Observability Export

Observability Export lets you send webhook delivery traces and metrics to your existing observability stack via OpenTelemetry (OTLP/HTTP). See your webhook pipeline health alongside the rest of your infrastructure in Grafana, New Relic, Honeycomb, Axiom, or any OTLP-compatible tool.

Plan Availability: Observability Export is available on Pro and Business plans.

• Pro: 2 export configs, traces + metrics
• Business: 10 export configs, traces + metrics + logs, resource filtering

How It Works

Hookbase instruments both directions of the webhook pipeline. One export config covers both, so inbound and outbound traces land in the same backend.

Inbound (receiving webhooks)

1. Ingest: when a webhook arrives, Hookbase creates a hookbase.ingest trace span and records payload size, duration, and event count metrics.
2. Delivery: when the webhook is delivered to your destination, Hookbase creates a hookbase.delivery span and records latency, success/failure, and retry metrics.

Both spans share a traceId, so you can follow the full journey in your observability tool, from webhook received to delivery completed.

Outbound (sending webhooks)

1. Send: each POST /v1/events (or /api/send-event) call creates a hookbase.outbound.send span. If the request has a W3C traceparent header, the span is a child of your span.
2. Delivery attempt: each attempt to each endpoint, retries included, creates a hookbase.outbound.delivery span under the send span. Its context goes to the endpoint in a traceparent header.

See Tracing Outbound Webhooks for the outbound attributes, metrics and header passthrough.

Telemetry is exported via fire-and-forget HTTP calls that never block your webhook processing. If your observability endpoint is down, Hookbase tracks consecutive errors and auto-disables the export after 10 failures.

Supported Providers

Setup

Via Dashboard

1. Go to Settings then Observability
2. Click Add Export
3. Select your provider — the endpoint auto-fills
4. Enter your auth credentials
5. Choose export types (traces, metrics, logs)
6. Set sampling rate (1-100%)
7. Click Create
8. Click the Test button to verify connectivity

Via API

curl -X POST "https://api.hookbase.app/api/observability" \
  -H "Authorization: Bearer {token}" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "Grafana Cloud",
    "provider": "grafana",
    "endpoint": "https://otlp-gateway-prod-us-west-0.grafana.net/otlp/",
    "authHeaders": {
      "Authorization": "Basic base64encodedcredentials"
    },
    "exportTypes": {
      "traces": true,
      "metrics": true,
      "logs": false
    },
    "samplingRate": 1.0
  }'

Provider Setup Guides

Grafana Cloud

1. Sign in at grafana.com and select your stack
2. Go to Connections then Add new connection and search OpenTelemetry (OTLP)
3. Copy the OTLP endpoint (includes your region)
4. Generate an API token under My Account then Security then API Keys
5. Build the Basic auth value: echo -n "instanceId:apiToken" | base64
6. In Hookbase, use Authorization as the header key and Basic <base64value> as the value

To view data: Explore then data source Tempo then Search then Service Name = hookbase

New Relic

1. Go to your profile then API Keys
2. Click Create a key and select type INGEST - LICENSE
3. Copy the key
4. In Hookbase, use Api-Key as the header key

To view data: Query Your Data then run SELECT * FROM Span WHERE service.name = 'hookbase' SINCE 1 hour ago

Honeycomb

1. Go to Account then Team Settings then API Keys
2. Create or copy an API key
3. In Hookbase, use x-honeycomb-team as the header key

To view data: Home then look for the hookbase dataset then Query tab

Axiom

1. Go to Settings then API Tokens
2. Create a token with ingest permissions
3. In Hookbase, use Authorization as the header key and Bearer <token> as the value

Span Attributes

Inbound spans include these attributes (outbound spans are listed in Tracing Outbound Webhooks):

Metrics

Sampling

Set a sampling rate between 1% and 100% to control data volume. At 100% (default), every event is exported. At 10%, roughly 1 in 10 traces sends telemetry. The decision is made by trace ID, so a trace is exported whole or not at all: you won't see a delivery span without its ingest span, or an outbound attempt without its send span. Sampling is applied per export config, so you can have a high-fidelity config for critical routes and a sampled config for high-volume sources.

Resource Filtering (Business)

On the Business plan, you can filter which resources export telemetry. Add source or route IDs (inbound) or application or endpoint IDs (outbound) to the resource filter. Only telemetry for those resources will be exported. This is useful for focusing observability on critical pipelines without paying for telemetry on every webhook.

Auto-Disable

If an export endpoint returns errors on 10 consecutive attempts, Hookbase automatically disables the config to avoid wasting resources. The config shows an "Auto-disabled" status in the dashboard. Fix the endpoint issue and re-enable it from Settings.

Security

• Auth headers are encrypted at rest using AES-256-GCM
• No webhook payload data is included in telemetry — only metadata (IDs, status codes, timing)
• Telemetry export never blocks webhook processing
• Each export has a 5-second timeout