Tag Archives: Applications

The Network Does Too Much

Posted on by
4

I’m at Networking Field Day this week and it’s good to be back in person around other brilliant engineers and companies. One of the other fun things that happens at Networking Field Day is that I get to chat with folks that help me think about things in new ways and come up with awesome ideas for networking blog posts.

One of the ones that was discussed quickly this week really got me thinking again about fragility and complexity. Thanks to Carl Fugate for reminding me about it. Essentially, networks are inherently unstable because they are doing far too much heavy lifting.

Swiss Army Design

Have you heard about the AxeSaw Reddit? It’s a page dedicated to finding silly tools that attempt to combine too many things together into one package that make the overall tool less useful. Like making a combination shovel and axe that isn’t easy to operate because you have to hold on to the shovel scoop as the handle for the axe and so on. It’s a goofy take on a trend of trying to make things too compact at the sake of usability.

Networking has this issue as well. I’ve talked about it before here but nothing has really changed in the intervening time since that post five years ago. The developers that build applications and systems still rely on the network to help solve challenges that shouldn’t be solved in the network. Things like first hop reachability protocols (FHRP) are perfect examples. Back in the dark ages, systems didn’t know how to handle what happened when a gateway disappeared. They knew they needed to find a new one eventually when the MAC address age timed out. However, for applications that couldn’t wait there needed to be a way to pick up the packets and keep them from timing out.

Great idea in theory, right? But what if the application knew how to handle that? Things like Cisco CallManager have been able to designate backup servers for years. Applications built for the cloud know how to fail over properly and work correctly when a peer disappears or a route to a resource fails. What happened? Why did we suddenly move from a model where you have to find a way to plan for failure with stupid switching tricks to a world where software just works as long as Amazon is online?

The answer is that we removed the ability for those stupid tricks to work without paying a high cost. You want to use Layer 2 tricks to fake NHRP? If it’s even available in the cloud you’re going to be paying a fortune for it. AWS wants you to use their tools that they optimize for. If you want to do things the way you’ve always done you can but you need to pay for that privileges.

With cost being the primary driver for all things, increased costs for stupid switching tricks have now given way to better software development. Instead of paying thousands of dollars a month for a layer 2 connection to run something like HSRP you can instead just make the application start searching for a new server when the old one goes away. You can write it to use DNS instead of IP so you can load balance or load share. You can do many, many exciting and wonderful things to provide a better experience that you wouldn’t have even considered before because you just relied on the networking team to keep you from shooting yourself in the foot.

Network Complexity Crunch

If the cloud forces people to use their solutions for reliability and such, that means the network is going to go away, right? It’s the apocalypse for engineer jobs. We’re all going to get replaced by a DevOps script. And the hyperbole continues on.

Reality is that networking engineers will still be as needed as highway engineers are even though cars are a hundred times safer now than in the middle of the 20th century. Just because you’ve accommodated for something that you used to be forced to do doesn’t mean you don’t need to build the pathway. You still need roads and networks to connect things that want to communicate.

What it means for the engineering team is an increased focus on providing optimal reliable communications. If we remove the need to deal with crazy ARP tricks and things like that we can focus on optimizing routing to provide multiple paths and ensure systems have better communications. We could even do crazy things like remove our reliability of legacy IP because the applications will survive a transition when they aren’t beholden to IP address or ARP to prevent failure.

Networking will become a utility. Like electricity or natural gas it won’t be visible unless it’s missing. Likewise, you don’t worry about the utility company solving issues about delivery to your home or business. You don’t ask them to provide backup pipelines or creative hacks to make it work. You are handed a pipeline that has bandwidth and the service is delivered. You don’t feel like the utility companies are outdated or useless because you’re getting what you pay for. And you don’t have to call them every time the heater doesn’t work or you flip a breaker. Because that infrastructure is on your side instead of theirs.

Tom’s Take

I’m ready for a brave new world where the network is featureless and boring. It’s an open highway with no airbags along the shoulder to prevent you from flying off the road. No drones designed to automatically pick you up and put you on the correct path to your destination if you missed your exit. The network is the pathway, not the system that provides all the connection. You need to rely on your systems and your applications to do the heavy lifting. Because we’ve finally found a solution that doesn’t allow the networking team to save the day we can absolutely build a world where the clients are responsible for their own behavior. The network needs to do what it was designed to do and no more. That’s how you solve complexity and fragility. Less features means less to break.

The Long And Winding Network Road

Posted on by
Reply

How do you see your network? Odds are good it looks like a big collection of devices and protocols that you use to connect everything. It doesn’t matter what those devices are. They’re just another source of packets that you have to deal with. Sometimes those devices are more needy than others. Maybe it’s a phone server that needs QoS. Or a storage device that needs a dedicated transport to guarantee that nothing is lost.

But what does the network look like to those developers?

Work Is A Highway

When is the last time you thought about how the network looks to people? Here’s a thought exercise for you:

Think about a highway. Think about all the engineering that goes into building a highway. How many companies are involved in building it. How many resources are required. Now, think of that every time you want to go to the store.

It’s a bit overwhelming. There are dozens, if not hundreds, of companies that are dedicated to building highways and other surface streets. Perhaps they are architects or construction crews or even just maintenance workers. But all of them have a function. All for the sake of letting us drive on roads to get places. To us, the road isn’t the primary thing. It’s just a way to go somewhere that we want to be. In fact, the only time we really notice the road is when it is in disrepair or under construction. We only see the road when it impacts our ability to do the things it enables.

Now, think about the network. Networking professionals spend their entire careers building bigger, faster networks. We take weeks to decide how best to handle routing decisions or agonize over which routing protocols are necessary to make things work the way we want. We make it our mission to build something that will stand the test of time and be the Eighth Wonder of the World. At least until it’s time to refresh it again for slightly faster hardware.

The difference between these two examples is the way that the creators approach their creation. Highway workers may be proud of their creation but they don’t spend hours each day extolling the virtues of the asphalt they used or the creative way they routed a particular curve. They don’t demand respect from drivers every time someone jumps on the highway to go to the department store.

Networking people have a visibility problem. They’re too close to their creation to have the vision to understand that it’s just another road to developers. Developers spend all their time worrying about memory allocation and UI design. They don’t care if the network actually works at 10GbE or 100 GbE. They want a service that transports packets back and forth to their destination.

The Old New Network

We’ve had discussion in the last few years about everything under the sun that is designed to make networking easier. VXLAN, NFV, Service Mesh, Analytics, ZTP, and on and on. But these things don’t make networking easier for users. They make networking easier for networking professionals. All of these constructs are really just designed to help us do our jobs a little faster and make things work just a little bit better than they did before.

Imagine all the work that goes into electrical power generation. Imagine the level of automation and monitoring necessary to make sure that power gets from the generation point to your house. It’s impressive. And yet, you don’t know anything about it. It’s all hidden away somewhere unimpressive. You don’t need to describe the operation of a transformer to be able to plug in a toaster. And no matter how much that technology changes it doesn’t impact your life until the power goes out.

Networking needs to be a utility. It needs to move away from the old methods of worrying about how we create VLANs and routing protocols and instead needs to focus on disappearing just like the power grid. We should be proud of what we build. But we shouldn’t make our pride the focus of hubris about what we do. Networking professionals are like highway workers or electrical company employees. We work hard behind the scenes to provide transport for services. The cloud has changed the way we look at the destination for those services. And it’s high time we examine our role in things as well.

Tom’s Take

Telco workers. COBOL programmers. Data entry specialists. All of these people used to be the kings and queens of their field. They were the people with the respect of hundreds. They were the gatekeepers because their technology and job roles were critical. Until they weren’t any more. Networking is getting there quickly. We’ve been so focused on making our job easy to do that we’ve missed the point. We need to be invisible. Just like a well built road or a functioning electrical grid. We are not the goal of infrastructure. We’re just a part of it. And the sooner we realize that and get out of our own way, we’re going to find that the world is a much better place for everyone involved in IT, from developers to users.

Cloud Apps And Pathways

Posted on by
1

jam

Applications are king. Forget all the things you do to ensure proper routing in your data center. Forget the tweaks for OSPF sub-second failover or BGP optimal path selection. None of it matters to your users. If their login to Seibel or Salesforce or Netflix is slow today, you’ve failed. They are very vocal when it comes to telling you how much the network sucks today. How do we fix this?

Pathways Aren’t Perfect

The first problem is the cloud focus of applications. Once our packets leave our border routers it’s a giant game of chance as to how things are going to work next. The routing protocol games that govern the Internet are tried and true and straight out of RFC 1771(Yes, RFC 4271 supersedes it). BGP is a great tool with general purpose abilities. It’s becoming the choice for web scale applications like LinkedIn and Facebook. But it’s problematic for Internet routing. It scales well but doesn’t have the ability to make rapid decisions.

The stability of BGP is also the reason why it doesn’t react well to changes. In the old days, links could go up and down quickly. BGP was designed to avoid issues with link flaps. But today’s links are less likely to flap and more likely to need traffic moved around because of congestion or other factors. The pace that applications need to move traffic flows means that they tend to fight BGP instead of being relieved that it’s not slinging their traffic across different links.

BGP can be a good suggestion of path variables. That’s how Facebook uses it for global routing. But decisions need to be made on top of BGP much faster. That’s why cloud providers don’t rely on it beyond basic connectivity. Things like load balancers and other devices make up for this as best they can, but they are also points of failure in the network and have scalability limitations. So what can we do? How can we build something that can figure out how to make applications run better without the need to replace the entire routing infrastructure of the Internet?

GPS For Routing

One of the things that has some potential for fixing inefficiency with BGP and other basic routing protocols was highlighted during Networking Field Day 12 during the presentation from Teridion. They have a method for creating more efficiency between endpoints thanks to their agents. Founder Elad Rave explains more here:

I like the idea of getting “traffic conditions” from endpoints to avoid congestion. For users of cloud applications, those conditions are largely unknown. Even multipath routing confuses tried-and-true troubleshooting like traceroute. What needs to happen is a way to collect the data for congestion and other inputs and make faster decisions that aren’t beholden to the underlying routing structure.

Overlay networking has tried to do this for a while now. Build something that can take more than basic input and make decisions on that data. But overlays have issues with scaling, especially past the boundary of the enterprise network. Teridion has potential to help influence routing decisions in networks outside your control. Sadly, even the fastest enterprise network in the world is only as fast as an overloaded link between two level 3 interconnects on the way to a cloud application.

Teridion has the right idea here. Alternate pathways need to be identified and utilized. But that data needs to be evaluated and updated regularly. Much like the issues with Waze dumping traffic into residential neighborhoods when major arteries get congested, traffic monitors could cause overloads on alternate links if shifts happen unexpectedly.

The other reason why I like Teridion is because they are doing things without hardware boxes or the need to install software anywhere but the end host. Anyone working with cloud-based applications knows that the provider is very unlikely to provide anything outside of their standard offerings for you. And even if they manage, there is going to be a huge price tag. More often than not, that feature request will become a selling point for a new service in time that may be of marginal benefit until everyone starts using it. Then application performance goes down again. Since Teridion is optimizing communications between hosts it’s a win for everyone.

Tom’s Take

I think Teridion is on to something here. Crowdsourcing is the best way to gather information about traffic. Giving packets a better destination with shorter travel times means better application performance. Better performance means happier users. Happier users means more time spent solving other problems that have symptoms that aren’t “It’s slow” or “Your network sucks”. And that makes everyone happier. Even grumpy old network engineers.

Disclaimer

Teridion was a presenter during Networking Field Day 12 in San Francisco, CA. As a participant in Networking Field Day 12, my travel and lodging expenses were covered by Tech Field Day for the duration of the event. Teridion did not ask for nor where they promised any kind of consideration in the writing of this post. My conclusions here represent my thoughts and opinions about them and are mine and mine alone.

 

The Complexity Conundrum

Posted on by
3

NailPuzzle

Complexity is the enemy of understanding. Think about how much time you spend in your day trying to simplify things. Complexity is the reason why things like Reddit’s Explain Like I’m Five exist. We strive in our daily lives to find ways to simplify the way things are done. Well, except in networking.

Building On Shifting Sands

Networking hasn’t always been a super complex thing. Back when bridges tied together two sections of Ethernet, networking was fairly simple. We’ve spent years trying to make the network do bigger and better things faster with less input. Routing protocols have become more complicated. Network topologies grow and become harder to understand. Protocols do magical things with very little documentation beyond “Pure Freaking Magic”.

Part of this comes from applications. I’ve made my feelings on application development clear. Ivan Pepelnjak had some great comments on this post as well from Steve Chalmers and Derick Winkworth (@CloudToad). I especially like this one:

Derick is right. The application developers have forced us to make networking do more and more faster with less requirement for humans to do the work to meet crazy continuous improvement and continuous development goalposts. Networking, when built properly, is a static object like the electrical grid or a plumbing system. Application developers want it to move and change and breathe with their needs when they need to spin up 10,000 containers for three minutes to run a test or increase bandwidth 100x to support a rollout of a video streaming app or a sticker-based IM program designed to run during a sports championship.

We’ve risen to meet this challenge with what we’ve had to work with. In part, it’s because we don’t like being the scapegoat for every problem in the data center. We tire of sitting next to the storage admins and complaining about the breakneck pace of IT changes. We have embraced software enhancements and tried to find ways to automate, orchestrate, and accelerate. Which is great in theory. But in reality, we’re just covering over the problem.

Abstract Complexity

The solution to our software networking issues seems simple on the surface. Want to automate? Add a layer to abstract away the complexity. Want to build an orchestration system on top of that? Easy to do with another layer of abstraction to tie automation systems together. Want to make it all go faster? Abstract away!

“All problems in computer science can be solved with another layer of indirection.”

This is a quote from Butler Lampson often attributed to David Wheeler. It’s absolutely true. Developers, engineers, and systems builders keep adding layers of abstraction and indirection on top of complex system and proclaiming that everything is now easier because it looks simple. But what happens why the abstraction breaks down?

Automobiles are perfect example of this. Not too many years ago, automobiles were relatively simple things. Sure, internal combustion engines aren’t toys. But most mechanics could disassemble the engine and fix most issues with a wrench and some knowledge. Today’s cars have computers, diagnostics systems, and require lots of lots of dedicated tools to even diagnose the problem, let alone fix it. We’ve traded simplicity and ease of repairability the appearance of “simple” which conceals a huge amount of complexity under the surface.

To refer back to the Lampson/Wheeler quote, the completion of it is, “Except, of course, for the problem of too many indirections.” Even forty years ago it was understood that too many layers of abstraction would eventually lead to problems. We are quickly reaching this point in networking today. With all the reliance on complex tools providing an overwhelming amount of data about every point of the network, we find ourselves forced to use dashboards and data lakes to keep up with the rapid pace of changes dictated to the network by systems integrations being driven by developer desires and not sound network systems thinking.

Networking professionals can’t keep up. Just as other systems now must be maintained by algorithms to keep pace, so too does the network find itself being run by software instead of augmented by it. Even if people wanted to make a change they would be unable to do so because validating those changes manually would cause issues or interactions that could create havoc later on.

Simple Solutions

So how do we fix the issues? Can we just scrap it all and start over? Sadly, the answer here is a resounding “no”. We have to keep moving the network forward to match pace with the rest of IT. But we can do our part to cut down on the amount of complexity and abstraction being created in the process. Documentation is as critical as ever. Engineers and architects need to make sure to write down all the changes they make as well as their proposed designs for adding services and creating new features. Developers writing for the network need to document their APIs and their programs liberally so that troubleshooting and extension are easily accomplished instead of just guessing about what something is or isn’t supposed to be doing.

When the time comes to build something new, instead of trying to plaster over it with an abstraction, we need to break things down into their basic components and understand what we’re trying to accomplish. We need to augment existing systems instead of building new ones on top of the old to make things look easy. When we can extend existing ideas or augment them in such as way as to coexist then we can worry less about hiding problems and more about solving them.

Tom’s Take

Abstraction has a place, just like NAT. It’s when things spiral out of control and hide the very problems we’re trying to fix that it becomes an abomination. Rather than piling things on the top of the issue and trying to hide it away until the inevitable day when everything comes crashing down, we should instead do the opposite. Don’t hide it, expose it instead. Understand the complexity and solve the problem with simplicity. Yes, the solution itself may require some hard thinking and some pretty elegant programming. But in the end that means that you will really understand things and solve the complexity conundrum.

DockerCon Thoughts – Secure, Sufficient Applications

Posted on by
1

containerssuspended

I got to spend a couple of days this week at DockerCon and learn a bit more about software containers. I’d always assumed that containers were a slightly different form of virtualization, but thankfully I’ve learned my lesson there. What I did find out about containers gives me a bit of hope about the future of applications and security.

Minimum Viable App

One of the things that made me excited about Docker is that the process isolation idea behind building a container to do one thing has fascinating ramifications for application developers. In the past, we’ve spent out time building servers to do things. We build hardware, boot it with an operating system, and then we install the applications or the components thereof. When we started to virtualize hardware into VMs, the natural progression was to take the hardware resource and turn it into a VM. Thanks to tools that would migrate a physical resource to a virtual one in a single step, most of the first generation VMs were just physical copies of servers. Right down to phantom drivers in the Windows Device Manager.

As we started building infrastructure around the idea of virtualization, we stopped migrating physical boxes and started building completely virtual systems from the ground up. That meant using things like deployment templates, linked clones, and other constructs that couldn’t be done in hardware alone. As time has rolled on, we have a method of quickly deploying virtual resources that we could never do on purely physical devices. We finally figured out how to use virtual platforms efficiently.

Containers are now at the crossroads we saw early on in virtualization. As explained by Mike Coleman (@MikeGColeman), many application developers are starting their container journey by taking an existing app and importing it directly into a container. It’s a bit more involved than the preferred method, but Mike mentioned that even running the entire resource pool in a container does have some advantages. I’m sure the Docker people see container adoption as the first step toward increased market share. Even if it’s a bit clumsy at the start.

The idea then moves toward the breakdown of containers into the necessary pieces, much as it did with virtual machines years ago. Instead of being forced to think about software as a monolithic construct that has to live on a minimum of one operating system, developers can break the system down into application pieces that can execute one program or thread at a time on a container. Applications can be built using the minimum amount of software constructs needed for an individual process. That means that those processes can be spread out and scaled up or down as needed to accomplish goals.

If your database query function is running as a containerized process instead of running on a query platform in a VM then scaling that query to thousands or tens of thousands of instances only requires spinning up new containers instead of new VMs. Likewise, scaling a web-based app to accommodate new users can be accomplished with an explosion of new containers to meet the need. And when the demand dies back down again, the containers can be destroyed and resources returned to the available pool or turned off to save costs.

Segment Isolation

The other exciting thing I saw with containers was the opportunity for security. The new buzzword of the day in security and networking is microsegmentation. VMware is selling it heavily with NSX. Cisco has countered with a similar function in ACI. At the heart of things microsegmentation is simply ensuring that processes that shouldn’t be talking to each other won’t be talking to each other. This prevents exposure by having your app database server visible on the public Internet, for instance.

Microsegmentation is great in overlay and network virtualization systems where we have to take steps to prevent systems from talking to each other. That means policies and safeguards in place to prevent communications. It’s a great way to work on existing networks where the default mode is to let everything on the same subnet talk to everything else. But what if the default was something different.

With containers, there is a sandbox environment for each container to talk to other containers in the same area. If you create a named container network and attach a container to it, that container gains a network interface on that particular named network. It won’t be able to talk to other containers on different networks without creating an explicit connection between the two networks. That means that the default mode of communications for the containers is restricted out of the box.

Imagine how nice it will be to create a network that isn’t insecure by default. Rather than having to unconnected all the things that shouldn’t speak, you can spend your time building connections between the things that should be speaking. That means a little mistake or forgotten connection will prevent communications instead of opening it up. That means much less likelihood that you’re going to cause an incident.

There are still some issues with scaling the networking aspect of Docker right now. The key/value store doesn’t provide a lot of visibility and definitely won’t scale up to tens or hundreds of thousands of connections. My hope is that down the road Docker will implement a more visible solution that can perform drag-and-drop connectivity between containers and leave an audit trail so networking pros can figure out who connected what and how that exposed everything. It also makes it much easier when the connection between the devices has to be explicit to prove intent or malice. But those features are likely to come down the road as Docker builds a bigger, better management platform.

Tom’s Take

I think Docker is doing things right. By making developers look at the core pieces they need to build apps and justify why things are being done the way they’ve always been done, containers are allowing for flexibility and new choices to be made. At the same time, those choices are inherently more secure because resources are only shared when necessary. It’s the natural outgrowth of sandboxing and Jails in the OS from so many years ago. Docker has a chance to make application developers better without making them carry the baggage of years of thinking along with them to a new technology.

BGP: The Application Networking Dream

Posted on by
2

bgp

There was an interesting article last week from Fastly talking about using BGP to scale their network. This was but the latest in a long line of discussions around using BGP as a transport protocol between areas of the data center, even down to the Top-of-Rack (ToR) switch level. LinkedIn made a huge splash with it a few months ago with their Project Altair solution. Now it seems company after company is racing to implement BGP as the solution to their transport woes. And all because developers have finally pulled their heads out of the sand.

BGP Under Every Rock And Tree

BGP is a very scalable protocol. It’s used the world over to exchange routes and keep the Internet running smoothly. But it has other power as well. It can be extended to operate in other ways beyond the original specification. Unlike rigid protocols like RIP or OSPF, BGP was designed in part to be extended and expanded as needs changes. IS-IS is a very similar protocol in that respect. It can be upgraded and adjusted to work with both old and new systems at the same time. Both can be extended without the need to change protocol versions midstream or introduce segmented systems that would run like ships in the night.

This isn’t the first time that someone has talked about running BGP to the ToR switch either. Facebook mentioned in this video almost three years ago. Back then they were solving some interesting issues in their own data center. Now, those changes from the hyperscale world are filtering into the real world. Networking teams are seeking to solve scaling issues without resorting to overlay networks or other types of workarounds. The desire to fix everything wrong with layer 2 has led to a revelation of sorts. The real reason why BGP is able to work so well as a replacement for layer 2 isn’t because we’ve solved some mystical networking conundrum. It’s because we finally figured out how to build applications that don’t break because of the network.

Apps As Far As The Eye Can See

The whole reason when layer 2 networks are the primary unit of data center measurement has absolutely nothing to do with VMware. VMware vMotion behaves the way that it does because legacy applications hate having their addresses changed during communications. Most networking professionals know that MAC addresses have a tenuous association to IP addresses, which is what allows the gratuitous ARP after a vMotion to work so well. But when you try to move an application across a layer 3 boundary, it never ends well.

When web scale companies started building their application stacks, they quickly realized that being pinned to a particular IP address was a recipe for disaster. Even typical DNS-based load balancing only seeks to distribute requests to a series of IP addresses behind some kind of application delivery controller. With legacy apps, you can’t load balance once a particular host has resolved a DNS name to an IP address. Once the gateway of the data center resolves that IP address to a MAC address, you’re pinned to that device until something upsets the balance.

Web scale apps like those built by Netflix or Facebook don’t operate by these rules. They have been built to be resilient from inception. Web scale apps don’t wait for next hop resolution protocols (NHRP) or kludgy load balancing mechanisms to fix their problems. They are built to do that themselves. When problems occur, the applications look around and find a way to reroute traffic. No crazy ARP tricks. No sly DNS. Just software taking care of itself.

The implications for network protocols are legion. If a web scale application can survive a layer 3 communications issue then we are no longer required to keep the entire data center as a layer 2 construct. If things like anycast can be used to pin geolocations closer to content that means we don’t need to worry about large failover domains. Just like Ivan Pepelnjak (@IOSHints) says in this post, you can build layer 3 failure domains that just work better.

BGP can work as your ToR strategy for route learning and path selection because you aren’t limited to forcing applications to communicate at layer 2. And other protocols that were created to fix limitations in layer 2, like TRILL or VXLAN, become an afterthought. Now, applications can talk to each other and fail back and forth as they need to without the need to worry about layer 2 doing anything other than what it was designed to do: link endpoints to devices designed to get traffic off the local network and into the wider world.

Tom’s Take

One of the things that SDN has promised us is a better way to network. I believe that the promise of making things better and easier is a noble goal. But the part that has bothered me since the beginning was that we’re still trying to solve everyone’s problems with the network. We don’t rearrange the power grid every time someone builds a better electrical device. We don’t replumb the house overtime we install a new sink. We find a way to make the new thing work with our old system.

That’s why the promise of using BGP as a ToR protocol is so exciting. It has very little to do with networking as we know it. Instead of trying to work miracles in the underlay, we build the best network we know how to build. And we let the developers and programmers do the rest.