Datasets:
roanbrasil
/
devops-gitops-corpus

Dataset card Data Studio Files
xet
 Community
1
text
stringlengths
0
59.1k
| `voltOpsSync.exportTimeoutMillis` | number | 30000 | Export timeout (ms) |
### Serverless Runtime
VoltAgent automatically detects serverless environments (Cloudflare Workers, Vercel Edge, Deno Deploy). For serverless apps, use `serverlessHono`:
```typescript
import { VoltAgent, serverlessHono } from "@voltagent/core";
import { honoServer } from "@voltagent/server";
new VoltAgent({
agents: { supportAgent },
serverless: serverlessHono(),
voltOpsClient: new VoltOpsClient({
publicKey: process.env.VOLTAGENT_PUBLIC_KEY!,
secretKey: process.env.VOLTAGENT_SECRET_KEY!,
}),
});
```
<|endoftext|>
# source: VoltAgent__voltagent/website/observability/tracing/node-based.md type: docs
---
title: Node-Based View
---
# Node-Based View
The node-based view shows the same trace as a visual flow. It helps you understand the overall path at a glance and see how agents, tools, memory, and model calls connect.
Use this view when you want the big picture, then drill into a specific node for details.
<video controls loop muted playsInline style={{width: '100%', height: 'auto'}}>
<source src="https://cdn.voltagent.dev/docs/observability/tracing/node-based.mp4" type="video/mp4" />
Your browser does not support the video tag.
</video>
<|endoftext|>
# source: VoltAgent__voltagent/website/observability/tracing/users.md type: docs
---
title: User Analytics
---
# User Analytics
The Users page groups traces by user so you can see who is impacted and where issues concentrate. It helps you go from "something feels off" to a specific user, session, and trace.
<video controls loop muted playsInline style={{width: '100%', height: 'auto'}}>
<source src="https://cdn.voltagent.dev/docs/observability/tracing/tracing-users.mp4" type="video/mp4" />
Your browser does not support the video tag.
</video>
<br/>
<br/>
## Key Metrics
Quick read of volume, success rate, latency, and cost over time. Use it to decide if the problem is quality, performance, or spend, and to link regressions to model switches.
<img
src="https://cdn.voltagent.dev/docs/observability/tracing/tracing-charts.gif"
alt="Trace filters overview"
style={{ maxWidth: "100%", borderRadius: "12px", display: "block", margin: "0 auto" }}
/>
<br/>
## User Feedback
Feedback turns user sentiment into signals you can track. Use it to confirm whether fixes improved the experience and to see which issues users feel most.
<img
src="https://cdn.voltagent.dev/docs/observability/tracing/tracing-feedback.gif"
alt="Trace filters overview"
style={{ maxWidth: "100%", borderRadius: "12px", display: "block", margin: "0 auto" }}
/>
<br/>
Use the charts to see feedback volume, sentiment trends, score clusters, dominant feedback keys, and when users leave comments.
## Conversations
Conversations group traces by session so you can follow a user flow end to end.
Use the cards (status, time range, model, cost) to pick the most suspicious session, then filter traces to that path.
![conversation](https://cdn.voltagent.dev/docs/observability/tracing/tracing-conversation-2.png)
<|endoftext|>
# source: VoltAgent__voltagent/website/observability/tracing/overview.md type: docs
---
title: Overview
---
# VoltOps Tracing
Tracing shows how a single user request runs from start to finish. As a developer, you use it to answer: