Docs/Overview
Docs/Serverless Compute

Serverless Compute & Sandboxing

Deploy backend APIs and dynamic Server-Side Rendered (SSR) web applications without managing virtual machines, provisioning clusters, or configuring reverse proxies. Exovon Compute scales from zero to parallel microcontainers automatically within Google Cloud Mumbai (asia-south1).

Sub-250ms Wake-Up

Optimized standalone microkernel containers scale to zero when idle and resume in under 250ms when incoming traffic arrives.

gVisor Microkernel

Every container executes inside a dedicated user-space microkernel, completely virtualizing system calls from host hardware.

Streaming & SSE

Native support for chunked HTTP responses, Server-Sent Events (SSE), WebSockets, and real-time AI token streaming.

Serverless Execution Lifecycle

Exovon Compute eliminates idle infrastructure expenses by decoupling compute billing from persistent uptime:

1
Dormant State (Scale to Zero)

Cost: ₹0.00

When an application receives zero requests for 15 minutes, its containers hibernate. Compute, CPU, and RAM allocations drop to zero until the next incoming connection.

2
Ingress Wake-Up & Execution

< 250ms

The Edge Router intercepts an incoming HTTP request, spins up an isolated microkernel container instance inside our Mumbai cluster, and proxies the request to port 8080. You are never billed for container initialization time; compute metering begins strictly when the application receives the HTTP request.

3
Horizontal Burst Autoscaling

80 Concurrency / Node

Each container instance processes up to 80 concurrent HTTP requests simultaneously. When traffic exceeds capacity, the orchestrator automatically provisions additional parallel container replicas in <500ms.

Compute Specifications & Quotas

Hardware resources allocated per container instance across plan tiers:

TiervCPU AllocationContainer RAMRequest TimeoutMax Replicas
Hobby (Free)0.5 vCPU (shared)512 MB10 seconds1 instance (scales to 0)
Cloud Starter1 vCPU (shared)1 GB30s HTTP / 15m StreamsUp to 4 instances
Pro1 vCPU (shared)1 GB60s HTTP / 30m StreamsUp to 4 instances
HeavyDedicated up to 8 vCPUs4 GB+ Dedicated120s HTTP / 60m StreamsDedicated Fleet

Streaming Protocols & WebSockets

Exovon Compute provides native support for streaming protocols, including HTTP chunked transfer encoding, Server-Sent Events (SSE), and bidirectional WebSockets:

Interactive Interactive Workloads

Exovon is optimized for event-driven, scale-to-zero serverless architecture. WebSockets are ideal for:

  • Collaborative canvas and live document sync (Yjs, Liveblocks)
  • Live visitor session telemetry and ephemeral in-app chats
  • Real-time token streaming from generative AI models

24/7 Daemon Guidance

An open WebSocket treats the container as actively in-flight and pauses scale-to-zero hibernation. For non-stop daemon listeners (e.g. continuous Discord bots, perpetual IoT sensors, continuous market tickers), dedicated compute instances or decoupled message queues are architecturally recommended.

Server-Sent Events (SSE) Route Implementation

app/api/chat/route.ts
// app/api/chat/route.ts
export const runtime = "nodejs";

export async function POST(req: Request) {
  const encoder = new TextEncoder();
  const stream = new ReadableStream({
    async start(controller) {
      controller.enqueue(encoder.encode("Streaming tokens from Mumbai (asia-south1)...\n"));
      controller.close();
    },
  });

  return new Response(stream, {
    headers: {
      "Content-Type": "text/event-stream",
      "Cache-Control": "no-cache",
      "Connection": "keep-alive",
    },
  });
}

Frequently Asked Questions

Do you bill for container startup time when scaling from zero?

No. You are never billed for wake-up boot time (<250ms). Container image decompression, microkernel initialization, and TCP port 8080 health checks are platform orchestration overhead absorbed entirely by Exovon. Billable compute begins strictly when your application receives the incoming HTTP request.

Do you bill for network transit latency?

No. Network latency (transit time between the visitor and Mumbai compute) is completely free. You only pay for the volume of data egressed from origin. Edge response buffering ensures a slow client on a weak mobile connection will never artificially extend your container's billed runtime.

Was this documentation page helpful?