In this report I look Cisco’s SD-WAN transition from IWAN to Viptela. I’ll highlight why IWAN has a ‘silent EOL’ and Viptela is the primary SD-WAN solution for Cisco’s Enterprise BU.
Table of Contents
- IWAN is put into ‘cold storage’ to protect existing customer investment and sustain in-flight deals to protect short-term revenue.
- Viptela is THE Cisco SD-WAN solution for the Enterprise BU.
- Viptela will run on ISR routers, eventually.
- There are plans for interim delivery, but customers should wait until the final solution is delivered.
- Viptela as a company and product is an excellent fit for Cisco. Integration should be smooth and relatively rapid in a Cisco frame of reference.
Cisco spent $660M to buy Viptela. 1 A deal of this size must produce a return and Cisco must move quickly as possible to monetize the acquisition.
The IWAN product family has earned a poor reputation over time for being difficult to deploy. Repeated reports on buggy IOS software (particularly PfR), the complexity of its design/setup, and serious issues with APIC-EM, and more, have been consistent.
Cisco invested in a few SD-WAN companies, which I speculate was to monitor the market 2. Realizing that SD-WAN was growing quickly and that Cisco IWAN was capturing limited market share, Cisco is rumored to have approached a number of SD-WAN companies in mid-2016 for preliminary discussions. Cisco announced the Viptela purchase on May 1 2017, and closed the deal three months later for $610MM. 3
The move happened with no rumors or leaks and surprised most analysts.
Key Factors In The Viptela Acquisition
In my view, the following factors are relevant:
- Cloud management and operations. A key part of the Viptela solution is the cloud-managed platform that aligns with Chuck Robbins’ oft-stated goal of making Cisco a cloud company. Even UCS servers now have a “cloud” strategy.4.
- Viptela is one of the few SD-WAN companies that operates as “router replacement” and is more closely based on existing “IP routing” models than other solutions. This makes it easier to sell and requires less retraining than cloud-centric SD-WAN products. Ease of sale is very important to Cisco’s business model, which relies on a reseller base of variable quality.
- Many Viptela executives have previously worked for Cisco in senior roles. This should smooth the Viptela integration into the difficult “Game of Thrones” business environment inside Cisco.
- Viptela needed to sell. Lots of rumors and scuttlebutt about Viptela’s internal situation lead me to conclude that the startup had incentives to sell after previously expecting to go all the way to IPO.
IWAN Is On Life Support
There are good reasons to sustain IWAN in the short term but terminate any ongoing or future development.
- Internal politics. Cisco is ruthless in culling staff, which leads to salary-at-risk situations. It’s common enough for people to speak up and defend their product/business units. This can lead to confusing messaging in public because Cisco’s marketing teams are distributed across all the business units, leading to conflicting information.
- Reputation damage. Many customers have spent large sums on rolling out IWAN, believing it to be a long-term strategy from Cisco. There is a negative impact on customer trust if IWAN is discontinued. Cisco’s reputation is being damaged by the Intel Atom clock issue, product failures, and repeated quality issues. Customers who have large, existing investments in Cisco ISR/IOS with 10 year strategic goals will be unhappy to see their assets killed off.
- IWAN is based on existing IOS technology that many customers will easily accept because it’s what they know. Customers who don’t like or want cloud-based management or GUI interfaces (and there are a lot of them) will see IWAN as the preferred solution.
- IWAN features are existing features and may continue to be developed. IWAN uses dozens of existing features in IOS software for an SD-WAN solution. These features will be maintained as part of IOS to support existing customers who are hand-operating their networks.
- Cisco must protect its cash flow to keep investors and share price. It is likely that there are many IWAN deals in the pipeline; losing those deals to competitors or seeing them go on hold has an outsized impact on a shrinking business.
Customers have substantively rejected the IWAN strategy. Conversations with customers suggest that many shifted to Meraki (incidentally creating a potential revenue loss), while others went to competitors. IWAN’s technology problems were causing reputation loss as Cisco struggled to make the technology stable. A key factor here is resellers backing away from IWAN because it wasn’t profitable to install, which could break the resellers’ businesses.
The Cisco announcement on completing the acquisition makes it clear that Viptela is the preferred solution (emphasis is mine) 1:
“Cisco will commit significant engineering resources to bring next-generation SD-WAN solutions to market. For customers and partners that require pure play enterprise grade SD-WAN, cloud networking and WAN network-as-a-service functionality, Cisco’s SD-WAN solution based on Viptela will be the preferred solution.“
Keeping IWAN on life support costs relatively little (important to the cost-focused Cisco of today). As a hodge-podge of existing software components inside of Cisco IOS ( GRE, DMVPN IPSec, NBAR2, PfR) many of them are sustained as standard features. For example, DMVPN IPSec is a core competent of IWAN but has been much more widely used as static public WAN overlay for more than a decade.
Overall, I am confident in stating that IWAN is on ‘life support’ for business reasons and will move into the EOS/EOL lifecycle over time.
The Viptela Rollout
I’ve heard a number of rumors from a wide range of sources about the integration of Viptela into the Cisco product family. I list those that I feel are ‘confirmed’ :
- Viptela will run on the existing ISR routers as a native operating system. Viptela may run in a container on ISR IOS in the short term to offer faster rollout, but this is not the long term plan.
- The current Viptela hardware is an Intel x86 based server with a handful of Ethernet ports. The ISR product family is also an x86 server with less CPU/RAM resources and non-standard architecture, and may struggle to handle the extra workload.
- The leadership at Cisco plans to lead with Viptela for all SD-WAN sales opportunities as soon as possible because of its cloud integration and operations. This is a key business goal for Cisco to increase subscription revenues and to be “a cloud company.” Cisco has stated that subscription revenues will increase total sales per customer.
- The current selling price of the ISR platform with IOS conflicts with the low cost of competitors and is likely causing difficulties with internal road maps. Discounting a Viptela/ISR solution likely causes revenue problems in other BUs. This is a big deal when the Enterprise BU is shrinking in revenue already.
- Training of resellers/channel on Viptela is major undertaking. Using the ACI launch as a guide, Cisco is likely to spend hundreds of millions of dollars in roadshows, training programs, collateral delivery, and personal outreach to create demand from their sales channel. Cisco has a well developed model and relies on resellers to take products to customers. However, reseller quality is variable, change-resistant due to lack of resources, and large numbers are struggling to be profitable.
- Migrating the Viptela cloud platform to a Cisco-supported technology platform is a major undertaking. The Enterprise IOS software teams have a track record of poor quality software in recent years. Internal upheavals such headcount reductions and realignments may cause delays, poor quality, and worse. Customers are advised to wait and not enter the market early.
- It may be linked with revenue loss to Meraki, where customers are quickly adapting to cloud-managed networks.
- Cisco staffers have strong reactions to seeing the IOS legacy in the Enterprise WAN being dumped. This process is taking substantial resources (meeting, presentations, budget planning, etc.) as the business units re-evaluate the impact to their plans.
The realization that IOS is no longer core to the future of Cisco Enterprise WAN products is a tough pill for many Cisco staffers to swallow. Some are struggling to accept this as their product strategies are modified, teams are redeployed, and budgets changed. It’s a #BigCompanyProblem to make a large change like this, and Viptela will have a substantial political fight to establish its future inside Cisco.
Embracing cloud-centric SD-WAN by buying Viptela is a decisive move that shows Cisco’s Enterprise BU is willing to make hard decisions. Of note, the signs suggest that Cisco may be willing to slaughter the IOS sacred cow on the altar of progress.
Cisco IWAN may not be dead 5 to keep customers happy, but I’m equally sure that all investment and planning for future products have either ceased or are being wound up.
Based on previous strategic transitions, we will know more when Cisco commences a major campaign to educate customers. SD-WAN is not easily understood and generating demand will be vital to generating a return on the $660M purchase price. At this point, a more aggressive sales effort will be made to switch IWAN opportunities to Viptela. Therefore, I advise customers to put any WAN upgrades on hold until the Viptela product launch.
Cisco is a slow moving, risk-averse company. Investors are pleased with its ability to regularly deliver cash dividends even as its total revenue has declined over the last few years. Executives will not risk stable, predictable cash flow for fear of impacting the share price and their own very generous options packages. Cisco has the resources and profits to sustain a dead-end strategy for years, as has been clearly demonstrated with the Catalyst 6500/6800 product line and more recently the Nexus 7000 family.
I remain committed to the view that SD-WAN using Public WAN/Internet will be the dominant enterprise WAN strategy. Further, that transition will be rapid. I acknowledge that the WAN has been very slow to change with over 30 years of stagnation with expensive technology. I doubt that a majority of WANs even run QoS a decade after the technology was mainstream in the IOS platform.
However, the financial incentive for enterprises to replace the private WAN with Internet connections, in part or entirely, is so compelling that it will drive a more rapid transition than previous technology changes.
- Cisco Completes Viptela Acquisition – Cisco – Retrieved 7 Sep, 2017 ↩
- VeloCloud Raises $35M to Meet Accelerating Global Demand for Cloud-Delivered SD-WAN – Retrieved 7 Sep, 2017 ↩
- Cisco grabs-up SD-WAN player Viptela for $610M | Network World – Retrieved 7 Sep, 2017 ↩
- Project Starship – Starting The Journey – Retrieved 7 Sep, 2017 ↩
- Cisco’s IWAN isn’t dead | Network World – Retrieved 7 Sep, 2017 ↩