Fluence Cloudless is a decentralized serverless solution deeply rooted in the decentralized physical infrastructure (DePin) paradigm. Compared to most serverless solutions, Fluence Cloudless allows you to achieve more functionality with less code, without worrying about the underlying infrastructure. Unlike centralized cloud providers, Cloudless allows you to choose capacity providers (such as data centers) from the market according to your application needs, customizing the availability, performance, service level agreements (SLAs), and budget requirements of your serverless infrastructure. Due to its DePin foundation, Cloudless can scale at a very low cost without sacrificing performance and service levels. In addition, Cloudless supports proof systems for function execution, such as Proof of Execution (PoE) and Proof of Correctness (PoP), which revolutionize various audits and accurate billing.
In summary, to make Fluence Cloudless useful to you, you need to complete four basic steps:
- Compile your business logic into Marine Wasm to create and manage your compute functions.
- Create Aqua scripts to orchestrate the compute functions.
- Create developer proposals to package your compute artifacts and deployment requirements.
- Submit the developer proposals to the market for matching.
Cloudless is available on a runtime called Marine, with more runtimes under development, providing high-performance, available, and portable compute functions across capacity providers. Additionally, Aqua allows you to orchestrate compute functions into cloudless deployments.
Once you have written your business logic in Rust and compiled it to the wasm-wasi target, you can create your cloudless deployments by simply specifying the cloudless distribution (i.e., the number of function instances, duration, resource types such as CPU and RAM, location, etc.) and the developer proposal, where you can specify the resource fees and deployment costs you are willing to pay.
Business logic written in Rust
Compiled to wasm32-wasi
Token purchase
Deployment specifications
Willingness to pay
Hosted payment
Sending matching capacity providers
Matching proposals
Matching proposals
Matching proposals
Hosted and executed deployment
Hosted and executed deployment
Hosted and executed deployment
Provider proposal
Compute market
Developer
Code
Compute functions
Cloudless deployment
Payment gateway
Cloudless distribution
Developer proposal
Capacity provider_1
Capacity provider_2
Capacity provider_k
Figure 1: Fluence Cloudless Development and Deployment Process Diagram
Before submitting a deployment, you need to obtain FLT and USDC tokens from a (fiat currency) gateway provider. The USDC tokens are used for hosting during each deployment and periodically compensating capacity providers based on valid proof of execution, demonstrating the correct execution of cloudless functions (including CPU and memory models, execution time, data payload, etc.).
Of course, during development on the testnet, you can request enough testnet tokens tFLT and tUSDC for free to support your development work.
Information
To launch the Fluence mainnet, billing follows an epoch model called the subscription model, rather than an execution-based model. This means that you pay capacity providers for hosting your deployments for a certain period of time, regardless of the number of executions or load. This is a temporary limitation.
An important part of the Fluence Cloudless developer experience is the Fluence CLI, which supports all aspects of development and deployment required for creating and managing cloudless applications.
Comparative positioning
In terms of functionality and development, Cloudless is very similar to other serverless solutions such as AWS Lambda, Azure Functions, and Google Cloud Functions. Unlike centralized serverless cloud providers, Fluence Cloudless leverages the power and portability of Wasm to provide a high-performance compute runtime, hosted and executed in a customizable, highly available capacity grid, eliminating single points of failure, access control, high exit barriers, and excessive leasing. See Table 1.
Table 1: High-level serverless provider comparison
Fluence Lambda AWS Lambda Azure Functions Google Cloud Functions
Runtime Wasm, more coming soon Multiple Multiple
Languages Rust, more coming soon Multiple Multiple
Workflow management Aqua Step Functions Azure Logic App
Orchestration √ √ √
Endpoint access P2P or HTTP gateway HTTP HTTP
Capacity (data centers) Transparent and optional Black box Black box
CLI Fluence CLI AWS CLI Azure CLI, Azure PowerShell
Unlike centralized cloud providers, the Fluence protocol enables an open, permissionless peer-to-peer network and compute market, facilitating crowdsourcing of (data center) capacity. Undoubtedly, the Fluence network is supported by a large number of Tier 4 data centers with top-tier CPU and RAM configurations and highly competitive service level agreements. The Fluence compute market is based on Inter-Planetary Consensus (IPC), a modern high-performance PoS blockchain compatible with EVM.
While developers can ignore the complexity of the peer-to-peer network, including network security solutions such as DDoS and other attack prevention or mitigation solutions, one notable difference between the Fluence protocol and other serverless solutions is that the Fluence network inherently does not support HTTP request and response calls. Function orchestration happens at the peer-to-peer level. However, this can be easily addressed using existing Fluence solutions at the network entry and exit points, such as HTTP gateways.
Use cases
Due to its on-demand, scalable, and pay-per-execution model, serverless computing is inherently suitable for event-driven architectures and systems. Some of the use cases supported by Fluence Cloudless are summarized in Table 2.
Table 2: Example use cases for Fluence Lambda
Type Application
Data processing Data preparation and processing
Process control Task scheduling
Messaging Sending notifications
API Mobile and web applications
Serverless workflows Orchestrating and composing functions
In conclusion, Fluence Cloudless supports all the event and time-triggered serverless development you are accustomed to.