NFD13: SolarWinds Presents Its New NetPath Tool

I don’t want to put words into Peter Welcher’s mouth. But in his post about SolarWinds’ NetPath tool, he straight up says, “[i]t turns out, I was very impressed with the new NetPath tool!” Not a lot of room for ambiguity there! SolarWinds presented at Networking Field Day this past November, and spent the entire session going over NetPath. Peter really enjoyed the presentation, not just for what NetPath could do, but also learning the journey SolarWinds took to refine and develop the tool.

See in the Fog with Ixia CloudLens

David Varnum draws a great analogy. In a lot of ways, flying through cloud and managing cloud infrastructure and applications are similar. Both don’t seem to bad to navigate from the outside, but once inside, you lose perspective. This requires both a pilot and systems engineer to have precise instrumentation to properly navigate where they want to go. With Ixia’s CloudLens, David sees someone finally providing that instrumentation.

Apstra’s Ethereal Network State

Apstra has a really interesting idea. What if you could design your network based on what you wanted to do with it? Instead of chaffing with the constraints of vendors and hardware, Apstra provides an abstraction layer to allow you to do this. Ethan Banks saw their presentation at Networking Field Day. He wrangled with the idea that Apstra is simply providing configuration management. But instead, they are providing a solution that takes a look at the network as a whole, something network engineers rarely can do. Instead, you state what you want to do with the network, and the Apstra Operating System tells you how to make that happens. It then knows the intent of the network, and is able to heal and remediate to keep that intent in place. Ethan thinks they’re entering into a crowded market, but that kind of model could really allow them to stand out.

NFD13: Forward Networks Comes Out of Stealth to Impress

Peter Welcher reviews what he saw from Forward Networks once they came out of stealth. Peter seemed impressed with how Forward is able to put together a database model of network configurations that can be searched and indexed independent of actually operating on the network. This can then be used for fast troubleshooting, as well as testing configurations to make sure they’re operating within a desired state. Overall, Peter sees this as a way to not spread you senior enterprise talent too thin. While there isn’t any automated remediation backed in, the powerful forecasting tools they present make it possible to better utilize talent in your organization.

Generating Maps of Your Traffic

For a network engineer, it sometimes feels impossible to avoid traceroute. Tim Miller thinks it can be a valuable tool to see where traffic is getting dropped, but it’s not without its issues. He’s highlighted some other solutions in previous posts, but the one he’s looking at today is SolarWinds. Their NetPath tool has gone from a lab toy to an official feature of their Network Performance Monitor solution in a little under a year. Tim finds it a really impressive tool. Even though it requires Windows-based polling appliances in a network, a Linux guy like Tim can still be tempted. It gives historical information layer on top of what you would find with a traceroute, and adds multipathing. Overall, Tim sees this as a very practical tool to help disentangle issues in increasingly complex networks.

On Network Blindness

Notable beard accomplisher and Apstra systems engineer Derick Winkworth shares some thoughts about network blindness. He compares it to face blindness, where people can see all the individual components, but cannot recognize them together. In much the same way, many network engineers build automation into their systems, without recognizing what the purpose of that network is, suffer from that same kind of affliction. Apstra developed with Apstra Operating System in response to this. It provides an abstraction over hardware to create networks with intentionality. The benefit of this is that it allows you to built a network around what you want to do, not change your behaviors based on the network. Derick gets into the nuts in bolts in the piece, but conceptually, its a refreshing take.

If you enjoy the piece, make sure to check out all of Apstra’s videos from Networking Field Day.

Ixia Works Out Its Network Trust Issues

Rich Stroffolino looks at what Ixia presented at Networking Field Day last month. Their product portfolio is pretty packed, but focused around network visibility. The presentation have Rich a new appreciation for the problem. Ixia has a comprehensive system of network probes and packet brokers to ensure zero-packet loss for monitoring solutions. Overall it’s an impressive offering.

Capture, Filter, See – Ixia Vision ONE

Ixia is a company some may not associate with network packet brokers, but that changed when they acquired Anue Systems in 2012. Ethan Banks wrote up his impressions on this based on what he saw at Networking Field Day in November. He seemed particularly impressed by their Ixia Vision ONE visibility tool. Sure it has all the features you could want, but for Ethan the most important part was that it was easy to get working right away. With the increasing complexity of networks, raw capability simply isn’t enough. Ixia differentiates itself with it’s ease of use here.

Enterprise Focused SD-WAN with Viptela

Rich Stroffolino looks at what Viptela presented at Networking Field Day last month. Overall, he found there approach interesting. Instead of being service provider focused for SD-WAN, Viptela designed their solution specifically for the enterprise. This allows them to address a lot of business needs directly. One of these is for multi-tenant locations, where SD-WAN can separate traffic without having to install a whole separate infrastructure. Overall, the approach opened up Rich’s ideas of what SD-WAN can do.

Trust But Verify: Lossless End-To-End Visibility from Ixia

Phil Gervasi looked at Ixia’s “Trust But Verify” approach to network monitoring. It’s an interesting approach, most other solutions simply assume that network traffic is being received by monitoring tools. Ixia goes beyond this. Instead of relying on SPAN ports, which drop traffic when a switch is overloaded, they use a series of packet brokers and network taps to make sure lossless data is being received by your monitoring solution. That’s right, Ixia proposes to not lose a single packet in doing this. That’s a tall order, check out Phil’s piece to see how Ixia is pulling it off.

Forward Thinkers, Forward Networks

Rob Coote shares his impressions from what he saw with Forward Networks at Networking Field Day last month. He really highlights how Forward’s software modeling of network performance could impact, not just the performance of the network, but the worth routines of network engineers. By effectively giving you a network lab to tinker with in software, their solution theoretically eliminates the “wait-and-see” approach to changes in a network. Rob really hopes they are able to move their solution beyond just monitoring to remediation. But he makes a really great point on the very human impact Forward Networks could have.

The Quest for Verification with Forward Networks

Forward Networks gave a presentation fresh out of stealth mode at Networking Field Day, and it certainly made an impression with Rich Stroffolino. He outlines how the company is doing their network monitoring. They model all possible places a packet can go on a network in a constantly updating software model. This allows you to not only react when problems occur, but also for better planning and provisioning, since you can model traffic very accurately in the software model. We’ll wait to see how their solution gets deployed in an actual enterprise, but on a theoretical level it’s fascinating.

Forward Networks – go ahead, break it

Amy Arnold laments the plight of the network engineer. The agonize over network design, try to come up with every conceivable failover scenario, and then deal with the consequences. Some have the aid of a lab to help test their configuration, most don’t. That’s why what Forward Networks presented at Networking Field Day was so interesting. It allows for you to model over your network in software, and then break it in every conceivable way. Forward’s model shows every a packet can possibly go, allowing the engineer to see exactly how a scenario will play out. She was justifiably concerned about how their product will be priced going forward, but otherwise it seems like a valuable tool in the engineer’s arsenal.

Shiny new NetPath Services

SolarWinds showed off their latest with NetPath at Networking Field Day. Amy Arnold seemed impressed by their solution. NetPath isn’t just a traceroute visualization tool, it uses “real” network traffic from Windows-based pollers on the network. This allows an engineer to get a better sense of how traffic flows, without worrying about packets being dropped (as much) by devices on the network. Amy says it best, “any tool that expands insight into what packets are doing is a beautiful thing.”

State of the Industry: Network Analytics

Gestalt IT just debuted a new feature, a weekly State of the Industry post. For their first week, they’re looking at the state of network analytics. They take a look at two competing methodologies to the problem. The first is SolarWinds NetPath tool, which sits in the network. The SolarWinds approach seems to take the ideas behind existing tools, and looks to perfect them. The other method is Forward Networks, which is presenting a top-down approach to do live mapping purely in software. Both are interesting, and point to further investment and development in the space going into 2017.

Why design simplicity is bad for your network

Design simplicity sounds appealing. After all, it would be easier to understand, manage, and theoretically expand. But Kevin Myers wrote a piece on why this can ultimately be a failing. He was having a discussion at Network Field Day about the differences in an LTE network versus an enterprise LAN. LTE just seems to work, even though it’s serving a vary large user base. Kevin notes that this is because enterprise networks aren’t often designed by engineers with their intended purpose in mind, rather a vendor supplies the network and the engineer is in charge of implementing within that given design. These are often instructed to be simple, but as businesses merge and needs change, the network designed to be simple is often unable to scale easily to a new complex environment. It’s an interesting read that touches on why a lot of enterprise technology decisions have more to do with culture than anything else.

OpenFlow Is Dead. Long Live OpenFlow.

Tom Hollingsworth takes a look at the curious life of OpenFlow. This once hyped panacea has found a completely new life from its original purpose of replacing the forwarding plane programming method of switches. Tom compares it to the development of Viagra as originally being intended for high blood pressure. Company’s like NEC have taken OpenFlow, with their ProgrammableFlow derivative, and adapted it to a whole new set of purposes, in this case mitigating the spread of infections within networks. It’s always interesting to see an established tool reimagined with a new purpose.

Coming to SD-WAN: The Build vs. Buy Decision

Presenting at Networking Field Day earlier this month, VeloCloud is offering a rather unique solution for SD-WAN. Instead of presenting themselves as a solution that an enterprise would build and deploy internally, VeloCloud takes a different approach. They’ve partnered with a number of Tier 1 and 2 Service Providers, integrating their service within their offerings, instead of using their offering as leverage for lower rates for customers.

Bob McCouch has an writeup about the pros and cons of this approach, as well as some thoughts on some of VeloCloud’s particular innovations. It’s a really thoughtful look at the tradeoffs implicit in this setup.

SolarWinds NPM 12 NetPath

Jody Lemoine got a look at SolarWinds’ NetPath product at Networking Field Day this month. For a product in its first official release, four months out of the lab, Jody thought it was a well implemented solution. He particularly like how NetPath moved beyond the confines of the enterprise network, into what’s happening with carriers and the destination networks. If you too seek to know the truth about your network, check out the rest of Jody’s piece.

Apstra intends greatness beyond Sparta

Apstra has a really interesting pitch. What if instead of building your network around how each vendor’s node and appliance talked to one another and what capabilities it had, you could design the network the way you wanted it to work first, and use an abstraction layer to make sure all the individual pieces played nice with one another? That’s what Apstra is proposing with their Apstra Operating System (AOS). David Varnum gave it a look at Tech Field Day, and shares his enthusiasm for their intent-driven approach.