Better Stack

Ship logs to Better Stack Telemetry

Send logs to Better Stack (formerly Logtail). Logs arrive with a dt timestamp, level, message, and the full meta object — queryable in Live tail and with SQL.

Setup

In Better Stack, create a source (platform: JavaScript / HTTP) and copy its source token. Newer sources also show a dedicated ingesting host.
Set the environment variables:

BETTER_STACK_SOURCE_TOKEN=your-source-token
BETTER_STACK_INGESTING_HOST=https://s123456.eu-nbg-2.betterstackdata.com # if your source shows one

Wire the transport:

import { Elysia } from 'elysia'
import logixlysia from 'logixlysia'
import { createBetterStackTransport } from 'logixlysia/better-stack'

const app = new Elysia()
  .use(
    logixlysia({
      config: {
        transports: [createBetterStackTransport()]
      }
    })
  )
  .get('/', () => 'ok')
  .listen(3000)

Trigger a request and watch Live tail.

Environment Variables

Variable	Required	Description
`BETTER_STACK_SOURCE_TOKEN`	Yes	Source token, sent as a Bearer token
`BETTER_STACK_INGESTING_HOST`	No	Dedicated ingesting URL (default `https://in.logs.betterstack.com`)

Options

const betterStack = createBetterStackTransport({
  endpoint: 'https://s123456.eu-nbg-2.betterstackdata.com'
})

Option	Type	Default	Description
`sourceToken`	`string`	`BETTER_STACK_SOURCE_TOKEN`	Source token
`endpoint`	`string`	`https://in.logs.betterstack.com`	Ingesting URL

Plus the shared batching options: maxBatchSize, flushIntervalMs, timeout, retries — see the overview.

Payload

Each log posts as one JSON object in a batched array:

{
  "dt": "2026-08-22T12:00:00.000Z",
  "level": "INFO",
  "message": "GET /users",
  "request": { "method": "GET", "url": "http://localhost:3000/users" },
  "status": 200,
  "durationMs": 12.4,
  "context": { "requestId": "0d5e…" }
}

Nested fields are queryable directly, e.g. context.requestId or request.method in SQL and Live tail filters.

Troubleshooting

401 — the source token is wrong, or the source was deleted; copy a fresh token from the source settings.
Logs land in the wrong source — each source has its own token; double-check which one is in the environment.
Newer source, nothing arrives — sources with a dedicated ingesting host reject the legacy default endpoint; set BETTER_STACK_INGESTING_HOST.