Appearance

Sampling & diagnostics ​

Control event volume and debug ingest issues.

Sampling ​

Send only a fraction of events when volume is high.

Node.js ​

typescript
const client = new AmlexiaClient({
  sdkKey: process.env.AMLEXIA_SDK_KEY!,
  sampleRate: 0.25, // 25% of track() calls sent
});

Python ​

python
client = AmlexiaClient(sdk_key="...", sample_rate=0.25)

Go / Ruby ​

Set SampleRate / sample_rate on client options (see package docs).

Sampling is applied per track() call before buffering. Use 1 (default) in development; lower in high-QPS production if needed.

WARNING

Error spikes may be under-represented at very low sample rates. Keep critical paths at 1 or use a dedicated low-volume project for alerts.

Diagnostic mode ​

Logs buffer size, flush success/failure to stderr — no data sent to Amlexia beyond normal events.

Node.js ​

typescript
const client = new AmlexiaClient({
  sdkKey: process.env.AMLEXIA_SDK_KEY!,
  diagnostic: true,
});

After flush:

typescript
const state = client.getDiagnosticState();
// { enabled, eventsBuffered, lastFlushAt, lastError }

Python ​

python
client = AmlexiaClient(sdk_key="...", diagnostic=True)

Health check ​

Verify SDK key and ingest reachability without sending events.

bash
npx amlexia health
amlexia health   # Python

Programmatic (Node):

typescript
import { checkIngestHealth } from '@amlexiahq/node';

const result = await checkIngestHealth('https://ingest.amlexia.com');

CLI in CI ​

yaml
# GitHub Actions example
- run: npx amlexia health
  env:
    AMLEXIA_SDK_KEY: ${{ secrets.AMLEXIA_SDK_KEY }}
    AMLEXIA_INGEST_URL: https://ingest.amlexia.com