The Day the Internet Died
For many (not me), it was one of the most terrifying hours in the history of their lives. From about 10:45 AM EST to 11:50 AM EST on Sunday, September 20th, 2015–mainly throughout the Northeast of the United States of America (as compared to the United States of Mexico…did you know that is Mexico’s official name?)–various websites and services like Netflix (OMFG!! Personally, I don’t give a shit about that), Amazon (including the Amazon Echo and Instant Video), Nest products, Reddit (another one I don’t give a shit about), Medium, IMDB, SocialFlow (another one I don’t care about…actually, I don’t care about any these), and others weren’t available.
You read that right. You couldn’t watch movies and shows on Netflix or Amazon. You couldn’t visit sub-Reddits. You couldn’t read posts on Medium. You couldn’t talk to Alexa!
Your world was effectively over, wasn’t it?
The cause of this (supposedly) was that one of Amazon’s web servers (AWS)–particularly the DynamoDB server group in Virginia–was down for the count for a little while.
The following day (today, 9/21/15), Skype was also down for many around the world–not just in the USA. MY GOD! What a terrible couple of days for the World Wide Web! And thus, what a terrible couple days for you!
For me personally, it didn’t mean much to my usual workflow and my personal free time. I keep local copies of all of my entertainment media (terabyte after terabyte of Stargate, Star Trek, Star Wars, Babylon 5, original Battlestar Galactica, Buck Rogers, and terabytes of mp3’s), and as much as is possible, I don’t run or use cloud services (outside of using OneDrive for backing up all of that data…they offer unlimited online storage, you know). I recognize that makes me a rarity in today’s world, but I take “being rare” as a point of pride.
But enough about me, what does all of this illustrate? Well, as I talk about often on my science and tech podcast–Sovryn Tech–and on the Dark Android Blog, there is a concept that I named (not really created) and promote which is called: “DAPS“. DAPS is an acronym for “Decentralization, Anonymity, Privacy, and Security”. While many times I will highlight dangers to and ways of improving the “APS” of that acronym, the “D” part of it doesn’t always come to fore.
Well, with the recent “internet outages” taking various forms over the past couple of days, decentralization becomes all the more important, since its antithesis is the entire reason these outages occurred. Centralization–in the sense of the internet–is best described with a real case being done on the internet itself: Google. As Alphabet’s premiere company, Google has been creating and integrating more and more services (Google Docs, YouTube, Google+, Google Photos, Google Keep, Google Play Music, etc.) to the point that just about anything you could possibly want to do online (and even offline) is to be done through Google’s own products and services, and all through your Google account. This is an example of centralization.
But what happens if you lose access to your Google account through some means? Or what happens if something happens to Alphabet/Google (you might think it implausible, but crazier things have happened)? All of that work, all of that data, all of those pictures, all of it…gone. Theoretically forever. And as analogy, this is what can–and does–happen when various websites and web services (and even web products) rely on servers and servers alone. This is what happens when you rely on the cloud and the cloud alone.
I could go on and on, because the potential for the inability to engage your work and your enjoyment in life when it relies on centralized servers goes far beyond what I’ve briefly described. Consider games and apps that REQUIRE YOU to connect to some company’s server to even begin being used (ie: DRM), regardless of whether or not the complete game or app is locally stored on your device or computer. You won’t be able to use those anymore. And there are other examples of the flaws and problems of central servers. And I think some of these companies are waking up to these concerns as Amazon and even Alphabet/Google allows for local storage of movies, music, documents, and other content. But let’s be clear, that is merely a band-aid on a gigantic bleeding wound-of-an-internet that really needs a whole new body: but the body of a hydra that no matter how many heads you cut off, the thing can never die.
Another analogy for you in case it all wasn’t coming across. Consider the cloud/World Wide Web as a hard drive. No IT or technology professional would ever tell you to connect a bunch of hard drives together in a RAID in the formation of RAID 0 if what you’re storing or using is important. Sure, RAID 0 will give you more speed and more storage space, but if one drive goes bad, all the drives in the RAID 0 go bad. Those professionals would recommend RAID 1 (or the like), which makes a mirror copy of everything you do, and stored throughout a myriad of hard drives. This is the concept of “resiliency”, and while you don’t get the same speeds you could with a RAID 0…what does speed matter when you’ve experienced a total loss of data and services? It doesn’t matter for shit. Resiliency is perhaps the most important concept (along with encryption, perhaps) to build into any technology, and the World Wide Web (which is distinct from the internet infrastructure itself) “as is” is not conducive to resiliency.
There are projects out there that are conducive to internet-style resiliency, and they follow the proven and solid ideas and implementations of P2P (peer-to-peer technologies). There are “decentralized internets” (yes, that’s plural) being created today. Maelstrom. ZeroNet. MaidSafe. And others. And some people have been calling for this with real solutions for a while. These projects need our support. Every time these widespread website, web service, and web-connected product outages happen, it’s a doleful reminder of how frail our interconnected world is. Jumping on these ideas now would cut the frailty off at the pass and allow for so much more than resilient data and communication. The possibilities are endless when you get away from centralization (in all of its forms).
For me personally? Again, none of these outages really bother me. In fact, anecdotally, this morning I noticed that Facebook and Gmail were both down and it changed practically nothing for me. I’m not sure what was behind that or how localized it was, but what I quickly noticed was that this website you are on right now wasn’t down at all. Many of the sites that I visit were still up (SIDE NOTE: Throughout all of what I’ve described in this blog post, my joyous torrent downloads have never skipped a beat. Hooray for P2P, baby!). And that was largely possible because I don’t rely upon “one company to rule them all” as far as my workflow and entertainment goes. That is because I’ve even integrated decentralization into my habits (as in, I don’t rely on Facebook to find my info). Like I said, much of what I do personally is on local storage. And I’m not saying that is the way one should do things–like I said, I’m a rarity–but I am saying that the way things are done on the World Wide Web today is a massive failure waiting to happen. And the proof is in the pudding.