QFX5700 Switch Datasheet

The QFX5700 provides enhanced scale with up to 1.24 million routes, 80,000 firewall filters, and 160,000 media access control (MAC) addresses. It supports high numbers of egress IPv4/IPv6 rules by programming matches in egress ternary content addressable memory (TCAM) along with ingress TCAM.

Today’s cloud-native applications have critical dependency on buffer size to prevent congestion and packet drops. The QFX5700 has 132 MB shared packet buffer that is allocated dynamically to congested ports.

• Automation and programmability: The QFX5700 supports several network automation features for plug-and-play operations, including zero-touch provisioning (ZTP), Network Configuration Protocol (NETCONF), Juniper Extension Toolkit (JET), Junos telemetry interface, operations and event scripts, automation rollback, and Python scripting.
• Cloud-level scale and performance: The QFX5700 supports best-in-class cloud-scale L2/L3 deployments with a latency as low as 900 ns and superior scale and performance. This includes L2 support for 160,000 MAC addresses and Address Resolution Protocol (ARP) learning, which scales up to 64,000 entries at 500 frames per second. It also includes L3 support for 1.24 million longest prefix match (LPM) routes and 160,000 host routes on IPv4. Additionally, the QFX5700 supports 610,000 LPM routes and 80,000 host routes on IPv6, 128-way equal-cost multipath (ECMP) routes, and a filter that supports 80,000 ingress and 18,000 egress exactly match filtering rules. The QFX5700 supports up to 128 link aggregation groups, 4096 VLANs, and Jumbo frames of 9216 bytes. Junos OS Evolved provides configurable options through a CLI, enabling each QFX5700 to be optimized for different deployment scenarios.
• VXLAN overlays: The QFX5700 is capable of both L2 and L3 gateway services. Customers can deploy overlay networks to provide L2 adjacencies for applications over L3 fabrics. The overlay networks use VXLAN in the data plane and EVPN or Open vSwitch Database (OVSDB) for programming the overlays, which can operate without a controller or be orchestrated with an SDN controller.
• RoCEv2: As a switch capable of transporting data as well as storage traffic over Ethernet, the QFX5700 provides an IEEE data center bridging (DCB) converged network between servers with disaggregated flash storage arrays or an NVMe-enabled storage-area network (SAN). The QFX5700 offers a full-featured DCB implementation that provides strong monitoring capabilities on the top-of-rack switch for SAN and LAN administration teams to maintain clear separation of management. The RDMA over Converged Ethernet version 2 (RoCEv2) transit switch functionality, including priority-based flow control (PFC) and Data Center Bridging Capability Exchange (DCBX), are included as part of the default software.
• Junos Evolved features: The QFX5700 switch supports features such as L2/L3 unicast, EVPN-VXLAN, BGP add-path, RoCEv2 and congestion management, multicast, 128-way ECMP, dynamic load balancing capabilities, enhanced firewall capabilities, and monitoring.
• Junos OS Evolved Architecture: Junos OS Evolved is a native Linux operating system that incorporates a modular design of independent functional components and enables individual components to be upgraded independently while the system remains operational. Component failures are localized to the specific component involved and can be corrected by upgrading and restarting that specific component without having to bring down the entire device. The switches control and data plane processes can run in parallel, maximizing CPU utilization, providing support for containerization, and enabling application deployment using LXC or Docker.
• Retained state: State is the retained information or status pertaining to physical and logical entities. It includes both operational and configuration state, comprising committed configuration, interface state, routes, hardware state, and what is held in a central database called the distributed data store (DDS). State information remains persistent, is shared across the system, and is supplied during restarts.
• Feature support: All key networking functions such as routing, bridging, management software, and management plane interfaces, as well as APIs such as CLI, NETCONF, JET, Junos telemetry interface, and the underlying data models, resemble those supported by the Junos operating system. This ensures compatibility and eases the transition to Junos Evolved.

The QFX5700 can be deployed as a universal device in cloud data centers to support 100GbE server access and 400GbE spine-and leaf configurations, optimizing data center operations by using a single device across multiple network layers (see Figure 1). The QFX5700 can also be deployed in more advanced overlay architectures like an EVPN-VXLAN fabric. Depending on where tunnel terminations are desired, the QFX5700 can be deployed in the EVPN-VXLAN Edge Routed Bridging architecture or in Bridged Overlay (BO).

Juniper offers complete flexibility and range of data center fabric designs that cater to data centers of different sizes, scale built by Cloud Operators, Service Providers and Enterprises. Here’re the main data center fabric design options where QFX5700 can be used for high port density server leaf, spine node or border-leaf node:

• Architecture 1: ERB - Edge Routed Bridging EVPN-VXLAN with distributed anycast IP gateway architecture supporting L2 and L3 for Enterprises and 5G Telco-Cloud. This type of design offers a combination of L2 stretch between multiple leaf/ToR switches and L2 active/active multihoming to the server with MAC-VRF EVI L2 virtualization support as well as L3 IP VRF virtualization at the leaf/ToR through the Type-5 EVPN-VXLAN. This type of design in DC use-case can be used to connect in redundant and optimized way the servers/compute nodes, Blade Centers, IP storage nodes running ROCEv2, as well as other appliances.
• Architecture 2: BO - Bridged Overlay EVPN-VXLAN design using MAC-VRF instances and different EVPN service-types (vlan-aware, vlan-bundle, vlan-based). In this case an external to the fabric first hop IP gateway can be used – for example at the firewall or external existing DC gateway routers. In this design the DC fabric is offering L2 active/active multihoming using ESI-LAG and fabric wide L2 stretch between the leaf ToR nodes.  
• Architecture 3: Seamless Data Center Interconnect (DCI) for ERB fabric design – DCI border-leaf design with seamless T2/T2 EVPN-VXLAN to EVPN-VXLAN tunnel stitching (RFC 9014) and T5/T5 EVPN-VXLAN tunnel stitching support. With this design, the data center benefits from geographical redundancy for the application deployed in private cloud DC. In this case, the QFX5700 can also be used as a border-leaf node.
• Architecture 4: Collapsed Spine design with ESI-LAG support and anycast IP – in this case the pair of QFX5700 switches is deployed with a back-to-back connect, without spine layer. The L2 active/active multihoming using ESI-LAG is used for the server NIC high availability as well as anycast IP gateway.

Juniper^® Apstra provides operators with the power of intent -based network design to help ensure changes required to enable data center services can be delivered rapidly, accurately, and consistently. Operators can further benefit from the built-in assurance and analytics capabilities to resolve Day 2 operations issues quickly.

Apstra key features are:

• Automated deployment and zero-touch deployment
• Continuous fabric validation
• Fabric life-cycle management
• Troubleshooting using advanced telemetry

For more information on Apstra, see www.juniper.net/us/en/products/network-automation/apstra/apstra-system.html.

The QFX5700 switches can be deployed in campus distribution/core layer networks using 10GbE/25GbE/40GbE/100GbE ports to support technologies such as EVPN multihoming and campus Fabric.

Juniper offers complete flexibility in choosing any of the following validated EVPN-VXLAN designs that cater to networks of different sizes, scale, and segmentation requirements:

• EVPN multihoming (collapsed core or distribution): A collapsed core architecture combines the core and distribution layers into a single switch, turning the traditional three-tier hierarchal network into a two-tier network. EVPN Multihoming on a collapsed core eliminates the need for Spanning Tree Protocol (STP) across campus networks by providing link aggregation capabilities from the access layer to the core layer. This topology is best suited for small to medium distributed enterprise networks and allows for consistent VLANs across the network. This topology uses ESI (Ethernet Segment Identifier) LAG (Link Aggregation) and is a standards-based protocol.
• Campus Fabric Core distribution: When EVPN VXLAN is configured across core and distribution layers, it becomes a campus Fabric Core Distribution architecture, which can be configured in two modes: centrally or edge routed bridging overlay. This architecture provides an opportunity for an administrator to move towards campus-fabric IP Clos without fork-lift upgrade of all access switches in the existing network, while bringing in the advantages of moving to a campus fabric and providing an easy way to scale out the network.
• Campus Fabric IP Clos: When EVPN VXLAN is configured on all layers including access, it is called the campus fabric IP Clos architecture. This model is also referred to as “end-to-end,” given that VXLAN tunnels are terminated at the access layer. The availability of VXLAN at access provides us with the opportunity to bring policy enforcement and microsegmentation to the access layer (closest to the source) using standards-based Group Based Policy (GBP) to segment traffic even within a VLAN. GBP tags are assigned dynamically to clients as part of Radius transaction by Mist Cloud NAC. This topology works for small-medium and large campus architectures that need macro and microsegmentation.

In all these EVPN-VXLAN deployment modes, QFX5700 switches can be used in Distribution or core as seen in Figure 2. All three topologies are standards-based and hence are inter-operable with 3rd party vendors. 

Juniper Mist Wired Assurance brings cloud management and Mist AI to campus fabric. It sets a new standard that moves away from traditional network management towards AI-driven operations, while delivering better experiences to connected devices. The Juniper Mist Cloud streamlines deployment and management of campus fabric architectures by allowing:

• Automated deployment and zero touch deployment (ZTD)
• Anomaly detection
• Root cause analysis

For more information, read the Juniper Mist Wired Assurance datasheet.

The QFX5700 can be used in L2 fabrics and L3 networks. You can choose the architecture that best suits your deployment needs and easily adapt and evolve as requirements change over time. The QFX5700 serves as the universal building block for these switching architectures, enabling data center operators to build cloud networks in their own way.

Layer 3 fabric: For customers looking to build scale-out data centers, a Layer 3 spine-and-leaf Clos fabric provides predictable, nonblocking performance and scale characteristics. A two-tier fabric built with QFX5700 switches as leaf devices and Juniper Networks QFX10000 modular switches in the spine can scale to support up to 128 40GbE ports or 128 25GbE and/or 10GbE server ports in a single fabric.

Junos OS Evolved ensures a high feature and bug fix velocity and provides first-class access to system state, allowing customers to run DevOps tools, containerized applications, management agents, specialized telemetry agents, and more.

The QFX5700 supports Junos telemetry interface, a modern telemetry streaming tool that provides performance monitoring in complex, dynamic data centers. Streaming data to a performance management system lets network administrators measure trends in link and node utilization and troubleshoot issues such as network congestion in real time.

Junos telemetry interface provides:

• Application visibility and performance management by provisioning sensors to collect and stream data and analyze the application and workload flow path through the network
• Capacity planning and optimization by proactively detecting hotspots and monitoring latency and microbursts
• Troubleshooting and root cause analysis via high frequency monitoring and correlating overlay and underlay networks

• MAC addresses per system: 160,000
• VLAN IDs: 4000 (QFX5700)
• Number of link aggregation groups (LAGs): 128
• Number of ports per LAG: 64
• Firewall filters: 

• IPv4 unicast routes: 1.24 million prefixes; 160,000 host routes
• IPv6 unicast routes: 610,000 prefixes; 80,000 host routes
• ARP entries: 32,000 (tunnel mode); 64,000 (non-tunnel mode)
• Neighbor Discovery Protocol (NDP) entries: 32,000 (tunnel mode); 64,000 (non-tunnel mode)
• Generic routing encapsulation (GRE) tunnels: 1000
• Jumbo frame: 9216 bytes
• Traffic mirroring: 8 destination ports per switch

• STP—IEEE 802.1D (802.1D-2004)
• Rapid Spanning Tree Protocol (RSTP) (IEEE 802.1w); MSTP (IEEE 802.1s)
• Bridge protocol data unit (BPDU) protect
• Loop protect
• Root protect
• RSTP and VLAN Spanning Tree Protocol (VSTP) running concurrently
• VLAN—IEEE 802.1Q VLAN trunking
• Routed VLAN interface (RVI)
• Port-based VLAN
• Static MAC address assignment for interface
• MAC learning disable
• Link Aggregation and Link Aggregation Control Protocol (LACP) (IEEE 802.3ad)
• IEEE 802.1AB Link Layer Discovery Protocol (LLDP)
  
  

• LAG load sharing algorithm—bridged or routed (unicast or multicast) traffic:
  • IP: Session Initiation Protocol (SIP), Dynamic Internet Protocol (DIP), TCP/UDP source port, TCP/UDP destination port
  • L2 and non-IP: MAC SA, MAC DA, Ether type, VLAN ID, source port

• Static routing
• OSPF v1/v2
• OSPF v3
• Filter-based forwarding
• Virtual Router Redundancy Protocol (VRRP)
• IPv6
• Virtual routers
• Loop-free alternate (LFA)
• BGP (Advanced Services or Premium Services license)
• IS-IS (Advanced Services or Premium Services license)
• Dynamic Host Configuration Protocol (DHCP) v4/v6 relay
• VR-aware DHCP
• IPv4/IPv6 over GRE tunnels (interface-based with decap/ encap only)
  
  

• Internet Group Management Protocol (IGMP) v1/v2
• Multicast Listener Discovery (MLD) v1/v2
• IGMP proxy, querier
• IGMP v1/v2/v3 snooping
• MLD snooping
• Protocol Independent Multicast PIM-SM, PIM-SSM, PIM-DM

• Secure interface login and password
• Secure boot
• RADIUS
• TACACS+
• Ingress and egress filters: Allow and deny, port filters, VLAN filters, and routed filters, including management port filters and loopback filters for control plane protection
• Filter actions: Logging, system logging, reject, mirror to an interface, counters, assign forwarding class, permit, drop, police, mark
• SSH v1, v2
• Static ARP support
• Storm control, port error disable, and autorecovery
• Control plane denial-of-service (DoS) protection
• Image rollback

• L2 and L3 QoS: Classification, rewrite, queuing
• Rate limiting:
  • Ingress policing: 1 rate 2 color, 2 rate 3 color
  • Egress policing: Policer, policer mark down action
  • Egress shaping: Per queue, per port
• 12 hardware queues per port (8 unicast and 4 multicast)
• Strict priority queuing (LLQ), shaped-deficit weighted round-robin (SDWRR), weighted random early detection (WRED)
• 802.1p remarking
• Layer 2 classification criteria: Interface, MAC address, Ether type, 802.1p, VLAN
• Congestion avoidance capabilities: WRED
• Trust IEEE 802.1p (ingress)
• Remarking of bridged packets

• EVPN support with VXLAN transport
• All-active multihoming support for EVPN-VXLAN (ESI-LAG aka EVPN-LAG)
• MAC-VRF (EVI) multiple EVPN service-type support: vlan-based, vlan-aware, vlan-bundle
• ARP/ND suppression aka proxy-arp/nd
• Ingress multicast Replication
• IGMPv2 snooping support fabric wide: using EVPN route type-6
• IGMPv2 snooping support for L2 multihoming scenarios: EVPN route type-7 and type-8
• IP prefix advertisement using EVPN with VxLAN encapsulation

• Explicit congestion notification (ECN)
• Data Center Bridging Quantized Congestion Notification (DCQCN)
• Priority-based flow control (PFC)—IEEE 802.1Qbb
• Priority-based flow control (PFC) using Differentiated Services code points (DSCP) at Layer 3 for untagged traffic
• Remote Direct Memory Access (RDMA) over converged Ethernet version 2 (RoCEv2)

• Bidirectional Forwarding Detection (BFD)
  
  

• Switched Port Analyzer (SPAN)
• Remote SPAN (RSPAN)
• Encapsulated Remote SPAN (ERSPAN)
• sFlow v5
• Junos telemetry interface

• Role-based CLI management and access
• CLI via console, telnet, or SSH
• Extended ping and traceroute
• Junos OS Evolved configuration rescue and rollback
• SNMP v1/v2/v3
• Junos OS Evolved XML management protocol
• High frequency statistics collection
• Automation and orchestration
• Zero-touch provisioning (ZTP)
• Python
• Junos OS Evolved event, commit, and OP scripts
• Juniper Apstra Management, Monitoring, and Analytics for Data Center Fabrics
• Juniper Mist Wired Assurance for Campus

• IEEE 802.1D
• IEEE 802.1w
• IEEE 802.1
• IEEE 802.1Q
• IEEE 802.1p
• IEEE 802.1ad
• IEEE 802.3ad
• IEEE 802.1AB
• IEEE 802.3x
• IEEE 802.1Qbb
• IEEE 802.1Qaz
• T11 Standards
• INCITS T11 FC-BB-5

• CAN/CSA-C22.2 No. 60950-1 Information Technology Equipment—Safety
• UL 60950-1 Information Technology Equipment—Safety
• EN 60950-1 Information Technology Equipment—Safety
• IEC 60950-1 Information Technology Equipment—Safety (All country deviations)
• EN 60825-1 Safety of Laser Products—Part 1: Equipment Classification
• UL 62368-1 Second Edition
• UL IEC 62328-1 Second Edition

• FCC 47 CFR Part 15
• ICES-003 / ICES-GEN
• EN 300 386 V1.6.1
• EN 300 386 V2.1.1
• EN 55032
• CISPR 32
• EN 55024
• CISPR 24
• EN 55035
• CISPR 35
• IEC/EN 61000 Series
• AS/NZS CISPR 32
• VCCI-CISPR 32
• BSMI CNS 13438
• KN 32 and KN 35
• KN 61000 Series
• TEC/SD/DD/EMC-221/05/OCT-16
• TCVN 7189
• TCVN 7317
  

• Common Language Equipment Identifier (CLEI) code

  Restriction of Hazardous Substances (ROHS) 6/6

     China Restriction of Hazardous Substances (ROHS)

  Registration, Evaluation, Authorization and Restriction of Chemicals (REACH)

  Waste Electronics and Electrical Equipment (WEEE)

      Recycled material

      80 Plus Silver PSU Efficiency

Juniper Networks is the leader in performance-enabling services that are designed to accelerate, extend, and optimize your high-performance network. Our services allow you to maximize operational efficiency while reducing costs and minimizing risk, achieving a faster time to value for your network. Juniper Networks ensures operational excellence by optimizing the network to maintain required levels of performance, reliability, and availability. For more details, please visit https://www.juniper.net/us/en/products.html.

QFX5700 supports varying port speeds at 400G, 100G, 50G, 40G, 25G,  and 10G with different transceiver options of direct attach copper cables, active optical cables, break out cables (DACBO and AOCBO). Up-to-date information on supported optics can be found on the Hardware Compatibility Tool at https://apps.juniper.net/hct/product/?prd=QFX5700.

Juniper Networks believes that connectivity is not the same as experiencing a great connection. Juniper's AI-Native Networking Platform is built from the ground up to leverage AI to deliver the best and most secure user experiences from the edge to the data center and cloud. Additional information can be found at Juniper Networks (www.juniper.net) or connect with Juniper on X (Twitter), LinkedIn, and Facebook.