Datasets:
roanbrasil
/
devops-gitops-corpus

Dataset card Data Studio Files
xet
 Community
1
text
stringlengths
0
59.1k
```bash
pnpm netlify deploy --prod
```
Netlify prints the production URL when the deploy finishes. If you need build steps (TypeScript, bundling) add them to your Netlify build command.
## Observability notes
- The example keeps the in-memory span/log storage by default, exposed through `/observability/*` routes.
- When VoltOps credentials are present, OTLP exports run via fetch with retry; the Lambda execution waits for the response, so keep an eye on cold-start durations.
- VoltOps Console still falls back to HTTP polling because WebSockets are not available in the Netlify Function runtime.
## Feature limitations
- **Model Context Protocol** still requires a Node server. The current Netlify Function runtime does not expose stdio transports.
- **libSQL memory adapter** needs a TCP socket. For Netlify Functions use the in-memory adapter or an external Postgres/Supabase database.
import Tabs from '@theme/Tabs';
import TabItem from '@theme/TabItem';
### Memory configuration examples
<Tabs>
<TabItem value="in-memory" label="In-memory (default)" default>
```ts
import { Memory, InMemoryStorageAdapter } from "@voltagent/core";
const memory = new Memory({
storage: new InMemoryStorageAdapter(),
});
```
</TabItem>
<TabItem value="postgres" label="PostgreSQL">
```ts
import { Memory } from "@voltagent/core";
import { PostgresMemoryAdapter, PostgresVectorAdapter } from "@voltagent/postgres";
const memory = new Memory({
storage: new PostgresMemoryAdapter({
connectionString: env.POSTGRES_URL,
}),
vector: new PostgresVectorAdapter({
connectionString: env.POSTGRES_URL,
}),
});
```
</TabItem>
<TabItem value="supabase" label="Supabase">
```ts
import { Memory } from "@voltagent/core";
import { SupabaseMemoryAdapter } from "@voltagent/supabase";
const memory = new Memory({
storage: new SupabaseMemoryAdapter({
url: env.SUPABASE_URL,
serviceRoleKey: env.SUPABASE_SERVICE_ROLE_KEY,
}),
});
```
</TabItem>
</Tabs>
Use `netlify logs --target=production` to tail production logs and confirm your agent runs as expected.
<|endoftext|>
# source: VoltAgent__voltagent/website/deployment-docs/overview.md type: docs
---
title: Deployment Overview
description: Deployment options for VoltAgent apps across serverful and serverless runtimes.
slug: /
---
You can run VoltAgent in classic Node.js servers or in serverless (edge) runtimes. This section explains the options and links to detailed guides.
:::tip VoltOps Deploy
Looking for the easiest way to deploy? [VoltOps Deploy](./voltops.md) offers one-click GitHub integration, automatic builds, and managed infrastructure for your VoltAgent projects.
:::
## Supported scenarios
- **Server (Node.js)** – use `@voltagent/server-hono` (or another HTTP layer) and deploy on any host such as Fly.io, Render, AWS, Railway. Note: For IPv6-enabled platforms like Railway and Fly.io, configure `hostname: "::"` for dual-stack networking.
- **Serverless (edge runtimes)** – run VoltAgent on platforms like Cloudflare Workers, Vercel Edge, or Deno Deploy for low latency responses while using the shared serverless provider.
- **Serverless Functions** – deploy to Node-based functions such as Netlify Functions when you need Node compatibility but prefer managed cold starts over dedicated servers.
- **Hybrid** – keep heavy work on a Node server and expose lightweight endpoints from the edge.
## When to pick which?
- Choose **Node.js** if you need long-running tasks, heavy state, or many open connections.
- Choose **Serverless (edge)** when global reach and very low latency are more important than local disk access or Node-specific libraries.
- **Observability** works in both modes. On serverless runtimes, VoltAgent falls back to HTTP polling instead of WebSocket streaming.
## Network Configuration
When deploying to Node.js servers, you may need to configure the network binding depending on your platform: