VMWare acquires VeloCloud – what will this mean for SD-WAN?

The move pitches VMWare directly against Cisco in key enterprise branch router market.

VMWare announced on November 2 that it intends to acquire VeloCloud Networks. The value of the proposed acquisition has not yet been announced, and VMWare expects it to close in Q4 of its FY18 which ends on February 2, 2018.

This acquisition appears to be receiving a more positive message overall in the market than Cisco’s acquisition of Viptela earlier this year. In Viptela’s case, being acquired by Cisco immediately raised questions about where this product would fit alongside the existing IWAN and Meraki products, its impact on the ecosystem of IWAN-related products (Glue Networks, LiveAction, etc.) and what the final product mix would look like. Many of these questions have subsequently been answered, but the immediate reaction was one of uncertainty and concern.

In this case, there is less concern about stepping on the toes of an existing product. VMWare has a very credible foothold in data center networking in its NSX product line, but the VeloCloud product would extend its capabilities towards the WAN, and provide a router alternative at the edge of the network. The lack of overlap is important, and the timing appears to be good as more enterprises look at virtualizing the network edge.

A further step away from dedicated SD-WAN hardware at the edge

One of the big promises of SD-WAN has always been that many of the products can run on generic x86 hardware. However, almost all SD-WAN vendors have made hardware a key part of their solution as it’s easier for the enterprises that are already familiar with using appliances at the edge. This can be a huge challenge for enterprises operating in many geographies – the SD-WAN vendors have had to start from scratch on hardware certification, global distribution, break-fix maintenance agreements and all of the other challenges that come with new hardware. Established vendors like Cisco and Riverbed have a big head start in this area.

Moving to a software-based SD-WAN, running on generic hardware, starts to look much more compelling for these enterprises. Use a standard Dell or HPE server, procured through the regular in-country channels, and run the SD-WAN edge on this hardware instead. Need a firewall too? This can share the same hardware. Same for WAN Optimization, a small Windows server for AD DNS/DHCP, and other lightweight functions that may be required. Putting generic compute resource at the edge starts to look compelling if it can be used for such a long list of functions.

In fact, VeloCloud had an interesting announcement of its own in this area in the last week; its partnership with Telstra is based on the use of Juniper universal CPE hardware instead of VeloCloud’s own appliances.

This is obviously VMWare’s core business. It isn’t too much of a stretch to see how this technology could be used to extend some of the functionality in NSX out of the data centers and into smaller branch offices. Even before that happens, the VeloCloud solution appear to integrate fairly well with the VMWare ESX platform, and that will likely continue to improve.

What about Cisco?

It is clear that this move will be seen as a threat by Cisco. There is a real risk that more enterprises realize they don’t need a dedicated hardware router at the edge, and move to an orchestrated, software-based SD-WAN edge instead. Physical WAN interfaces were once a major obstacle for other vendors; the ubiquity of Ethernet has neutralized that. Comprehensive routing protocol support was the other; it is fascinating to see how quickly the major SD-WAN vendors (including VeloCloud) have matured in this department as they started picking up global carrier clients. It’s now realistic for an SD-WAN edge to replace the WAN router functionality in many use cases, including a complex eBGP configuration with MPLS vendors.

Cisco’s roadmap for the Viptela technology is starting to become clearer, and it’s looking like there will be some serious competition between these vendors in the next 12 months. However, Cisco’s proprietary hardware architecture at the edge (for now at least), and the price point attached to it, is going to be seen as a major obstacle for some enterprises.

Good news for the enterprise

Overall, I see this news as positive for the enterprise considering SD-WAN. There is a high likelihood of the VeloCloud product continuing to exist in something close to its current form, and the intense competitive environment between VMWare, Cisco and Riverbed will benefit the end user. 2018 promises to be another exciting year for SD-WAN!

Copyright © 2017 IDG Communications, Inc.

The 10 most powerful companies in enterprise networking 2022