Much has been written about the potential of serverless platforms to change the way developers - whether they are independent or part of a team - build and deploy web applications. For instance, serverless development platforms remove many of the most challenging infrastructure demands that can hinder rapid application development.
Recently, the Head of Business at Velo - Navot Volk, discussed these issues in an opinion piece on DZone. In the post, Volk helps developers understand the current serverless landscape and what it means for them as professionals.
“While serverless platforms are by their very nature more restrictive, once you accept those restrictions it can make life much easier. Security, backups, servers — all of these infrastructure challenges are simply handed off to the serverless provider.
“Serverless is even more attractive when we take into account all of the demands we place on developers. For most developers, we are asking way too much and making their lives way too hard. They are tasked with developing for the frontend and backend, working according to the constraints of DevOps teams, or independently handling infrastructure challenges. This doesn’t even begin to account for the demands of device and platform fragmentation.”
Velo itself is a serverless platform that allows developers to quickly build web applications. Recently, new capabilities have been added to Velo to allow users to use their own local development tools and collaborate across teams.
Adding these features, along with other capabilities, required solutions for a number challenges. In the post, Volk details how the Velo team solved some of these problems.
“Now, to actually deliver this experience, we had to solve a big problem. We looked at traffic across our own service, especially at low traffic sites, which make up a huge part of the industry. We saw that most instances that are running right now are idle over 86 percent of the time and not handling even a single user request.
“So, we are paying incredible amounts of money for computers that are doing nothing. Now, there’s a good reason for that. When a request arrives, you need to be able to respond fast. If it takes two minutes or even two seconds to start an instance, no one is going to wait. So, you need to have the instance up and running beforehand.
“We had to find a way to manage our grid and infrastructure in such a way that a user’s service is up when it's needed and down when it’s not needed. This means there is zero latency for cold starts. With other environments, you need to start an environment, wait for it to be up and running, and only then can you start working with that environment.”
Read the full post over on DZone.
To try Velo, register here. Alternatively, if you already have a Wix site click Dev Mode and click Turn on Dev Mode.