SDxCentral’s own Matt Palmer took some time to chat with Brocade’s Kelly Herrell, senior VP and general manager of software networking, to find out what the acquisition of SteelApp has brought to the company, as well as what the future holds for merging SteelApp’s technology into Brocade’s NFV product line.
Herrell: The Brocade ADX has a large hardware installed base that is growing through build-outs, primarily within some long-time, highly-valued customers. We service and support them as those build-outs continue.
Last year Brocade initiated our surge to capture the opportunity in the New IP – architectural changes required within IP networks as a result of megatrends such as cloud, mobile, social, and big data. For ADCs, a big part of this change is the shift of system architectures away from proprietary hardware, and to open virtual software on standard servers. This is an aggressive trend; virtual ADCs are capturing 100 percent of the market growth, and are even beginning to cannibalize hardware ADCs. It was a powerful strategic opportunity for Brocade as industry disruptor.
That’s why last year we extended Brocade’s ADX product line to include vADX, a virtual ADC with an Intel-optimized data plane. The new price and performance advantages of the vADX were staggering – currently running 80 Gb/s on only eight cores. Customers recognized the benefits quickly, especially when compared to legacy hardware vendor price points.
While that activity was underway, the opportunity to acquire SteelApp suddenly presented itself. SteelApp had unique bragging rights: Top-three market share for vADCs, and also the only vADC included in Gartner’s Magic Quadrant for ADCs – in fact, labeled a “visionary.”
We quickly realized how SteelApp could turbocharge our vADC strategy. From a product and technology standpoint, its sweet spot is a strong focus on the application. This Layer 7-intensive capability earned it an established position in enterprises and cloud providers. So when we looked at this we said, “The best possible scenario would be to acquire SteelApp’s Layer 7 application centricity and the associated route to market, and then double engineering resources on it instantly by applying our vADX team’s skills.” We’ll extend the new offering by including the vADX feature sets and incredible data plane scalability.
The result is a best in class vADC roadmap where one plus one equals three. We’re very happy with that conclusion. The combined team is hard at work and customers love the direction.
A question that always comes up when a company consolidates products within a market segment is what is the combined product roadmap? What does this mean for ADX, vADX, and SteelApp customers? What should customers of all three products do?
Herrell: For the physical ADX there is no change; customers can continue to purchase it and we’ll service and support them as we always have, same for customers using virtual ADX. For new customers, we present the SteelApp vADC and its expanded roadmap. That is Brocade’s vADC solution, and it will be a superset of the SteelApp and vADX functionality. As such it will enable clear transition for all customers – physical or virtual ADX, and SteelApp alike.
If I’m a developer, what makes me say, “I need to look at SteelApp?” What are some of its use cases?
Herrell: The power-trend driving vADC adoption is the evolution of data centers to cloud architectures. This puts geography questions on app developers: Where are they going to deploy apps? Is it going to be in private cloud, public cloud, or a hybrid of the two? This is a massive shift compared to “old IP” where historically developers deployed large, monolithic apps only inside their own walls, and managed their delivery through large hardware ADCs.
Given this shift to cloud architectures, the more tools developers have to leverage, the more degrees of freedom they have to pursue their design objectives. They shouldn’t be constrained in where they place apps, or have to approach their design using old waterfall development methods. They need agile and open options – freedom of choice and elasticity of infrastructure.
Network infrastructure changes to adapt to new application environments. Cloud and virtualization are at the root of the application changes, and the DevOps model is accelerating that pace of change dramatically. Brocade’s unique vADC offering based on SteelApp is the best suited to this new environment.
Just consider the issue of app placement. SteelApp is already in numerous clouds – Amazon, Rackspace, Azure, Joyent – and more are coming. The exact same product can be deployed in a private data center. If your app is portable, so must be your vADC. And today’s apps need to be elastic; SteelApp offers pooled bandwidth service that can be expanded and contracted on demand. With SteelApp’s Web Application Firewall, apps can be managed with even better granularity and security.
The world of ADCs is morphing to adapt itself to the changing app environment very much, and Brocade’s vADC roadmap based on SteelApp is the superior choice.
“Controller” is a popular word used to define products today, leading to customer confusion. Brocade has its Vyatta OpenDaylight Controller and SteelApp has the SteelApp Service Controller. To help end-users better understand Brocade’s software suite, what does each do? When should a customer use one over another? Do you expect these “Controllers” to be integrated over time?
Herrell: The word “Controller” is so used, and misused, and reused. It’s not out of bad intentions, it’s just going through a rapid evolution and definitions are expanding.
Brocade is making this much simpler to understand. Brocade has a single SDN Controller, based on OpenDaylight. Below that, IP products have their own local control planes as always, and there is a range of automation and provisioning mechanisms.
For Brocade vADC, we refer to a function called the Service Director (which was formerly named “services controller”). It’s not a network controller; it’s a provisioning and metering mechanism for our pooled bandwidth offering. This enables on-demand spin-up and wind-down of SteelApp micro-ADCs, creating new value and eliminating the provisioning complexity in a data center or a cloud environment.
This is one of the things that makes Brocade’s vADC product roadmap so differentiated and compelling. Service Director allows you to use a few or a lot vADCs to achieve your result, so it’s really an innovative usage-based model.
By contrast, Brocade’s SDN Controller based on OpenDaylight is doing a much higher abstraction of services and capabilities, complimentary to all our IP and automation offerings. Network infrastructure is too complex and fragmented; it’s almost impossible to get the network to behave according to business requirements. However, those requirements can be modeled as business logic in an application that runs on top of Brocade SDN Controller, which then issues commands to the various network elements. This automation and programmability can be applied to all parts of the network – switches, routers, firewalls, VPNs, ADCs – whether they’re physical or virtual.
What do you see as the relationship going forward between the Vyatta platform and SteelApp?
Herrell: Vyatta has expanded dramatically since its early days as the pioneer in virtual routing. In the “old IP,” network products were identified by function – a router, a firewall, an ADC. In the New IP this distinction starts to dissolve as functions are coming together.
Brocade’s software strategy includes a unified service offering from Layer 2 to 7. Today we offer that capability in virtual functions – in fact we’re the only vendor in the world with this portfolio, and it can be deployed today wherever you want it, whether that’s a private data center, public cloud, or a hybrid. That’s really interesting for a variety of customers so that they can single-source their virtual network functions.
Longer-term, customers will require multi-function network services for virtual and cloud environments. There will be commonality at the data plane level, optimized for the hottest hardware. This approach allows network services to be ubiquitous across the server infrastructure. Being able to have those services that we can invoke for the customer is a very appealing proposition for the emerging software-defined data center environments.
What do you see as Brocade’s ambitions in security, over the next three to six quarters?
Herrell: Security is a very specialized and very niche area, and you see that in the scope and the size of the companies and the narrowness of what they’re pursuing. Security is certainly not one thing – it’s a thousand different things, and it’s constantly changing. We think that there is a security awareness that our products have to have and, in some cases, enabling capabilities that work better with security products. With that said, we are not a security company. We don’t plan to become a security company. But we will have a range of APIs that enable best-in-class security to be designed into network architectures.
It used to be a lot more delineated between the network infrastructure silo and the security silo. The software-defined data center is changing that dramatically, and it’s blurring those lines very much. And so we’re making sure that we have good, open, exposed interfaces into that blurry area so that the best-of-breed security partners can work with us and know that we’re not trying to use old-school methods. You’ll see some interesting innovations from our software roadmap that will allow for our virtual network services to be the best partner for security companies out there, because of that openness.
Brocade is also talking a lot about New IP. Can you provide our audience with an overview of what is New IP and how SteelApp fits into it?
Herrell: New IP is really about approaching IP networking from the perspective of the new environment – one defined by the massive growth of cloud, social, mobile, and big data. This changing environment brings fundamentally new demands on the network. It’s a brand new way to conceive of network designs, and creates the window for a new breed of IP solution designs based on openness and virtualization. This is why for the first time software plays a significant role in network infrastructure.
It’s an application-centric view of infrastructure, and that’s why vADC is such an important component for customers. SteelApp has an undeniable history of being application-centric so we’re proud to leverage that heritage as the foundation of Brocade’s broader vADC strategy.
Brocade is not going to dictate what hypervisor is selected. We’re not going to dictate what hardware is selected. We’re not even going to dictate what geography you’re going to deploy this in, or whose data center or which cloud provider you’re going to deploy it in. The New IP very much has that freedom attribute to it. Every day when we wake up, we’re thinking, “How do we advance our software capabilities to enable the New IP?” Clearly, one of our latest actions has been to embrace the vADC opportunity and we’re proud to build upon the SteelApp foundation in that pursuit.
If someone is interested in SteelApp how can they get started?
Herrell: You can visit the following links to request trials for various SteepApp services: Brocade SteelApp Traffic Manager, Brocade SteelApp Services Controller, Brocade SteelApp Web Application Firewall.