Juniper® Validated Design: JVD Test Report Brief: Data Center Interconnect over IPoDWDM

Introduction

The traditional Dense Wave Division Multiplexing (DWDM) networks utilize transponders to convert Ethernet signals into a DWDM signal suitable for DWDM transport. The Converged Optical Routing Architecture (CORA) integrates DWDM optics directly into Juniper routers and switches. This integration allows DWDM optics in a router to connect directly to a DWDM multiplexer, eliminating the need for a separate transponder. In this model, the Internet Protocol (IP) and Optical network management operate as a single domain controller, which simplifies operations, lowers operational expenses, increases network efficiency, and lowers power consumption. It also allows the router to monitor the performance of the DWDM link and make routing decisions based on that performance, aiding in faster troubleshooting and reduced downtime.

This testing demonstrates Juniper 400G Coherent Optics capabilities and validates the Internet Protocol over Dense Wavelength-Division Multiplexing (IPoDWDM) solution.

Major Technical Attributes:

For Amplified Path: Receives minimum optical signal-to-noise ratio, Maximum chromatic dispersion.
For Unamplified Path: RX sensitivity, Telemetry, Configurability, Performance monitoring.
Juniper 400G Coherent Optics Tested: JCO400-QDD-ZR-M-HP, QDD-400G-ZR-M-HP.

Test Topology

Figure 1: Logical Topology

The logical topology illustrates an IP-MPLS transport network using OSPF, RSVP-based LSP, MP-iBGP sessions (family inet-vpn) with BFD for both amplified and unamplified paths. It shows DCI1 and DCI2 connected via a 50km link, with transceivers (TRX) and multiplexers (Mux) at each end. The setup includes various virtual routing instances (vrf1, vrf#2, vrf#100) and customer edge devices (STC1, STC2, STC3).

Figure 2: Amplified Path Topology

The amplified path topology details a setup involving ADTRAN OLS (Optical Line System), a 50/50 Optical Splitter/Combiner, and a Configurable CD Emulator. An ASE Source is used to emulate noise. This diagram shows the connection between DCI1 (PTX10001-36MR) and DCI2 (MX304) through amplified optical paths, highlighting the role of the transceivers (TRX) and the optical components.

Figure 3: Unamplified Path Topology

The unamplified path topology depicts connections using Fiber Optic Cable Pairs #1 and #2 between DCI1 (PTX10001-36MR), DCI3 (ACX7100-48L), and DCI2 (MX304). Each pair carries two wavelengths. A Variable Optical Attenuator is used to emulate span loss. This setup tests the performance without optical amplification.

Platforms Tested

Table 1 lists the platforms tested during the initial qualification of this JVD. For detailed information on supported platforms and OS versions, refer to the Validated Platforms and Software section in the JVD document.

Table 1: Devices Under Test
Role	Model	OS
DCI-Edge1	PTX10001-36MR	Junos OS Evolved Release 24.2R2
DCI-Edge2	MX304	Junos OS Release 24.2R2
DCI-Edge3	ACX7100-48L	Junos OS Evolved Release 24.2R2
Transceiver	JCO400-QDD-ZR-M-HP	N/A
Transceiver	QDD-400G-ZR-M-HP	N/A

Version Qualification History

This JVD was initially qualified on Junos OS Release 24.2R2 and Junos OS Evolved Release 24.2R2. For more details on all supported platforms and OS versions, see the Validated Platforms and Software section in the JVD document.

Scale Data

The scaling data is as follows:

MPLS: 100 RSVP based LSP between DCIs
Routing Instances: 100VPNv4
Route: 20000
100 eBGP sessions

Performance Data

Performance tests are performed as follows:

Amplified Path

For the amplified path, three optical transceivers are used on both DCI1 and DCI2 routers. All transceivers and Reconfigurable Optical Add-Drop Multiplexer (ROADM) ports are tuned to a specific wavelength or frequency. EDFA noise is emulated by injecting noise using an Amplified Spontaneous Emission (ASE) source. The noise is increased gradually using the Variable Optical Attenuator (VOA). Chromatic Dispersion (CD) is also emulated using Chromatic Dispersion Emulators (CDE).

The process continues until traffic loss is observed or the channel goes down. Noise is decreased until traffic is stable for 2 hours. Then, the Pre-Forward Error Correction Bit Error Rate (Pre-FEC BER), uncorrected Frame Error Rate (FER), and OSNR are measured. The measured OSNR is the minimum receive-OSNR requirement of the transceiver. An Optical Spectrum Analyzer (OSA) is used to verify the OSNR reported by the transceiver.

Table 2: Amplified Path Test Results with JCO400-QDD-ZR-M-HP as Rx
Speed/Media Code	CD	rOSNR (reported by the transceiver)			rOSNR (measured by an OSA)	Pre-FEC BER (reported by the transceiver)
		Min	Max	Avg		Min	Max	Avg
400GE ZR400-OFEC-16QAM	0 ns/nm	22.2 dB	22.4 dB	22.2 dB	21.2 dB	1.74E-02	1.83E-02	1.78E-02
	5 ns/nm	22.2 dB	22.4 dB	22.4 dB	21.5 dB	1.70E-02	1.84E-02	1.77E-02
	10 ns/nm	22.4 dB	22.7 dB	22.4 dB	22.0 dB	1.65E-02	1.80E-02	1.72E-02
	15 ns/nm	22.7 dB	22.9 dB	22.9 dB	22.6 dB	1.47E-02	1.60E-02	1.53E-02
	20 ns/nm	23.7 dB	23.9 dB	23.9 dB	24.4 dB	1.04E-02	1.16E-02	1.09E-02
3x100GE ZR300-OFEC-8QAM	0 ns/nm	18.9 dB	19.2 dB	19.0 dB	18.3 dB	1.58E-02	1.74E-02	1.66E-02
	20 ns/nm	19.4 dB	19.7 dB	19.4 dB	19.5 dB	1.29E-02	1.40E-02	1.34E-02
2x100GE ZR200-OFEC-QPSK	0 ns/nm	15.0 dB	15.0 dB	15.0 dB	14.2 dB	1.44E-02	1.59E-02	1.51E-02
	20 ns/nm	15.0 dB	15.0 dB	15.0 dB	14.5 dB	1.51E-02	1.63E-02	1.57E-02
1x100GE ZR100-OFEC-QPSK	0 ns/nm	12.0 dB	12.0 dB	12.0 dB	11.1 dB	1.49E-02	1.65E-02	1.57E-02
	20 ns/nm	12.0 dB	12.0 dB	12.0 dB	11.3 dB	1.46E-02	1.64E-02	1.54E-02

Table 3: Amplified Path Test Results with QDD-400G-ZR-M-HP as Rx
Speed/Media Code	CD	rOSNR (reported by the transceiver)			rOSNR (measured by an OSA)	Pre-FEC BER (reported by the transceiver)
		Min	Max	Avg		Min	Max	Avg
400GE ZR400-OFEC-16QAM	0 ns/nm	19.1 dB	19.4 dB	19.2 dB	21.6 dB	1.70E-02	1.83E-02	1.70E-02
	5 ns/nm	19.2 dB	19.6 dB	19.4 dB	22.0 dB	1.73E-02	1.83E-02	1.70E-02
	10 ns/nm	19.2 dB	19.5 dB	19.4 dB	22.2 dB	1.78E-02	1.87E-02	1.80E-02
	1
15 ns/nm	19.4 dB	19.7 dB	19.5 dB	22.5 dB	1.74E-02	1.84E-02	1.70E-02
3x100GE ZR300-OFEC-8QAM	20 ns/nm	19.4 dB	19.9 dB	19.7 dB	22.8 dB	1.70E-02	1.81E-02	1.70E-02
	0 ns/nm	18.4 dB	18.7 dB	18.5 dB	18.4 dB	1.73E-02	1.86E-02	1.70E-02
2x100GE ZR200-OFEC-QPSK	20 ns/nm	18.7 dB	19.0 dB	18.8 dB	19.0 dB	1.74E-02	1.82E-02	1.78E-02
	0 ns/nm	13.6 dB	13.9 dB	13.7 dB	13.7 dB	1.86E-02	1.92E-02	1.89E-02
1x100GE ZR100-OFEC-QPSK	20 ns/nm	14.0 dB	14.1 dB	14.0 dB	14.1 dB	1.80E-02	1.91E-02	1.80E-02
	0 ns/nm	10.6 dB	10.8 dB	10.7 dB	10.7 dB	1.82E-02	1.94E-02	1.80E-02
	20 ns/nm	10.7 dB	10.9 dB	10.8 dB	10.9 dB	1.81E-02	1.94E-02	1.80E-02

For both transceivers, increasing rOSNR penalties are observed as CD increases. This should be considered when designing the network.

Unamplified Path

For the unamplified/dark-fiber path, two optical transceivers are used on each DCI router, tuned to a specific wavelength or frequency. Two pairs of fiber optic cables are used, each carrying two wavelengths. In the unamplified path, signal loss occurs mainly due to light absorption by the fiber material, known as Span Loss. The Echola Variable Optical Attenuator (VOA) is used to emulate span loss.

The link is attenuated gradually until traffic loss is observed or the channel goes down. Noise is decreased until traffic is stable for 2 hours. The Pre-Forward Error Correction Bit Error Rate (Pre-FEC BER), uncorrected Frame Error Rate (FER), and Rx Power are measured. The measured Rx Power is the Rx Sensitivity of the transceiver, verified by a power meter.

The results are presented in two tables per transceiver. The first table shows Receive (Rx) Signal Power, and the second table shows Rx Total Power. Rx Signal Power is the power received from the desired wavelength/frequency. Rx Total Power is the power received from all wavelengths/frequencies. For this test case, there are 2 wavelengths/frequencies per cable. Since the Transmit (Tx) Power for both signals is equal, the Rx Total Power is 3dB higher than the Rx Signal Power.

Table 4: Unamplified Path Test Results – Rx Signal Power for JCO400-QDD-ZR-M-HP
Speed/Media Code	Rx Signal Power (reported by the transceiver)			Rx Signal Power (measured by a power meter)	Pre-FEC BER (reported by the transceiver)
	Min	Max	Avg		Min	Max	Avg
400GE ZR400-OFEC-16QAM	-26.64 dBm	-26.36 dBm	-26.50 dBm	-26.90 dBm	1.30E-02	1.41E-02	1.35E-02
3x100GE ZR300-OFEC-8QAM	-29.19 dBm	-28.95 dBm	-29.07 dBm	-30.00 dBm	1.22E-02	1.31E-02	1.27E-02
2x100GE ZR200-OFEC-QPSK	-33.03 dBm	-30.72 dBm	-31.30 dBm	-34.20 dBm	1.12E-02	1.32E-02	1.24E-02
1x100GE ZR100-OFEC-QPSK	-32.97 dBm	-32.72 dBm	-32.84 dBm	-37.80 dBm	1.15E-02	1.26E-02	1.21E-02

The JCO400-QDD-ZR-M-HP on 100GE Mode, ZR100-OFEC-QPSK can only link up at -32dBm Rx Power. The link can degrade to -37.80dBm after linking up. Thus, design the link at -32dBm Rx Sensitivity or higher.

Table 5: Unamplified Path Test Results – Rx Total Power for JCO400-QDD-ZR-M-HP
Speed/Media Code	Rx Total Power (reported by the transceiver)			Rx Total Power (measured by a power meter)	Pre-FEC BER (reported by the transceiver)
	Min	Max	Avg		Min	Max	Avg
400GE ZR400-OFEC-16QAM	-24.44 dBm	-23.87 dBm	-24.21 dBm	-24.00 dBm	1.30E-02	1.41E-02	1.35E-02
3x100GE ZR300-OFEC-8QAM	-27.70 dBm	-26.58 dBm	-27.20 dBm	-27.10 dBm	1.22E-02	1.31E-02	1.27E-02
2x100GE ZR200-OFEC-QPSK	-32.22 dBm	-29.59 dBm	-30.81 dBm	-31.30 dBm	1.12E-02	1.32E-02	1.24E-02
1x100GE ZR100-OFEC-QPSK	-36.99 dBm	-31.55 dBm	-33.60 dBm	-34.80 dBm	1.15E-02	1.26E-02	1.21E-02

Table 6: Unamplified Path Test Results -Rx Signal Power for QDD-400G-ZR-M-HP
Speed/Media Code	Rx Signal Power (reported by the transceiver)			Rx Signal Power (measured by a power meter)	Pre-FEC BER (reported by the transceiver)
	Min	Max	Avg		Min	Max	Avg
400GE ZR400-OFEC-16QAM	-20.82 dBm	-20.60 dBm	-20.71 dBm	-20.50 dBm	1.38E-02	1.43E-02	1.41E-02
3x100GE ZR300-OFEC-8QAM	-26.99 dBm	-25.14 dBm	-26.17 dBm	-24.60 dBm	1.62E-02	1.68E-02	1.65E-02
2x100GE ZR200-OFEC-QPSK	-40.00 dBm	-40.00 dBm	-40.00 dBm	-29.40 dBm	1.37E-02	1.43E-02	1.40E-02
1x100GE ZR100-OFEC-QPSK	-40.00 dBm	-40.00 dBm	-40.00 dBm	-32.30 dBm	6.14E-03	7.05E-03	6.43E-03

The QDD-400G-ZR-M-HP's Rx Signal Power Monitor is only accurate down to -21dBm. Any Rx Signal Power values below -21dBm are not guaranteed.

Table 7: Unamplified Path Test Results -Rx Total Power for QDD-400G-ZR-M-HP
Speed/Media Code	Rx Total Power (reported by the transceiver)			Rx Total Power (measured by a power meter)	Pre-FEC BER (reported by the transceiver)
	Min	Max	Avg		Min	Max	Avg
400GE ZR400-OFEC-16QAM	-17.72 dBm	-17.62 dBm	-17.67 dBm	-17.40 dBm	1.38E-02	1.43E-02	1.41E-02
3x100GE ZR300-OFEC-8QAM	-21.94 dBm	-21.80 dBm	-21.87 dBm	-21.60 dBm	1.62E-02	1.68E-02	1.65E-02
2x100GE ZR200-OFEC-QPSK	-26.78 dBm	-26.58 dBm	-26.65 dBm	-26.30 dBm	1.37E-02	1.43E-02	1.40E-02
1x100GE ZR100-OFEC-QPSK	-31.55 dBm	-30.97 dBm	-30.99 dBm	-29.20 dBm	6.14E-03	7.05E-03	6.43E-03

Channel Sweep Tests

Link stability and integrity are evaluated across various channel frequencies by performing controlled channel sweep tests across the Frequency Range: 191.300 THz to 196.100 THz with the following frequency step sizes:

50 GHz Grid
75 GHz Grid
100 GHz Grid
Flex Grid (6.25GHz)

For each frequency size, the tests verify that:

The optical link is successfully established and remained stable.
No alarms are triggered during the testing process.
No traffic loss is observed once the link is stabilized.

High-Level Features Tested

The high-level features tested successfully include OSPF, IBGP, EBGP, MPLS (RSVP and LDP), BFD, AE (LACP), FRR, and EVPN-VXLAN.

Event Testing

The following is the list of test cases tested successfully:

Restart DUT
FPC restart
PIC offline/online
PFE off/on
Restart apps
Laser off/on
Link Flap
Unplug/plug optics
Clear ospf
Clear bgp
Deactivate/activate AE member
Trigger Link Fault
GRES
NSR

Known Limitations

The known limitation detail is as follows:

Limitation: Changing bandwidth causes interface/optics to remain in a down state.
Workaround: A soft OIR (Online Insertion and Removal) of the optical module will restore the interface to an operational (up) state.

Contact Information

Corporate and Sales Headquarters

Juniper Networks, Inc.
1133 Innovation Way
Sunnyvale, CA 94089 USA
Phone: 888.JUNIPER (888.586.4737) or +1.408.745.2000
Fax: +1.408.745.2100
www.juniper.net

APAC and EMEA Headquarters

Juniper Networks International B.V.
Boeing Avenue 240
1119 PZ Schiphol-Rijk
Amsterdam, The Netherlands
Phone: +31.207.125.700
Fax: +31.207.125.701

	Juniper 400G Optical Transceivers and Cables Guide Explore Juniper Networks' 400G optical transceivers and cables with this comprehensive guide. Learn about QSFP-DD form factors, ZR/ZR+ optics, PAM4 modulation, DSP, FEC, installation, and maintenance for high-speed data center and metro networking solutions.
	Juniper Validated Design for Data Center Interconnect over IPoDWDM This document outlines the Juniper Validated Design (JVD) for Data Center Interconnect (DCI) using Internet Protocol over Dense Wavelength-Division Multiplexing (IPoDWDM). It details the use of Juniper ACX7000, MX Series, and PTX Series Routers, along with Juniper 400G Coherent Optics and ADTRAN Open Line System (OLS).
	Juniper Validated Design: 5G Mobile xHaul with Seamless MPLS Segment Routing Explore Juniper Networks' validated design for 5G xHaul networks, detailing seamless MPLS Segment Routing, ACX7000 series, MX series, and PTX series integration, performance, and QoS for mobile backhaul and fronthaul applications.
	Junos OS Evolved Release Notes 23.2R2 Release notes for Junos OS Evolved Release 23.2R2, detailing new and updated features, software limitations, and open issues for various Juniper Networks devices.
	Juniper Validated Design for Enterprise WAN Advanced Core and Edge Services A comprehensive guide to deploying Juniper Networks solutions for advanced enterprise WAN core and edge services, focusing on MPLS-based backbones, EVPN, and Segment Routing for enhanced connectivity, scalability, performance, and security.
	Juniper Routing Assurance Quick Start: Onboard Routers Learn how to onboard routers and monitor their health and performance using Juniper Routing Assurance with this quick start guide. Covers prerequisites, setup, site assignment, and insights for Juniper routers.
	APAC Cohesion: Juniper Automated WAN Systems (JAWS) Tech Roundup Q4-2022 A technical roundup of Juniper Automated WAN Systems (JAWS) and Paragon AI driven core network routing, covering PTX and MX series platforms, Express ASIC evolution, EANTC test results, SmartWall TDD capabilities, and licensing information.
	Junos OS Evolved Release 24.2R1 Release Notes Discover the new features, updated functionalities, software limitations, and known issues in Junos OS Evolved Release 24.2R1 for Juniper Networks ACX Series routers. This document details enhancements in areas such as Authentication and Access Control, Class of Service, EVPN, Interfaces, and more.