Worth Reading: Redundancy does not imply fault tolerance

It’s a tough life being the developer of a distributed datastore. Thanks to the wonderful work of Kyle Kingsbury (aka, @aphyr) and his efforts on Jepsen.io, awareness of data loss and related issues in the face of network delays and partitions is way higher than it used to be. Under the Jepsen tests, every single system tested other than ZooKeeper exhibited problems of some kind. But many teams are stepping up, commissioning tests, and generally working hard to fix the issues even if they’re not all there yet. —the morning paper

The post Worth Reading: Redundancy does not imply fault tolerance appeared first on 'net work.

Worth Reading: Fearing Facebook

I used to greatly admire Mark Zuckerberg. Now, I fear him. Here’s why. Facebook is now a supranational company. It spans the globe with about 2 billion subscribers, its 2016 revenue reached $27.6B—54% more than in 2015; it’s immensely profitable (37% of revenue) and cash-rich with $29.56 “in the bank”. Decades ago, European Communists complained about “multinationals”, companies whose spread over multiple countries allowed them to escape full accountability in any one location. —Monday Note

The post Worth Reading: Fearing Facebook appeared first on 'net work.

On the ‘net: PCEP

The original Path Control Element Protocol (PCEP) work dates from the early 2000’s, with the first IETF RFC (4655) being made informational in 2006—which means PCEP predates the time when SDNs were “cool.” PCEP was originally, because of the increasingly complex nature of computing Traffic Engineering (TE) paths through (primarily), Service Provider (SP) networks. Three specific developments drove the design, standardization, and deployment of PCEP —ECI Blog

The post On the ‘net: PCEP appeared first on 'net work.

Worth Reading: Fat Finger Folly

You see, many newly created production systems, when they first come online, focus more on functionality and less on supportability. They are built to provide a specific capability. Of course, the services must be maintained, but maintaining these services often involve manual tasks by operations individuals, tasks such as logging on to servers and rebooting them, or terminating processes, or updating a configuration file. For many production systems, this is how they are maintained. Individuals login to the production servers and high privileged or “superusers” and manually perform necessary actions. This is, unfortunately, a very common process in the industry. —The New Stack

The post Worth Reading: Fat Finger Folly appeared first on 'net work.

History!

The audio is a bit low on this video; it actually recorded a bit lower than this, and the process of amplifying made it poor quality. I’ve already separated the audio recorder from the camera; now I’m upgrading the mic, and playing with the settings on the recorder to make the audio better quality. I’m also playing with putting text and drawings on the side of the video, a concept I intend to use more often in the future.

So this is a bit of a play video, but with a somewhat serious topic: the importance of learning the history of network engineering. Some resources are included below.

A bit history of the Internet
Net Heads versus Bell Heads
On the History of the Shortest Path Problem
The Elements of Networking Style
Software Defined Networks has a great introductory section with a good bit of history

The post History! appeared first on 'net work.

Worth Reading: Comments on the Wikileaks CIA Vault #7

The CIA didn’t remotely hack a TV. The docs are clear that they can update the software running on the TV using a USB drive. There’s no evidence of them doing so remotely over the Internet. If you aren’t afraid of the CIA breaking in an installing a listening device, then you should’t be afraid of the CIA installing listening software. —Errata Security

The post Worth Reading: Comments on the Wikileaks CIA Vault #7 appeared first on 'net work.

Worth Reading: Web Scale Networking

In this episode of The New Stack Makers, we spoke to Josh Leslie, Chief Executive Officer, and JR Rivers, Chief Technology Officer and co-founder of Web-scale networking software provider Cumulus Networks. We discussed how to best define web-scale networking, and how Cumulus Linux is different from traditional networking solutions. They spoke to the need for inter-system communication for new workloads, and how that factors into building a software-defined data center. They told us about their company’s recent product release Cumulus Express, as well as its announcement around Facebook’s Backpack, an open network chassis. Our conversation included points about hyper-converged infrastructure, customer roadblocks, and the roadmap for Cumulus Networks. —The New Stack

The post Worth Reading: Web Scale Networking appeared first on 'net work.

Worth Reading: When your code has to work

Douglas Crockford famously declared browsers to be “the most hostile software engineering environment imaginable,” and that wasn’t hyperbole. Ensuring that our websites work across a myriad of different devices, screen sizes and browsers our users depend on to access the web is a tall order, but it’s necessary. If our websites don’t enable users to accomplish the key tasks they come to do, we’ve failed them. —Smashing Magazine

The post Worth Reading: When your code has to work appeared first on 'net work.

Tan Gate

The post Tan Gate appeared first on 'net work.

Worth Reading: The nuanced future of 5G

5G sounds like the successor to 4G cellular telephony, and indeed that is the intent. While the progression from 2G to 3G, to 4G and now 5G seems simple, the story is more nuanced. At CES last month I had a chance to learn more about 5G (not to be confused with the 5Ghz WiFi) as well as another standard, ATSC 3.0 which is supposed to be the next standard for broadcast TV. —CircleID

The post Worth Reading: The nuanced future of 5G appeared first on 'net work.

Worth Reading: Keys to solving wicked problems

The incredible pace of change of the Internet — from research laboratory inception to global telecommunication necessity — is due to the continuing pursuit, development and deployment of technology and practices adopted to make the Internet better. This has required continuous attention to a wide variety of problems ranging from “simple” to so-called “wicked problems”. Problems in the latter category have been addressed through collaboration. This post outlines key characteristics of successful collaboration activities. —CircleID

The post Worth Reading: Keys to solving wicked problems appeared first on 'net work.

Worth Reading: Bouncing back to private clouds

There is an adage, not quite yet old, suggesting that compute is free but storage is not. Perhaps a more accurate and, as far as public clouds are concerned, apt adaptation of this saying might be that computing and storage are free, and so are inbound networking within a region, but moving data across regions in a public cloud is brutally expensive, and it is even more costly spanning regions. —The Next Platform

The post Worth Reading: Bouncing back to private clouds appeared first on 'net work.

Worth Reading: Creating an IXP in Denver

The word Internet is short for internetwork. It’s just a network of networks. So the more places you can connect those networks, the more robust the whole system is. That’s what Internet Exchange Points (“IXPs”) are. They’re the connection points where networks can connect to each other, and they’re a crucial part of the infrastructure of the Internet. In Europe, IXPs have traditionally been independent and are often run as nonprofits, whereas in North America, they’ve typically been owned and operated by commercial colocation facility operators or Internet Service Providers (ISPs). In the last several years, though, there’s been a movement in the US to build more independent, community-focused IXPs. IX-Denver is part of that movement. —The Internet Society

The post Worth Reading: Creating an IXP in Denver appeared first on 'net work.

Leaky Abstraction: An Example

The paper in question is from 2008, and the remedies have already been implemented in every BGP implementation I know of (in fact, getting rid of the scanner is something that just went into FR Routing). The paper can be found here, and a presentation based on the paper can be found here. They use a different example than mine in the paper—it’s a bit more subtle.

In this case, the abstraction is treating an AS as a single “thing” to provide stability in the larger routing system. But any time you have an abstraction, including summarization and aggregation, there is always some possibility the abstraction will leak. Here the leak is an internal route metric change “leaking” into an actual change in the path through the network, although both paths still exist. There are more subtle changes, such as a route metric change causing an update to be sent even though the path itself hasn’t even changed, but I chose this example to make it easier to understand.

The post Leaky Abstraction: An Example appeared first on 'net work.

Worth Reading: The rise of the conversational interface

Eight trillion. That’s the number of text messages sent every year. For the record, that’s 23 billion texts sent every day. Or almost 16 million per minute. On July 7, 2016, Facebook reported that over one billion people now use Messenger, up 200 million since January 2016. As of February 1, 2016, WhatsApp logged one billion users on its platform. In case you didn’t know, Facebook owns WhatsApp. —A Little More Human

The post Worth Reading: The rise of the conversational interface appeared first on 'net work.

Worth Reading: Cisco’s identity crisis

Our Cisco team has been reaching out to get feedback on our relationship with Cisco and its products — a healthy practice for any vendor. I’ve tried to be open, honest, and consistent in all our talks. As I mentally review our conversations, I conclude I’ve been contradictory. On one hand, I’ve talked about how the industry is changing and Cisco’s products need to evolve in a software-defined marketplace. At the same time, I’ve decried their decision to move last-generation data center products to the campus portfolio to make way for newer technology. —Eyvonne Sharp

The post Worth Reading: Cisco’s identity crisis appeared first on 'net work.

Automation

The post Automation appeared first on 'net work.

Worth Reading: Ransomware for Dummies

Among today’s fastest-growing cybercrime epidemics is “ransomware,” malicious software that encrypts your computer files, photos, music and documents and then demands payment in Bitcoin to recover access to the files. A big reason for the steep increase in ransomware attacks in recent years comes from the proliferation of point-and-click tools sold in the cybercrime underground that make it stupid simple for anyone to begin extorting others for money. —Krebs on Security

The post Worth Reading: Ransomware for Dummies appeared first on 'net work.

Worth Reading: Five jobs robots will take first

Oxford University researchers have estimated that 47 percent of U.S. jobs could be automated within the next two decades. But which white-collar jobs will robots take first? First, we should define “robots” (for this article only) as technologies, such as machine learning algorithms running on purpose-built computer platforms, that have been trained to perform tasks that currently require humans to perform. —Shelly Palmer

The post Worth Reading: Five jobs robots will take first appeared first on 'net work.

CoDel and Active Queue Management

Buffering packets in a network is both good and bad. It is good because a buffer can hold packets from one stream while another stream’s packets are being processed, to take up and release short bursts of traffic, to hold and then release packets when there is a very short interruption on the wire (or during a route change), and in many other situations. However, queues are bad when there is a standing queue, which means a particular flow always has some number of packets in a particular queue along the path between the source and the destination. This normally occurs at the narrowest point along the path, or rather the link with the lowest bandwidth. In a previous post, I looked at BBR, a change to the way TCP computes its window sizes, that attempts to reduce the amount of traffic “in flight” between a sender and receiver to reduce the number of packets being held in a particular buffer along the way.

This post will consider another solution: CoDel. CoDel is essentially an improved tail drop mechanism that provides the correct signals to TCP to slow down its send rate, or rather to reduce the window size (and Continue reading