Help us improve your experience.

Let us know what you think.

Do you have time for a two-minute survey?

 
ON THIS PAGE
 

Solution Architecture

For the purposes of this JVDE, the 3-stage, 5-stage and Collapsed Fabric Data Center designs have been used for Data Center Interconnect (DCI) design. Hence the JVDE focusses on provisioning of the types of DCI using Apstra with Juniper Switches as discussed in Use Case and Reference Architecture section above.

For the purposes of this JVDE, the hardware components discussed in Juniper Hardware and Software Components required for the DCI functionality focuses on the Border Leaf switches. The interconnect switches used for connecting data centers can be any switch/router so long as they support the interconnect functionality. For the lab validation purposes, QFX10002-36Q are used, although configuration of these switches is out of scope of this document, configuration statement references have been provided where necessary.

Note: Each of the DCI designs were configured and tested in isolation. For instance, the over-the-top (OTT) design was configured and validated first. The OTT design is a mutually exclusive design as all of the VXLAN tunnels are formed and stretched across data centers and cannot be mixed with seamless stitching. Apstra also prevents mixing OTT with other Interconnect DCI design for the same reason. However, the Type 2 and Type 5 traffic can be configured and mixed as would be discussed later.

Juniper Hardware and Software Components

For this solution, the Juniper products and software versions are as below.

The design documented in this JVD is considered the baseline representation for the validated solution. As part of a complete solutions suite, we routinely swap hardware devices with other models during iterative use case testing. Each switch platform validated in this document goes through the same rigorous role-based testing using specified versions of Junos OS and Apstra management software.

Juniper Hardware Components

The focus in this JVDE is on the border leaf switch models that were used to interconnect each of these data center fabrics. The table below shows device models and their roles that were used for each of the Fabric Solution:

Table 1: Platform Positioning and Roles
Platforms and Roles
Solution Server Leaf Switches Border Leaf Switches Spine Super Spine
3-stage EVPN/VXLAN Data Center design (ERB) QFX5120-48Y-8C QFX5700 (EVO) QFX5220-32CD (EVO)  
QFX5110-48S PTX10001-36MR QFX5120-32C  
Collapsed Fabric Data Center design* QFX5120-48YM      
5-stage EVPN/VXLAN Data Center design (ERB) QFX5120-48YM QFX5130-48C (EVO) QFX5220-32CD (EVO) QFX5230-64CD (EVO)
QFX5130-32CD (EVO)   QFX5210-64C  
    QFX5120-32C  

* Switches in a collapsed fabric perform the roles of spine, leaf, and border leaf switches.

Note: If one of the devices in the hardware series is tested then all the rest of the variations should work. For instance, QFX5120-48YM covers all the rest of the variations such as QFX5120-48Y as the chipset used is same, however there are some exceptions such as QFX5130-48C and QFX5130-32CD. Contact your Juniper representative for more information.
Table 2: Juniper Software and Version
Juniper Software
Juniper Products Software or Image Version
Juniper Apstra 5.0.0-64
Junos OS 23.4R2-S4

Validated Functionality

This DCI JVDE was validated for below features and functionalities:

  • Deployment of DCI methods using Apstra:
    1. 3-stage Fabric to 3-stage Fabric data center OTT DCI design.
    2. 3-stage to Collapsed Fabric data center VXLAN-to-VXLAN Seamless Transition design.
    3. 3-stage to 5-Stage Fabric data center VXLAN Type 5 to VXLAN Type 5 Seamless transition.
  • Provisioning L2/L3 switches required to interconnect data centers.
  • The hardware and software listed in Table 1 support the features required for DCI EVPN/VXLAN deployment.
  • Equal-Cost Multipath (ECMP) feature Type 2 and Type 5 Route Peering.
  • EVPN route propagation using overlay eBGP sessions and all remote leaves can get to the routes in remote data center PODs/leaves. For instance, the servers in DC1 3-stage can reach the servers deployed in DC2.
  • Both IPV4 and IPV6 are enabled for routing between the data centers
  • BFD is enabled for overlay DCI using configlet.
  • IRB enabled within PODs and across data center PODs/leaves for inter-subnet forwarding.
  • Inter VRF route leaking configured to reach routes in DCI connected data center PODs/leaves
  • Configure MACSEC (using configlets in Apstra) between 3-stage data center with QFX5700 as border leaves and collapsed fabric with QFX5120-48YM as leaves.