FaaS – Frequently Asked Questions (FAQ)
What is Function as a Service (FaaS)?
Function as a Service (FaaS) is a serverless computing model that allows developers to deploy individual pieces of code called functions without managing servers or infrastructure. The platform automatically handles provisioning, scaling, execution, and lifecycle management, allowing developers to focus purely on application logic.
What does activating FaaS on E2E Cloud do?
Activating FaaS enables the serverless execution environment in your cloud account. Once activated, you receive the ability to create and deploy functions, invoke them through a public endpoint, and consume a predefined monthly compute allocation for CPU functions. Activation also provisions a dedicated outbound traffic IP for the account.
Why do I need to activate FaaS before creating functions?
FaaS activation initializes the serverless environment for your account. This step provisions platform resources such as compute allocation, networking configuration, and the outbound IP used for external communication from your functions.
What runtimes are supported in E2E Cloud FaaS?
E2E Cloud provides multiple runtime templates including Python, Node.js, Go, PHP, and C#. It also supports advanced frameworks like FastAPI, Flask, and Express.js, along with GPU-accelerated environments such as PyTorch and TensorFlow for AI workloads.
Can I run machine learning workloads on FaaS?
Yes. GPU-enabled runtimes are available that include frameworks such as PyTorch and TensorFlow. These environments allow developers to build AI inference pipelines, image processing systems, or other GPU-accelerated workloads directly within serverless functions.
What is the difference between CPU and GPU functions?
CPU functions are optimized for general-purpose workloads such as APIs, automation tasks, and lightweight processing. GPU functions are designed for compute-intensive operations such as machine learning inference, deep learning models, and high-performance image or data processing.
What is a function handler?
The function handler is the entry point of the function code. It receives the incoming request data and context information, processes the logic, and returns a response that is sent back to the caller.
How do I deploy a function on E2E Cloud?
Functions can be created through the E2E Cloud console by selecting a runtime template, writing code in the editor, and deploying it. Alternatively, code can be uploaded using a ZIP file that follows the platform's directory structure guidelines.
What are dedicated replicas?
Dedicated replicas are continuously running instances of a function that help handle incoming requests faster and maintain predictable latency under load. They are available only for CPU-based functions and remain active to reduce cold starts and ensure consistent performance during burst or high traffic.
How many dedicated replicas can a function have?
A CPU-based function can have up to 5 dedicated replicas configured through system settings. These always-ready instances handle requests without cold starts, ensuring faster response times and consistent performance.
What happens when a function receives a large number of requests?
The platform automatically scales function instances based on incoming traffic. Additional instances are started dynamically to handle the workload and are scaled down when demand decreases.
What is the function invocation URL?
Each deployed function receives a unique public endpoint. This URL allows the function to be invoked directly through HTTP requests from applications, services, or external systems.
What outbound IP address is used by functions?
All functions in an account use a dedicated outbound traffic IP provided during FaaS activation. This allows external systems to whitelist a single predictable IP for API integrations or security compliance.
Why is a dedicated outbound IP important?
A dedicated outbound IP ensures predictable network identity when functions communicate with external services. This is useful for integrations with APIs, databases, or systems that require IP-based access control.
How can I update my function code after deployment?
You can modify the code directly in the console editor or upload an updated ZIP file. After saving the changes, the function can be redeployed.
What configuration options are available for functions?
Functions can be configured with memory limits, execution timeout settings, environment variables, and replica settings. These configurations help optimize performance and resource usage.
What are environment variables in functions?
Environment variables allow you to store configuration values such as database URLs, service endpoints, or application settings separately from the function code. This makes the application easier to manage and update without modifying the code.
What does the function execution timeout mean?
Execution timeout defines the maximum time a function is allowed to run before it is automatically terminated. This prevents runaway processes and ensures efficient resource usage.
How can I monitor function activity?
The platform provides logs and metrics for each function. These logs help developers track execution details, debug errors, and analyze invocation behavior.
What are the possible states of a function?
Functions typically move through several states during their lifecycle, including deploying, running, and failed states. These states indicate whether the function is being deployed, actively running, or encountering issues during deployment or execution.
Why might a function fail during deployment?
A function may fail due to incorrect dependencies, syntax errors, unsupported libraries, or configuration issues. Logs provided by the platform help identify the root cause of deployment failures.
Can I deploy functions using my own container image?
Yes. The platform supports custom container images. Developers can provide a public container image URL and deploy their function using their own runtime environment while following platform security requirements.
What are common use cases for FaaS?
FaaS is commonly used for building APIs, automating workflows, processing data streams, running event-driven applications, performing image or file processing, and integrating services across distributed systems.
How does FaaS help reduce infrastructure management?
FaaS abstracts server management, capacity planning, scaling, and patching. The platform automatically handles these responsibilities so developers can focus entirely on application development and business logic.
Can functions interact with other cloud services?
Yes. Functions can connect with databases, object storage, messaging systems, APIs, and other cloud services. This enables developers to build complex distributed applications using lightweight serverless components.