We recently partnered with DellEMC to bring you a new IDC whitepaper focused on one of the most critical and relevant considerations for today’s data center leaders and operators. In How Network Disaggregation Facilitates Datacenter and IT Modernization, IDC discusses how digital transformation and the prioritization and modernization of applications are putting pressure on business infrastructure, specifically the network, to modernize and optimize for the digital era.
In brief, the IT world has changed rapidly around the network, demanding automation, scalability and agility. But the network has remained unchanged — monolithic, stagnant and inflexible. The implications for businesses now are not just technological support but economical as well. An inflexible network becomes expensive to scale at the speed of customer expectations and business innovation. Business innovation puts pressure on data centers to offer extensive automation of the entire network life cycle, from provisioning and deployment to day-to-day management and upgrades.
With IT transformation initiatives taking place everywhere to support application needs, the time is now to leverage those initiatives to reassess network architectures and operational models. Network disaggregation is a key architectural step forward to help organizations meet these challenges by offering unprecedented flexibility and agility to support end-to-end Continue reading
Behind each Cumulus customer is an awesome story about the future of open networking. These companies are forward thinkers who know that web-scale is the best solution for their network — enabling flexibility, reducing TCO, increasing efficiency — but it’s more than just our innovation. When we begin working with our customers, we get to see what innovations they’re driving, and recognize how we now fit into their story. We wanted to take a moment to highlight a few of our newer case studies that have been published and highlight the awesome work that these organizations have accomplished.
iNNOVO Cloud truly blew us away with one of the most energy efficient data centers in the world. They decided to run everything as an L3/IP fabric and not have large layer 2 domains (music to our ears). Partnering with OpenStack and Kubernetes, iNNOVO built an environment that brings efficiency and scale to their cloud and blockchain services. In fact, they reduced the time to configure switches by 83%. How’s that for increased efficiency? The best part is that their data center is all about sustainability and they’ve gone above and beyond your average infrastructure. Here’s a hint: think shipping Continue reading
Click here for our previous episode.
If you’ve been waiting for a tech-heavy deep dive, then you’re in luck. In this episode we’re getting REAL nerdy — so we decided to bring out the big guns and invite two of the industry’s biggest networking geeks to discuss hyper-converged infrastructure. Naturally, we got our fearless co-founder and CTO JR Rivers into the recording booth so he could share his wisdom (and crack a few jokes, as usual).
And who did we invite to go toe-to-toe with JR on networking knowledge? None other than the one and only Greg Ferro, co-founder of Packet Pushers! We couldn’t be more excited that Greg agreed to join us in the recording booth and share his industry insights.
So, what data center networking topic did we decide was meaty enough for these guys to chew on? Because Greg and JR are all about looking toward the future and analyzing what they see coming up on the horizon, this episode is dedicated to hyper-converged infrastructure (HCI). How is HCI changing the way we look at network architecture? We’ll discuss these topics Continue reading
Digital transformation means different things to different organizations. Perhaps it means shifting workloads to the cloud, digging deeper into data analytics, giving your employees more options for mobility or automating more processes. But one thing’s for sure: no matter what your digital transformation strategy looks like, your network is going to be either the hero or a hindrance to your digital transformation efforts.
According to Gartner, “the top networking challenge… is improving agility.” Not a big surprise there. But what is surprising is Gartner’s advice to “shift investments away from premium networking products toward [your] existing network personnel.” That’s right: the answer to improving your network’s speed and agility is not by buying expensive, proprietary monster switches and premium automation solutions. Rather, it’s by letting the people who best know your network decide the best way to make your network more agile.
Agile networks require a deep understanding of your organizational objectives. A “one size fits all” approach to networking just doesn’t work anymore. In order to prepare your network for digital transformation, it has to be both customized to fit your organization’s needs and be flexible enough to adapt when those needs change. Agility requires responsive, Continue reading
We are thrilled to announce that, with the Cumulus Linux 3.6.2 release, the Facebook Voyager packet-optical device is now generally available. That’s certainly an exciting development as the industry’s first packet-optical whitebox, but I actually wanted to discuss something else: how we were able to get support for the device up and running so quickly and what that means for Cumulus Linux. There’s a large number of devices that we support, which are listed on our Hardware Compatibility List. Those devices, however, are all “normal” Ethernet devices. Usually we can get those to market rapidly, but Voyager was more complicated.
First, as you may know, Facebook Voyager is a device with a Broadcom Tomahawk chip supporting 12 100 Gigabit Ethernet ports as well as an additional 4 200 Gigabit DWDM ports supported from 2 Acacia AC400 modules (for more details on the underlying specifics, see the Voyager Tech Docs). I mention the chipset explicitly as that’s one of the most critical – and time-consuming – components to support as we port to a new device. In this case, of course, we already supported the Tomahawk, so that was an immediate “leg up” on the work.
Those Continue reading
Time for another Cumulus content roundup! We’ve been really busy this summer, so there’s a little bit of everything in this post: videos, industry news articles, new podcast episodes and even an entire book! So if you’ve got room on your summer reading list, be sure to add EVPN in the Data Center. Or, if you’ve got too much to do and can’t find time to sit down and read, grab a pair of headphones and listen to the latest episode of Kernel of Truth while you work. The choice is yours!
Kernel of Truth episode 03 — Linux: the kernel, the community & beyond: You can’t name an open networking podcast “Kernel of Truth,” and NOT have an episode dedicated to the Linux kernel! Listen to our discussion about the Linux community and why Linux belongs in the data center.
EVPN in the Data Center: This eBook cuts through the fog and explains how you can deploy this technology seamlessly in your data center. You’ll discover why EVPN can be simpler to use in data centers than in service provider networks.
Click here for our previous episode.
On March 27, 2018, Cisco announced it was embracing disaggregation of the data center by allowing customers to run NX-OS on third-party switches and to use any network operating system on its Nexus switches. It’s certainly an interesting move, considering that they’re the company that claimed to have killed white-box networking.
…But does this model REALLY fit the definition of network disaggregation? What does true data center disaggregation look like? Why did Alanis Morissette name the song “Ironic” when none of the lyrics are examples of irony?? To answer these questions, I invited Ben Ritter (Consulting Engineer, Cumulus Networks) and Rama Darbha (Senior Consulting Engineer, who you’ll remember from our second episode — get ready for more #RamaRants!) into the recording booth so we can get to the bottom of this. In addition to breaking down the definition of data center disaggregation, Rama, Ben and I go full John Lennon and imagine a perfect world, where Cisco actually embraces the true spirit of disagreggation. How would this impact the industry? Imagine there’s no black box…it’s easy if Continue reading
We couldn’t be happier to report that Cumulus Networks has been placed on the Gartner Magic Quadrant for Data Center Networking in the Visionaries category — for the second year in a row. This acknowledgement and honor once again solidifies not only our place in the industry, but also open networking’s place as a whole. The report states, “Cumulus continues to pioneer a vision based on open components, disaggregation of switching hardware/software and automation.” We couldn’t agree more, and we’re excited to be driving the industry forward into efficiency, flexibility, simplicity and innovation.
Gartner’s 2018 Magic Quadrant for Data Center Networking gives an assessment of the current data center networking solutions for enterprises, and evaluates where they fall on ability to execute as well as completeness of vision. Without further ado, here is this year’s quadrant:
A few key highlights:
Major cloud providers are able to achieve impressive uptime by distributing the load across a large number of commodity servers. There’s no single master server that can fail and bring down the entire infrastructure. It’s not surprising then that so many healthcare networks are already relying on major cloud providers to run electronic medical records (EMR) and imaging applications and store patient data. So doesn’t it make sense to adopt the same approach for the network?
Healthcare networks face a unique challenge that many other networks don’t: how do you grow your network quickly and reliably while remaining compliant? As patient data grows, this tension between growth and compliance is increasing. Healthcare networks are feeling the pressure to move data faster, and this often requires adding more connections and switches, which entails routing and other configuration changes. These continual and rapid changes come at the cost of long, sometimes unplanned, outages. Planned outages are to be expected, but unexpected outages are a nightmare. Consequently, many healthcare networks have understandably opted for slower network growth to maintain a stable, reliable network.
Thanks to the limitations of traditional networks, network operators are accustomed to doing everything manually and slowly. But they want to perform configuration, troubleshooting and upgrades faster and with fewer mistakes. They’re ready and willing to learn a new approach, but they want to know what their options are. More importantly, they want to do it right. The good news is, regardless of your organization’s specific goals, you can operationalize Cumulus Linux to meet those objectives faster and more consistently. This post will help you understand your options for developing agile, speed-of-business workflows for:
And if you’re looking for a deeper, more technical dive into how to implement these network operations, download this white paper.
The biggest disadvantage of manual configurations is that they simply don’t scale. Implementing BGP across dozens of switches is a copy-and-paste endeavor that’s time-consuming and prone to error. Not only that, checking that the configuration took effect and works as expected requires hop-by-hop verification in addition to testing route propagation and IP connectivity. However, In a small network, there’s no shame in at least starting out doing everything by hand.
Cumulus Linux lets you use a Continue reading
Listen, you can’t name an open networking podcast “Kernel of Truth,” and NOT have an episode dedicated to the Linux kernel! So we got two of the brightest, most enthusiastic Linux experts we know into the recording booth and let them wax poetic about the language of the data center. As I soon found out, it’s harder to get Linux fans to STOP talking about Linux that it is to get them going — but hey, that just makes my job as host a lot easier! There’s nothing like listening to knowledgeable people discuss something they’re passionate about, and that’s what we’ve got in store for you.
In this episode, I’m joined by Roopa Prabhu, leader of the kernel team at Cumulus Networks, and Shrijeet Mukherjee, Cumulus’ former VP of Engineering. Specifically, our discussion revolves around the Linux kernel and Linux community. We get into some pretty interesting questions: why Linux in the data center? What has Cumulus contributed to the kernel? How has the prolific Linux community evolved? What the heck is a “boffin”?? I’m not a fan of spoilers, (thanks for ruining Avengers: Infinity War for me, Twitter!) so I’ll let you guys tune in and find Continue reading
Cumulus NetQ is on FIRE!!
Just one year ago, Cumulus launched a new product that fundamentally changes the way organizations validate and troubleshoot not just their network, but the entire Linux ecosystem as a whole. The product was named NetQ (think Network Query). It provides deep insight on the connectivity of all network devices either now or in the past — including all switches, Linux hosts, inside Linux hosts (Containers, direct interaction with container orchestration tools like Kubernetes, VMs, Openstack environment) and any other devices running a Linux-based operating system connected to the network. No more manual box-by-box troubleshooting, no more wondering what happened last night, no more pulling cables to find where the issue was stemming from, no more finger pointing, no more human-led misconfigurations and no more frustration of not having sight past the edge of the network.
Instead, Cumulus NetQ, the agent-based technology that runs on anything Linux, changes all that. NetQ brings the efficiencies of web-scale to network operations with an algorithmic, preventive, centralized telemetry system built for the modern automated cloud network. NetQ aggregates and maintains data from across all Linux nodes in the data center in a time-series database, making the fabric-wide events, Continue reading
When you think of your ideal campus network, the term “web-scale” may not immediately come to mind. After all, the term web-scale is something you’re more likely to associate with the cloud than with your network. But you might be surprised to learn that your ideal campus network fits the definition of a web-scale network to a T.
Fundamentally, a web-scale network functions as a single unit that can grow and change on demand, without requiring hands-on reconfiguration of multiple switches or mass hardware replacement. And because it functions as a single unit, a web-scale network can also give you full visibility into the health of your network, end-to-end.
The primary way web-scale networks achieve this flexibility and visibility is by decoupling or disaggregating the hardware and the network operating system (NOS) that runs on the hardware. Since the advent of specialized hardware networking devices, the operating system and hardware have been tightly coupled together. Proprietary NOSes often have platform-independent code that runs only on specialized hardware. Because of that, upgrading to a new software version often means buying new hardware. In some cases, that may be as simple as buying additional RAM to support the new version. In more Continue reading
“To boldly go where no one has gone before!”
Those words still echo in my mind as I remember watching the old Star Trek shows from yesteryear. It rings of adventure, of exploration, and of never settling for the known state of things.
It is these words that come to mind when I think of the new Voyager technology that is coming to market, which is designed to go boldly where no other technology in the Packet and Optical world has gone before. Voyager is the industry’s first combined routing, switching and optical platform all combined in a 1 RU footprint. The unique combo sets out to unifying both IP and optical to massively reduce complexity and costs. It will boldly transform the data center interconnect of today.
But it’s also the first open offering in the optical space. Cumulus is bringing its networking with S.O.U.L. (Simple. Open. Untethered. Linux) moxie and applying it to the transport and data center interconnect markets. This disaggregated solution dramatically reduces the cost of the current proprietary stack. It’s a solution with multiple players…
A few weeks ago, I set out to the beautiful city of Vancouver’s convention center, along with a boatload of rocket turtles and a stellar team. It was a great time with a wonderful scenic view of the ocean. I’ve been at Cumulus a few months now, but I can’t help but enjoy looking around, seeing the friends I’ve made in the industry, and the friends and companies Cumulus has worked with over the years. It is exciting to have thousands of people coming together at OpenStack Summit Vancouver to work on a shared goal.
This year, we were lucky enough to have our very own Pete Lumbis take the stage with David Iles of Mellanox to present our joint solution around the latest SDN revolution, which is centered on creating efficient virtualized data center networks using VXLAN & EVPN.
In the next few paragraphs, I’ll share some highlights of the event, some photos, and a recap of that exact discussion. There was a lot to learn and discover, and I’m excited to share the details.
On our first day, lots of things were going on — we Continue reading
Network monitoring, “Wonderwall” by Oasis, virtual test environments, Wu-Tang Clan (Cumulus Rules Everything Around Me!), validating configurations and cursed email chains. What do all of these things have in common? They’re all topics in Kernel of Truth’s second episode! Now, if you want to know HOW all of these seemingly random talking points fit together, you’ll have to listen for yourself, but the main focus of this discussion is Day 2 operations. Specifically, we get into important topics like:
Our guest panel consists of two networking ops experts from Cumulus Networks: Senior Consulting Engineer Rama Darbha (also known as “Tough Tiger Fist” according to the Wu-Tang name generator), who you’ll remember from our previous episode on network automation, and Technical Marketing Engineer Pete Lumbis (aka “Master Block Warrior”). These industry pros joined me (“Ungrateful Ambassador”) to provide first-hand experience and insight into why Day 2 operations deserve just as much attention as architectural design.
On another note, we’ve got some great news — Continue reading
No doubt about it: the prospect of adding another zero to the end of your top network speeds is exciting. And the reward of the immediately noticeable performance improvement never gets old. Speed makes a noticeable, and not just measurable, difference. And with the massive increase in the amount of data servers need to process, 100G is soon going to be a necessity for many organizations.
But increasing network speed is about more than pushing more bits across a wire. Faster networks enable you to squeeze more out of your physical rack space. You need fewer servers, fewer network connections, and – dare I say it – fewer switches. It’s true. A faster network lets you pack more computing into the same space.
Whether you plan to do a forklift upgrade to 100G or intend to replace one switch at a time, there are some key things you need to know to avoid getting locked into one switch vendor or losing backward compatibility with your existing equipment. In this post, I’m going to give you my top 5 tips for making transitioning to 100G networking a smooth one.
First, a little background. Continue reading
Virtual Routing and Forwarding (VRF) is a ubiquitous concept in networking, first introduced in the late 1990s as the control and data plane mechanism to provide traffic isolation at layer 3 over a shared network infrastructure. VRF for Linux is an excellent blog that describes the technology behind VRFs, especially as it pertains to the Linux kernel. With the introduction of support for leaking of routes, VRFs get to enjoy their isolation while also having the nous to mix and mingle.
You have a valid question there. That was certainly the initial use case for VRFs. Each VRF was intended to represent a customer of a service provider and isolation was a fundamental tenet. Each VRF had its own routing protocol sessions and IPv4 and IPv6 routing tables and route computation as well as packet forwarding was independent from other VRFs. All communication stayed within the VRF other than specific scenarios such as reaching the Internet. Hershey’s wouldn’t want to get too chatty with Lindt, right? No, VRFs weren’t meant to be gregarious.
As VRFs moved outside the realm of the service provider and started finding application elsewhere, such as in the Continue reading
It’s officially summer time, so we’re bringing you the HOTTEST new content from Cumulus Networks in this month’s content roundup! Whether you want to layer on the sunscreen and enjoy our content while basking in the sun, or stay safe and cool indoors with your laptop and AC, you’re bound to enjoy what we’ve got in store for you. We’ve got new videos and white papers, and even a brand new official Cumulus Networks podcast for you to check out!
Kernel of Truth Episode 01 – Networking Automation: “Kernel of Truth” is a Cumulus Networks podcast dedicated to bringing the best of open networking thought leadership straight to your ears. Listen to our very first episode where we discuss network automation and its impact on the industry!
5 Network automation tips and tricks for NetOps: In this white paper, we’ll give you five tips and tricks to get clarity around your automation decisions and reduce any friction that may be inhibiting (further) adoption of network automation. Check it out!
Joint solution overview: OpenStack and Cumulus Networks: By combining with Cumulus Linux, you can unify the entire stack on Linux, bringing together the OpenStack servers Continue reading
We often receive the following campus design question: “do you support switch stacking?” This is a fair question, as many of the legacy vendors have promoted stacking designs for the past decade. It’s popular enough that people ask for it, so we must support it, right?
Well, the popular option isn’t always the best one, and switch stacking designs are a very good example of that philosophy. So when people ask if we support stacking, we think to ourselves “heck, no” before politely telling them that we do not because better options exist.
“Perfection is attained, not when there is nothing more to add, but when there is nothing more to take away.”
At Cumulus Networks, we believe that simplicity is the corner-stone of network design.
Or, to say it another way, complex designs fail in complex ways (shoutout to Eric Pulvino for that quote!). Our former Chief Scientist, Dinesh Dutt, gave an excellent explanation around the importance of simple building blocks in his Tech Field Day 9 Presentation (6min 50 seconds in).
Let’s address a little history on switch stacking and then break down the major technical downfalls of a stacking design, the stacking protocol itself, Continue reading