1## Configuration Patterns
2 
3### Enable Workers Logs
4 
5```jsonc
6{
7  "observability": {
8    "enabled": true,
9    "head_sampling_rate": 1  // 100% sampling (default)
10  }
11}
12```
13 
14**Best Practice**: Use structured JSON logging for better indexing
15 
16```typescript
17// Good - structured logging
18console.log({ 
19  user_id: 123, 
20  action: "login", 
21  status: "success",
22  duration_ms: 45
23});
24 
25// Avoid - unstructured string
26console.log("user_id: 123 logged in successfully in 45ms");
27```
28 
29### Enable Workers Traces
30 
31```jsonc
32{
33  "observability": {
34    "traces": {
35      "enabled": true,
36      "head_sampling_rate": 0.05  // 5% sampling
37    }
38  }
39}
40```
41 
42**Note**: Default sampling is 100%. For high-traffic Workers, use lower sampling (0.01-0.1).
43 
44### Configure Analytics Engine
45 
46**Bind to Worker**:
47```toml
48# wrangler.toml
49analytics_engine_datasets = [
50  { binding = "ANALYTICS", dataset = "api_metrics" }
51]
52```
53 
54**Write Data Points**:
55```typescript
56export interface Env {
57  ANALYTICS: AnalyticsEngineDataset;
58}
59 
60export default {
61  async fetch(request: Request, env: Env): Promise<Response> {
62    // Track metrics
63    env.ANALYTICS.writeDataPoint({
64      blobs: ['customer_123', 'POST', '/api/v1/users'],
65      doubles: [1, 245.5], // request_count, response_time_ms
66      indexes: ['customer_123'] // for efficient filtering
67    });
68    
69    return new Response('OK');
70  }
71}
72```
73 
74### Configure Tail Workers
75 
76Tail Workers receive logs/traces from other Workers for filtering, transformation, or export.
77 
78**Setup**:
79```toml
80# wrangler.toml
81name = "log-processor"
82main = "src/tail.ts"
83 
84[[tail_consumers]]
85service = "my-worker" # Worker to tail
86```
87 
88**Tail Worker Example**:
89```typescript
90export default {
91  async tail(events: TraceItem[], env: Env, ctx: ExecutionContext) {
92    // Filter errors only
93    const errors = events.filter(event => 
94      event.outcome === 'exception' || event.outcome === 'exceededCpu'
95    );
96    
97    if (errors.length > 0) {
98      // Send to external monitoring
99      ctx.waitUntil(
100        fetch('https://monitoring.example.com/errors', {
101          method: 'POST',
102          body: JSON.stringify(errors)
103        })
104      );
105    }
106  }
107}
108```
109 
110### Configure Logpush
111 
112Send logs to external storage (S3, R2, GCS, Azure, Datadog, etc.). Requires Business/Enterprise plan.
113 
114**Via Dashboard**:
1151. Navigate to Analytics → Logs → Logpush
1162. Select destination type
1173. Provide credentials and bucket/endpoint
1184. Choose dataset (e.g., Workers Trace Events)
1195. Configure filters and fields
120 
121**Via API**:
122```bash
123curl -X POST "https://api.cloudflare.com/client/v4/accounts/{account_id}/logpush/jobs" \
124  -H "Authorization: Bearer <API_TOKEN>" \
125  -H "Content-Type: application/json" \
126  -d '{
127    "name": "workers-logs-to-s3",
128    "destination_conf": "s3://my-bucket/logs?region=us-east-1",
129    "dataset": "workers_trace_events",
130    "enabled": true,
131    "frequency": "high",
132    "filter": "{\"where\":{\"and\":[{\"key\":\"ScriptName\",\"operator\":\"eq\",\"value\":\"my-worker\"}]}}"
133  }'
134```
135 
136### Environment-Specific Configuration
137 
138**Development** (verbose logs, full sampling):
139```jsonc
140// wrangler.dev.jsonc
141{
142  "observability": {
143    "enabled": true,
144    "head_sampling_rate": 1.0,
145    "traces": {
146      "enabled": true
147    }
148  }
149}
150```
151 
152**Production** (reduced sampling, structured logs):
153```jsonc
154// wrangler.prod.jsonc
155{
156  "observability": {
157    "enabled": true,
158    "head_sampling_rate": 0.1, // 10% sampling
159    "traces": {
160      "enabled": true
161    }
162  }
163}
164```
165 
166Deploy with env-specific config:
167```bash
168wrangler deploy --config wrangler.prod.jsonc --env production
169```