The Emerging Technology of Podcast Hosting
As Simplecast prepares to release a new standard of podcast analytics and hosting, the best is about to get much better
Behind every podcast episode you publish or download there is a team of Simplecast engineers watching out for you. We know the delivery of your site, embed players, RSS feed and audio files are all critical to your success and we strive to provide you with the best and most reliable infrastructure out there. Our name says Simple, but the architecture of a truly powerful, redundant podcast hosting infrastructure is far from it—we sweat the hard stuff, so you don’t have to.
For over 5 years Simplecast has continued to build one of the most trusted names in podcast analytics, hosting, and distribution. As a powerful, time-tested platform, delivering billions of minutes of audio each year, Simplecast remains the easiest way for a podcaster to publish and distribute audio to the world. It’s no wonder we are one of the few approved partners of Apple Podcasts.
In the past year alone we’ve made great strides by giving you instant RSS feed uncaching, expanding and improving upon our global audio delivery, hiring a Head of Platform and Podcaster Success Lead, enabling live chat support, surfacing mapped city-level analytics and much, much more. But we’re not stopping there. We are constantly seeking ways to make things even faster and more reliable for you and your listeners! Here’s a look at what we’re doing behind the scenes.
Make It Better, Do It Faster
In any complex system, if you can achieve the same outcome with fewer variables and greater simplicity, you will have a better, faster system. Think Occam’s Razor but for programming. To accomplish this we looked at all the pieces of our current system and simplified how we deliver each portion without reducing user experience. This left us with insanely fast solutions. Let me explain a few of them…
With our embed players all we really need to do is show the end-user information about a particular podcast/episode and let it be played and shared. We were able to come up with a fully static solution served up via our global content delivery network (CDN). This means there is no server on the backend to slow things down. We can also harness the reliability of our global and redundant CDN.
Our current audio URLs are a pass-through URL that redirects to our CDN. While there were good reasons for this originally, it is also a point of failure and slows things down. We have developed a solution that eliminates and streamlines the process causing all requests to hit the CDN directly without having to pass-through a potentially slow, or unavailable, server interface.
Similar to audio files, RSS files are also dependent on our servers for delivering updates to the RSS feed. If an RSS file uncaches, and is requested, the CDN grabs the newest version from our servers. While a good solution, we will be eliminating the need for the CDN to hit our server, by continuously generating a static version served to our CDN.
Make Us Stronger
There are many things at the core of making a system reliable, but the heavy hitters are redundancy, scalability and simplicity.
Redundancy helps in the event a piece of our infrastructure fails. When a critical piece fails we can quickly switch to the backup.
While our CDN (what serves up all your audio files) is globally redundant we plan on making all other parts of our infrastructure globally redundant as well. What does this mean for you? Last year we saw the entire Northeast of the United States get taken down by cyber attacks. With global redundancy on all levels of our infrastructure, attacks of this magnitude would not affect our global infrastructure and Simplecast would function as normal.
It’s easy to scale slowly, but what happens when a podcast gets insanely popular overnight? Some things we’ve done to improve scalability is to remove unnecessary pieces.
We have also changed our codebase to Elixir which leverages the Erlang VM. Erlang is well known for it’s stability and scalability and gives us the ability for distributed computing, while also achieving massive scale.
In addition to a different codebase we are also rolling out a much better scaling solution. Our entire infrastructure will be able to handle massive continuous traffic as well as massive short lived bursts.
Reducing complexity in a system not only makes things faster, it also makes things more reliable. The fewer variables or moving parts, the fewer things that can go wrong. As we have reduced many of our points of failure and made our underlying system more simple we have made things more reliable.
A Look Ahead
One of the hardest parts of my job as a developer is waiting to release all the features and goodies I know we have in store for you. Over the past twelve months, we have been diligently engineering an all new Simplecast — a first-of-its-kind platform that will modernize podcasting as we know it.
We already had a stellar engineering team, but in recent weeks have added two incredible members, Brandon Bassett and Steven Crothers. I’m now confident we have one of the best engineering teams in the industry.
Brandon Bassett has been coding Elixir long before Elixir was what all the cool kids were using. He will be bringing his extensive skill and knowledge of Elixir and the Erlang VM to our team as he continues to build out the faster, simpler, better, more reliable backend of Simplecast.
With names like Namecheap, Bankrate and Amazon under his belt, Steven Crothers will be using his proven systems engineering experience to make the infrastructure and analytics of Simplecast simply the best in the industry. It should also be noted he makes some of the best spreadsheets you have ever seen.
With this quickly growing, big-brained team, we vow to ensure Simplecast remains the most trusted way for a podcaster to publish and distribute audio to the world. The future of Simplecast arrives this spring.