Serverless Roadmap: Hosted and Installable Platforms
The survey conducted for our “Guide to Serverless Technologies” asked a series of questions about the vendors and technologies that we will be using in the next 18 months to drive our coverage at The New Stack.
The first half of the roadmap dealt with the platforms and frameworks on which serverless architecture is deployed. The leading hosted platforms are the big three cloud providers’ FaaS offerings. AWS Lambda has twice as many users as Azure Functions and more than three times as many as Google Cloud Functions. Although on-premises deployments are not widespread, they are widely being considered. In fact, the percentage planning to use Kubeless, OpenFaaS, and Apache OpenWhisk rivals those looking at AWS, Microsoft and Google. In other words, the battle for the next wave of serverless adoption is alive and well.
A wide range of range of companies are actively being considered for their hosted serverless platforms. We look forward to seeing how many paying customers these vendors get in the next 18 months. Below are in-depth coverage of six of these options:
- Netlify Embeds Serverless Functionality into Its Web App Development Platform
- Twilio Previews a Serverless Capability, Called Functions, to Manage Messaging Apps
- StdLib: A Serverless Library for Building Developer Velocity
- Binaris Wants to Reduce Latency in Serverless Productions
- Spotinst: Making the Most of Cheaper Excess Compute Capacity
Interpreting the Data
As with most web-based surveys, this study does not claim to accurately describe market share. Instead, it just describes the current state of adoption among the 382 respondents that say they are using or planning to use serverless architecture in the next 18 months. High levels of future consideration are sometimes the result of brand recognition as opposed to concrete plans to use a specific product or service.
The New Stack took steps to limit the amount of vendor influence on the survey. For example, responses from identical companies or IP addresses were excluded from the study. Furthermore, UTM codes were used to track which respondents learned of the survey via a vendor. Although a few vendors may have received a one percentage point bump due to these respondents, this had a marginal impact on the total “in use + in plan” figures.
The full data set for the survey results is available online.
Feature image via Pixabay.