OPEX reduction for

v. CPE solution with v. MX as the SDN Gateway • v. MX as

Virtual Route Reflector Virtual RR on VMs On standard servers Virtual RR i. BG

Virtual BNG cluster in a data center v. MX as v. BNG cluster v.

v. MX Product Details 25 Copyright © 2014 Juniper Networks, Inc.

v. MX Components VCP (Virtualized Control Plane) VFP (Virtualized Forwarding Plane) • • •

VMX system architecture • Optimized data path from physical NIC to v. NIC via

Product Offering at FRS 28 Copyright © 2014 Juniper Networks, Inc.

FRS Product Offering • • • FRS at Q 2 2015 with JUNOS release

Reference server configuration CPU Intel Xeon 3. 1 GHz Cores Minimum 10 cores RAM

Performance 31 Test setup 8. 9 G 8. 9 G VMX 8. 9 G

Slides: 32

Download presentation

FULL PORTFOLIO COVERAGE ACX PTX North. Star Controller MX MX ACX MOBILE ACX HOME MOBILE MX MX MX HQ ACX MOBILE PTX MX Metro MX PTX Backbone PTX ACX HOME ACX PTX MOBILE MX HQ HQ ACX HOME Metro ACX AGGREGATION MX CORE NETWORK PTX EPC MX BRANCH MX PTX ACCESS PRE-AGGREGATION ACX PTX Metro ACX MX PTX Metro v. MX MOBILE SDN GW PTX MX BRANCH Contrail Controller CPE Metro ACX HOME Solution requirements • Solution cost - Port cost - Maintenance cost - Upgrade cost • Reliability • New service rollout speed 3 Copyright © 2014 Juniper Networks, Inc.

MX Portfolio Overview 9600 Gbps One TRIO Architecture One UNIVERSAL EDGE 4800 Gbps 2860 Gbps 1560 Gbps 520 Gbps 80 Gbps N x 10 Gbps VMX 5 MX 104 MX 240 * Current full duplex capacity is shown MX 480 MX 960 MX 2010 MX 2020 Copyright © 2014 Juniper Networks, Inc.

MPC 5 E 24 x 10 GE or 6 x 40 GE MPC 5 E Description 240 Gbps line card with flexible 10 GE/40 GE interface configuration options, increased scale and OTN support MIC 0/1 Interface Features • Interface combinations: • 24 x 10 GE (MIC 0 and MIC 1) • 6 x 40 GE (MIC 2 and MIC 3) • 12 x 10 GE and 3 x 40 GE (MIC 0 and MIC 3) • 12 x 10 GE and 3 x 40 GE (MIC 1 and MIC 2) • Port queues with optional 32 K queues upgrade license • 1 M queues option XM XL QSFP QSFP XQ MIC 2 XM MIC 3 Applications and Scale • Up to 10 M IP Routes (in hardware) • Full scale L 3 VPN and VPLS • Increased Inline IPFIX Scale 6 Copyright © 2014 Juniper Networks, Inc.

MPC 5 E 2 x 100 GE and 4 x 10 GE MPC 5 E Description 240 Gbps line card with providing 100 GE and 10 GE connectivity, increased scale and OTN support CFP 2 Interface Features CFP 2 • Interfaces: • 4 x 10 GE SFP+ • 2 x 100 GE CFP 2 • Port queues with optional 32 K queues upgrade license • 1 M queues option MIC 1 XM MIC 3 XL XQ MIC 0 XM MIC 2 Applications and Scale • Up to 10 M IP Routes (in hardware) • Full scale L 3 VPN and VPLS • Increased Inline IPFIX Scale 7 Copyright © 2014 Juniper Networks, Inc.

MPC 6 E Overview MPC 6 E Description 480 (520) Gbps modular line card for MX 2 K platform, increased scale and performance Interface Features • Interface cards supported: • 2 x 100 G CFP 2 w/ OTN (OTU 4) • 4 x 100 G CXP • 24 x 10 G SFP+ • 24 x 10 GSFP+ w/ OTN (OTU 2) • Port based queueing • Limited scale per-vlan queueing XM MIC 0 XL XM XM MIC 1 XL XM Applications and Scale • Up to 10 M IP Routes • Full scale L 3 VPN and VPLS • Increased Inline IPFIX Scale 8 Copyright © 2014 Juniper Networks, Inc.

Routing system upgrade scenario Highlights Driver package installed With Continuity Support May upgrade to new fabric or may use the same 8 x SFB 800 Gbps per slot New higher density MPCs added MPC 6 installed, 480 Gbps per slot 9 • Protects investments into hardware (SFBs) • Reduces software qualification efforts • No JUNOS upgrade required, driver installation is in service Future Router Happy network engineers working on something else (probably SDN-related) Copyright © 2014 Juniper Networks, Inc.

MX Data Center Gateway Internet DC Gateway EVPN • • 10 13. 2 R 1 First Implementation (MPLS encaps) 14. 1 R 2 VM Mobility Support 14. 1 R 4 Active / Active Support 14. 1 R 4 VXLAN encapsulation DC Fabric VXLAN • 14. 1 R 4 VMWare NSX Integration Copyright © 2014 Juniper Networks, Inc.

EVPN Advantages Link Efficiency Convergence L 3 and L 2 All Active forwarding with built-in L 2 Loop Prevention Leading high availability, convergence, fast reroute capabilities L 2 & L 3 Layers Tie-In Built-in the protocol Optimal Routing Ingress and Egress VM Mobility Optimizations MPLS / IP DC Fabric DC Gateway DC Fabric Virtual Machine Mobility 11 Copyright © 2014 Juniper Networks, Inc.

PTX Series Routers PTX 1000: Distributed Converged Supercore Power 2. 88 Tb Distributed Core Router Performance Powered by Combining Full IP with Express MPLS Express. Plus Deployability OS + SDN 13 *Courting Bits In & Bits Out. Flexible 288 x 10 Gb. E, 72 x 40 Gb. E, 24 x 100 Gb. E Industries Only Fixed Core Router Only 2 RU and 19” Rack Mountable JUNOS: 15 Years of Routing Innovations SDN: 25 Years Perfecting IP/MPLS-TE Traffic Optimization Algorithms Copyright © 2014 Juniper Networks, Inc.

PTX Series Routers PTX Product Family of Routers: Technical Specifications PTX 1 K PTX 3 K PTX 5 K Fixed 2. 88 Tbps per system 1 T/Slot 3 T/Slot 2. 88 Tbps ( 288 x 10 GE) 8 T (80 x 100 GE) 24 T (240 x 100 GE) Typical ~1. 35 KW ~6 k. W 13 k. W* Maximum 1. 65 KW ~7. 2 k. W 18 k. W* Height 2 RU 22 RU 36 RU Depth 31” 270 mm 33” No. of FPCs/PICs N/A 8/8 8/16 Type of FPCs supported N/A SFF-FPC 1/FPC 2/FPC 3 100 GE Density 24 80 240 10 GE Density 288 768 1536 2 HCY 15 Slot Capacity at FRS System Capacity 14 Timing Copyright © 2014 Juniper Networks, Inc.

Physical vs. Virtual Each option has its own strength, and it is created with different focus Physical 16 Virtual High throughput, high density Flexibility to reach higher scale in control plane and service plane Guarantee of SLA Agile, quick to start Low power consumption per throughput Low power consumption per control plan and service Scale up Scale out Higher entry cost in $ and longer time to deploy Lower entry cost in $ and shorter time to deploy Distributed or centralized model Optimal in centralized cloud-centric deployment Well development network mgmt system, OSS/BSS Same platform mgmt as Physical, plus same VM mgmt as a SW on server in the cloud Variety of network interfaces for flexibility Cloud centric, Ethernet-only Excellent price per throughput ratio Ability to apply “pay as you grow” model Copyright © 2014 Juniper Networks, Inc.

VMX overview • Efficient separation of control and data-plane Data packets are switched within v. TRIO Multi-threaded SMP implementation allows core elasticity Only control packets forwarded to JUNOS Feature parity with JUNOS (CLI, interface model, service configuration) NIC interfaces (eth 0) are mapped to JUNOS interfaces (ge-0/0/0) SNMP DCD TCP LCKernel v. TRIO CHASSISD VCP VFP RPD • • • Guest OS (JUNOS) Guest OS (Linux) Hypervisor x 86 Hardware 17 Copyright © 2014 Juniper Networks, Inc.

General consideration of v. MX deployment • v. MX behaves the same way as the physical MX, so it is capable of being the virtual CPE, or virtual PE, virtual BNG, … (*note-1) • Great option as lab equipment for general qualification or function validation • In cloud or NFV deployment, when virtualization is the preferred technology • Fast service enablement without overhead of installing new HW • Solution for scaling services when control plane scale is the bottleneck • When network functions are centralized in DC or cloud • When service separation is preferred by deploying different routing platforms Note-1: Please see product feature & Perf detail in later slides, also see other comments for where v. MX is more a suitable choice. 19 Copyright © 2014 Juniper Networks, Inc.

Agility example: Bring up a new service in a POP 1. Install a new v. MX to start offering a new service without impact to existing platform 4. Integrated the new service into existing PE when the service is mature 2. Scale out the service with v. MX quickly if traffic profile fits the requirements POP v. MX PE L 3 CPE 20 v. MX SP Network for VPN service MX 3. Add service directly to the physical MX GW or add more physical MX if service is successful and there is more demand with significant traffic growth PE L 3 CPE Copyright © 2014 Juniper Networks, Inc.

v. CPE solution with v. MX as the SDN Gateway • v. MX as SDN GW router providing support for BGP and overlay tunneling protocols • v. MX also address the VRF scaling issue for L 3 service chaining Service-chain-X 1 DC SDN GW L 3 CPE VPN-starbucks -core PE SDN GW VPN-starbucks -Hawaii PE SP Network for VPN service VPN-starbucks. NY L 3 CPE New York PE PE Las Vegas Service-chain-Y 1 DC VPN-starbucks-LA LA • v. CPE service • Virtualized services for VPN customer before access to internet or among VPN sites: NAT, Firewall, DPI, caching VPN-starbuckscore VPN-starbucks. NY New Jersy L 3 CPE DC GW Chicago L 3 CPE Honolulu 22 L 3 CPE DC VPN-starbucks-LA VPN-starbucks-NY VPN-starbucks-core VPN-starbucks-Hawaii Copyright © 2014 Juniper Networks, Inc.

Virtual Route Reflector Virtual RR on VMs On standard servers Virtual RR i. BG P Client 3 P i. BGP Client 1 Client n Client 2 23 • v. MX can be used as Route Reflector and deployed the same way as the physical RR in the network • v. MX can act as both v. RR or any typical router function with forwarding capability Copyright © 2014 Juniper Networks, Inc.

Virtual BNG cluster in a data center v. MX as v. BNG cluster v. MX Data Center or CO 10 K~100 K subscribers • • • 24 Potentially BNG function can be virtualized, and v. MX can help form a BNG cluster at the DC or CO (Roadmap item, not at FRS); Suitable to perform heavy load BNG control-plane work while there is little BW needed; Pay-as-you-grow model; Rapid Deployment of new BNG router when needed; Scale-out works well due to S-MPLS architecture, leverages Inter-Domain L 2 VPN, L 3 VPN, VPLS; Copyright © 2014 Juniper Networks, Inc.

v. MX Components VCP (Virtualized Control Plane) VFP (Virtualized Forwarding Plane) • • • Virtual JUNOS to be hosted on a VM Follows standard JUNOS release cycles Additional software licenses for different control plane applications such Virtual Route Reflector • VM that runs the packet forwading engine that is modeled after Trio ASIC Can be hosted on a VM (offer at FRS) OR run as a Linux container (bare-metal) in the future • 26 Copyright © 2014 Juniper Networks, Inc.

VMX system architecture • Optimized data path from physical NIC to v. NIC via SR-IOV (Single Root IO Virtualization). • v. Switch for VFP to VCP communication (internal host path) • Open. Stack (Icehouse) for VM management (Nova) and provisioning of infrastructure network connections (Neutron) Guest VM (Linux + DPDK) Guest VM (Free. BSD) VCP VFP SR-IOV Virtual NICs Hypervisor v. Switch Physical layer Cores Memory Physical NICs 27 Copyright © 2014 Juniper Networks, Inc.

FRS Product Offering • • • FRS at Q 2 2015 with JUNOS release 14. 1 R 4 Function: Provide feature parity with MX except function related to HA and QOS Performance: SR-IOV w/ PCI pass-through, along with DPDK integration Hypervisor support: KVM VM Implementation: VFP to VCP 1: 1 mapping Open. Stack integration (to be finalized) VFP VCP Juniper deliverable Hypervisor/Linux NIC drivers, DPDK Customer defined Server, CPU, NIC 29 Copyright © 2014 Juniper Networks, Inc.

Reference server configuration CPU Intel Xeon 3. 1 GHz Cores Minimum 10 cores RAM 20 GB OS • Ubuntu 14. 04 LTS (w/ libvirt 1. 2. 2, for better performance, upgrade to 1. 2. 6) • Kernel: Linux 3. 13. 0 -32 -generic • Libvert: 1. 2. 6 NICs Intel 82599 EB (for 10 G) QEMU-KVM Version 2. 0 Note: Initially requires minimum 10 Cores: 1 for RE VM, 7 for PFE VM which include 4 packet processing cores, 2 I/O cores, 1 for host 0 if processing), and 2 cores for RE and PFE emulations (QEMU/KVM) ; Later a version with smaller footprint, less # of cores or RAM required 30 Copyright © 2014 Juniper Networks, Inc.

Performance 31 Test setup 8. 9 G 8. 9 G VMX 8. 9 G • Setup: • Single instance of VMX with 6 ports of 10 G sending bidirectional traffic • 16 core total (among those, 6 for I/O, 6 for packet processing) • 20 G RAM total, 16 G memory for v. FP process • Basic routing is enabled, no filter configured • Performance • 60 G bi-directional traffic per VMX instance @ 1500 bytes • No packet loss • Complete RFC 2544 results to follow Tester Copyright © 2014 Juniper Networks, Inc.

Thank you

Related presentations