> Agent-readable docs index: /llms.txt. Download /docs.zip to grep all markdown files locally.

---
$schema: https://holocron.so/frontmatter.json
title: Cloudflare Workers OTel
description: Reference for Cloudflare Workers OpenTelemetry traces and logs, with payload examples and query recipes.
icon: lucide:cloud
---

import CloudflareWorkersOtel from '../../../docs/cloudflare-workers-otel-export.md'

# Cloudflare Workers OTel export via destinations

This document is a **querying reference** for agents and humans working with Cloudflare Workers telemetry inside Strada.

Use it to answer questions like:

* **Is this span from Cloudflare Workers?**
* **Why is this request slow?**
* **Did the Worker fail with an uncaught exception?**
* **Is this a D1, KV, R2, RPC, Queue, or Durable Object span?**
* **Which attributes should a SQL query filter on first?**

Cloudflare Workers can export **OpenTelemetry traces and logs** to any **OTLP HTTP endpoint** configured as a destination in the Cloudflare dashboard.

For Strada, that means these endpoints are the important ones:

* `POST /v1/traces`
* `POST /v1/logs`

Cloudflare does **not** currently export metrics through this destination flow.

## Quick answers

### Is a row from Cloudflare Workers?

Usually filter on:

```sql
ResourceAttributes['cloud.platform'] = 'cloudflare.workers'
```

or:

```sql
ResourceAttributes['cloud.provider'] = 'cloudflare'
```

### What are the most important attributes?

Start with these:

* `cloud.platform`
* `cloud.provider`
* `service.name`
* `faas.invocation_id`
* `cloudflare.ray_id`
* `cloudflare.handler_type`
* `cloudflare.entrypoint`
* `cloudflare.execution_model`
* `cloudflare.outcome`
* `cloudflare.cpu_time_ms`
* `cloudflare.wall_time_ms`
* `url.path`
* `http.request.method`
* `http.response.status_code`
* `cloudflare.colo`

### What attributes are best for slow requests?

Look at:

* span `Duration`
* `cloudflare.wall_time_ms`
* `cloudflare.cpu_time_ms`
* `cloudflare.response.time_to_first_byte_ms`
* `db.query.text`
* `cloudflare.d1.response.sql_duration_ms`
* `cloudflare.d1.response.rows_read`
* `cloudflare.kv.response.cache_status`
* `cloudflare.r2.operation`

### What attributes are best for failed Worker executions?

Look at:

* `cloudflare.outcome`
* `http.response.status_code`
* `cloudflare.ray_id`
* logs for surrounding context, especially:
  * request log bodies like `GET https://.../route`
  * app `console.error()` output
  * framework-generated error log lines such as `Spiceflow unhandled error: Error: ...`

Validated note: in real Cloudflare destination exports, **runtime crashes were reliably marked on traces**, while logs did **not** include standard `exception.*` fields.

Important documented outcomes include:

* `ok`
* `exception`
* `exceededCpu`
* `exceededMemory`

## What Cloudflare exports

Cloudflare Workers destinations export:

* **traces**
* **logs**

Cloudflare Workers destinations do **not** currently export:

* **metrics**

So for Strada, the relevant ingest paths are:

* `https://<project>-ingest.strada.sh/v1/traces`
* `https://<project>-ingest.strada.sh/v1/logs`

## Validated behavior from a real Cloudflare Worker

Strada now includes a real validation worker at `example-app-cloudflare/` used to test actual Cloudflare destination export.

The collector had to support **gzip-compressed OTLP** first, because Cloudflare destinations send:

* `content-encoding: gzip`

Without gzip decompression, the collector failed before any Cloudflare-specific behavior could be inspected.

### Route matrix used for validation

| Route            | What it does                                    | What Cloudflare exported                                                    | What matters for Strada          |
| ---------------- | ----------------------------------------------- | --------------------------------------------------------------------------- | -------------------------------- |
| `/ok`            | healthy request                                 | request trace + request log + app log                                       | baseline shape                   |
| `/caught`        | error caught and converted into HTTP 500        | request trace + request log + app log                                       | **not** a runtime crash          |
| `/throw`         | thrown inside Spiceflow route                   | request trace + request log + framework error log                           | `cloudflare.outcome` stayed `ok` |
| `/throw-async`   | async throw inside Spiceflow route              | request trace + request log + framework error log                           | `cloudflare.outcome` stayed `ok` |
| `/crash-runtime` | throw before `app.handle()` in worker `fetch()` | request trace with `cloudflare.outcome = exception` + request log + app log | **real runtime crash marker**    |

### Key validated findings

* **Framework-handled route errors are not Cloudflare runtime crashes.**
* **Real Worker runtime crashes are marked by `cloudflare.outcome = exception` on the root span.**
* A real runtime crash also returned the Cloudflare runtime error page body `error code: 1101` to the client.
* Real destination-exported logs did **not** include:
  * `exception.type`
  * `exception.message`
  * `exception.stacktrace`
  * `$workers.outcome`
  * `$metadata.error`
* The most reliable signal for materializing Cloudflare runtime crashes into `otel_errors` is the **trace root span fallback**, not logs.

## Example `wrangler.jsonc`

```json
{
  "name": "my-worker",
  "main": "src/index.ts",
  "compatibility_date": "2026-04-23",
  "observability": {
    "traces": {
      "enabled": true,
      "destinations": ["strada-traces"],
      "persist": false
    },
    "logs": {
      "enabled": true,
      "destinations": ["strada-logs"],
      "persist": false
    }
  }
}
```

## Example OTLP payloads

Cloudflare documents destination export as standard **OTLP over HTTP** to `/v1/traces` and `/v1/logs` endpoints. The exact full payload shape for Cloudflare-generated uncaught exception logs is not fully documented, but the envelopes are standard OTLP.

## Example traces payload

This example is representative of the **shape** Strada should expect when Cloudflare exports trace data.

```json
{
  "resourceSpans": [
    {
      "resource": {
        "attributes": [
          { "key": "cloud.provider", "value": { "stringValue": "cloudflare" } },
          { "key": "cloud.platform", "value": { "stringValue": "cloudflare.workers" } },
          { "key": "service.name", "value": { "stringValue": "my-worker" } },
          { "key": "faas.name", "value": { "stringValue": "my-worker" } },
          { "key": "cloudflare.colo", "value": { "stringValue": "SFO" } }
        ]
      },
      "scopeSpans": [
        {
          "spans": [
            {
              "traceId": "4bf92f3577b34da6a3ce929d0e0e4736",
              "spanId": "00f067aa0ba902b7",
              "name": "fetch",
              "kind": 1,
              "startTimeUnixNano": "1713890000000000000",
              "endTimeUnixNano": "1713890000200000000",
              "attributes": [
                { "key": "faas.trigger", "value": { "stringValue": "http" } },
                { "key": "cloudflare.ray_id", "value": { "stringValue": "87d3abc123" } },
                { "key": "cloudflare.handler_type", "value": { "stringValue": "fetch" } },
                { "key": "cloudflare.outcome", "value": { "stringValue": "ok" } },
                { "key": "cloudflare.cpu_time_ms", "value": { "doubleValue": 12 } },
                { "key": "cloudflare.wall_time_ms", "value": { "doubleValue": 48 } },
                { "key": "url.full", "value": { "stringValue": "https://example.com/api/users?id=1" } },
                { "key": "url.path", "value": { "stringValue": "/api/users" } },
                { "key": "http.request.method", "value": { "stringValue": "GET" } },
                { "key": "http.response.status_code", "value": { "intValue": "200" } }
              ]
            }
          ]
        }
      ]
    }
  ]
}
```

## Example logs payload

This example reflects the **validated shape** seen from a real Cloudflare destination export. For runtime crashes, the exported OTLP logs contained request logs and app/framework log lines, but **not** standard `exception.*` attributes.

```json
{
  "resourceLogs": [
    {
      "resource": {
        "attributes": [
          { "key": "cloud.provider", "value": { "stringValue": "cloudflare" } },
          { "key": "cloud.platform", "value": { "stringValue": "cloudflare.workers" } },
          { "key": "service.name", "value": { "stringValue": "my-worker" } }
        ]
      },
      "scopeLogs": [
        {
          "logRecords": [
            {
              "timeUnixNano": "1713890000100000000",
              "severityText": "ERROR",
              "body": { "stringValue": "Spiceflow unhandled error: Error: uncaught cloudflare validation error" },
              "attributes": [
                { "key": "cloudflare.ray_id", "value": { "stringValue": "87d3abc123" } }
              ]
            }
          ]
        }
      ]
    }
  ]
}
```

## Querying strategy for agents

## Filter early

For Cloudflare data, start with:

```sql
WHERE ResourceAttributes['cloud.platform'] = 'cloudflare.workers'
```

Then narrow by one of:

* `SpanAttributes['cloudflare.handler_type']`
* `SpanAttributes['cloudflare.entrypoint']`
* `SpanAttributes['cloudflare.execution_model']`
* `SpanAttributes['cloudflare.outcome']`
* `SpanAttributes['url.path']`
* `SpanAttributes['http.request.method']`

## Good first questions

### Slow request

Filter by:

* `url.path`
* `http.request.method`
* `Duration`
* `cloudflare.wall_time_ms`
* `cloudflare.cpu_time_ms`

Then inspect child spans like D1, KV, R2, `fetch`, or Durable Object storage.

### Worker crash or uncaught exception

Filter by:

* `cloudflare.outcome = 'exception'`

Then inspect:

* same trace
* related logs with same `TraceId`
* exception fields on logs if present

### D1 slowness

Filter by:

* `db.system.name`
* `db.operation.name`
* `cloudflare.d1.response.sql_duration_ms`

### Durable Object RPC failures

Filter by:

* `cloudflare.jsrpc.method`
* `cloudflare.entrypoint`
* `cloudflare.outcome`

## Example queries

### Failed Worker invocations

```sql
SELECT
  Timestamp,
  ServiceName,
  SpanName,
  SpanAttributes['cloudflare.outcome'] AS outcome,
  SpanAttributes['cloudflare.handler_type'] AS handler_type,
  SpanAttributes['cloudflare.entrypoint'] AS entrypoint,
  SpanAttributes['cloudflare.ray_id'] AS ray_id
FROM otel_traces
WHERE ResourceAttributes['cloud.platform'] = 'cloudflare.workers'
  AND SpanAttributes['cloudflare.outcome'] != 'ok'
ORDER BY Timestamp DESC
LIMIT 100
```

### Slow requests

```sql
SELECT
  Timestamp,
  ServiceName,
  SpanName,
  Duration,
  SpanAttributes['http.request.method'] AS method,
  SpanAttributes['url.path'] AS path,
  SpanAttributes['cloudflare.cpu_time_ms'] AS cpu_ms,
  SpanAttributes['cloudflare.wall_time_ms'] AS wall_ms
FROM otel_traces
WHERE ResourceAttributes['cloud.platform'] = 'cloudflare.workers'
  AND SpanAttributes['cloudflare.handler_type'] = 'fetch'
ORDER BY Duration DESC
LIMIT 100
```

### Uncaught exception traces

```sql
SELECT
  Timestamp,
  ServiceName,
  SpanName,
  TraceId,
  SpanAttributes['cloudflare.handler_type'] AS handler_type,
  SpanAttributes['cloudflare.entrypoint'] AS entrypoint,
  SpanAttributes['cloudflare.ray_id'] AS ray_id
FROM otel_traces
WHERE ResourceAttributes['cloud.platform'] = 'cloudflare.workers'
  AND SpanAttributes['cloudflare.outcome'] = 'exception'
ORDER BY Timestamp DESC
LIMIT 100
```

### D1 bottlenecks

```sql
SELECT
  Timestamp,
  ServiceName,
  SpanName,
  SpanAttributes['db.operation.name'] AS operation,
  SpanAttributes['cloudflare.d1.response.sql_duration_ms'] AS sql_ms,
  SpanAttributes['cloudflare.d1.response.rows_read'] AS rows_read,
  SpanAttributes['cloudflare.d1.response.rows_written'] AS rows_written
FROM otel_traces
WHERE ResourceAttributes['cloud.platform'] = 'cloudflare.workers'
  AND SpanAttributes['cloudflare.binding.type'] = 'd1'
ORDER BY toFloat64OrZero(SpanAttributes['cloudflare.d1.response.sql_duration_ms']) DESC
LIMIT 100
```

### Logs with runtime-crash context

```sql
SELECT
  Timestamp,
  ServiceName,
  SeverityText,
  Body,
  TraceId,
  LogAttributes['cloudflare.ray_id'] AS ray_id
FROM otel_logs
WHERE ResourceAttributes['cloud.platform'] = 'cloudflare.workers'
  AND TraceId = '<trace-id-from-failing-root-span>'
ORDER BY Timestamp DESC
LIMIT 100
```

### Logs with standard exception fields if present

Some app code or SDKs may still emit normal `exception.*` fields. Query them separately when needed:

```sql
SELECT
  Timestamp,
  ServiceName,
  SeverityText,
  Body,
  TraceId,
  LogAttributes['exception.type'] AS exception_type,
  LogAttributes['exception.message'] AS exception_message
FROM otel_logs
WHERE ResourceAttributes['cloud.platform'] = 'cloudflare.workers'
  AND (
    mapContains(LogAttributes, 'exception.type')
    OR mapContains(LogAttributes, 'exception.message')
  )
ORDER BY Timestamp DESC
LIMIT 100
```

## Attribute reference

This section lists the documented Cloudflare Workers trace attributes with one attribute per section. Each entry explains:

* what it means
* why it matters
* common query usage

## Resource and invocation identity

### `cloud.provider`

**Meaning:** Cloud provider name. Cloudflare sets this to `cloudflare`.

**Why it matters:** Good coarse filter across multi-cloud telemetry.

**Common query use:**

```sql
ResourceAttributes['cloud.provider'] = 'cloudflare'
```

### `cloud.platform`

**Meaning:** Cloud runtime platform. Cloudflare sets this to `cloudflare.workers`.

**Why it matters:** Best top-level filter for Cloudflare Workers telemetry.

**Common query use:**

```sql
ResourceAttributes['cloud.platform'] = 'cloudflare.workers'
```

### `faas.name`

**Meaning:** Worker name in serverless naming.

**Why it matters:** Helps identify the function or Worker service.

**Common query use:** compare with `service.name` when normalizing service identity.

### `faas.invocation_id`

**Meaning:** Unique identifier for a specific Worker invocation.

**Why it matters:** Useful for correlating spans and logs from a single execution.

**Common query use:** drill into one specific invocation.

### `faas.version`

**Meaning:** Deployed version tag of the Worker.

**Why it matters:** Helpful for release correlation and regressions.

### `faas.invoked_region`

**Meaning:** Region where the Worker was invoked.

**Why it matters:** Useful for regional performance or routing issues.

### `service.name`

**Meaning:** Worker service name.

**Why it matters:** One of the most important grouping dimensions in Strada.

**Common query use:**

```sql
ServiceName = 'my-worker'
```

### `cloudflare.colo`

**Meaning:** Three-letter Cloudflare colo code like `SFO` or `LHR`.

**Why it matters:** Very useful for edge-specific debugging.

### `cloudflare.script_name`

**Meaning:** Worker script name.

**Why it matters:** Another service identity field, often similar to `service.name`.

### `cloudflare.script_tags`

**Meaning:** Tags associated with the Worker deployment.

**Why it matters:** Useful when deployments are grouped by internal tags.

### `cloudflare.script_version.id`

**Meaning:** Cloudflare version identifier of the deployed Worker.

**Why it matters:** Important for deploy regression analysis.

### `cloudflare.invocation.sequence.number`

**Meaning:** Ordering number used by Cloudflare when timestamps are identical.

**Why it matters:** Helps with precise ordering in logs and spans from the same invocation.

### `telemetry.sdk.language`

**Meaning:** SDK language. For Workers docs this is `javascript`.

**Why it matters:** Mostly informational.

### `telemetry.sdk.name`

**Meaning:** SDK name. Cloudflare docs say this is `cloudflare`.

**Why it matters:** Helpful to distinguish Cloudflare runtime-generated telemetry from app SDK telemetry.

## Root invocation outcome and lifecycle

### `faas.trigger`

**Meaning:** Invocation trigger like `http`, `cron`, `queue`, or `email`.

**Why it matters:** Important top-level classifier.

### `cloudflare.ray_id`

**Meaning:** Cloudflare request ID.

**Why it matters:** One of the best identifiers for debugging specific requests.

### `cloudflare.handler_type`

**Meaning:** Handler type like `fetch`, `scheduled`, `queue`, `email`, or `alarm`.

**Why it matters:** Very important for narrowing the kind of Worker execution.

### `cloudflare.entrypoint`

**Meaning:** Invoked entrypoint, for example a Durable Object class.

**Why it matters:** Important for Durable Object and RPC exploration.

### `cloudflare.execution_model`

**Meaning:** Execution model such as `stateless` or `stateful`.

**Why it matters:** Useful for distinguishing normal Workers from Durable Object-like execution.

### `cloudflare.outcome`

**Meaning:** Outcome of the invocation, such as `ok`, `exception`, `exceededCpu`, or `exceededMemory`.

**Why it matters:** One of the most important attributes in the whole document.

**Common query use:**

```sql
SpanAttributes['cloudflare.outcome'] = 'exception'
```

### `cloudflare.cpu_time_ms`

**Meaning:** CPU time used by the invocation in milliseconds.

**Why it matters:** Great for separating CPU-bound from IO-bound slowness.

### `cloudflare.wall_time_ms`

**Meaning:** Total wall time used by the invocation in milliseconds.

**Why it matters:** Great for slow-request analysis.

## HTTP and fetch request attributes

### `network.protocol.name`

**Meaning:** Network protocol name.

**Why it matters:** Usually minor, but useful for debugging protocol behavior.

### `network.protocol.version`

**Meaning:** Network protocol version.

**Why it matters:** Can help explain transport differences.

### `url.full`

**Meaning:** Full URL.

**Why it matters:** Useful for debugging individual requests.

**Caution:** Often high-cardinality.

### `url.scheme`

**Meaning:** URL scheme such as `https`.

**Why it matters:** Mostly informational.

### `url.path`

**Meaning:** Request path.

**Why it matters:** One of the best attributes for grouping endpoints.

### `url.query`

**Meaning:** Request query string.

**Why it matters:** Useful for debugging, but often high-cardinality.

### `server.port`

**Meaning:** Target server port.

**Why it matters:** Mostly relevant for outbound `fetch()` spans.

### `server.address`

**Meaning:** Target server address.

**Why it matters:** Useful for outbound dependency debugging.

### `user_agent.original`

**Meaning:** Raw user agent string.

**Why it matters:** Useful for browser or client analysis.

### `http.request.method`

**Meaning:** HTTP method like `GET` or `POST`.

**Why it matters:** One of the most common filters for request analysis.

### `http.request.header.content-type`

**Meaning:** Request content type header.

**Why it matters:** Useful for API debugging.

### `http.request.header.content-length`

**Meaning:** Request content length header.

**Why it matters:** Helpful for payload-size issues.

### `http.request.header.accept`

**Meaning:** Request `Accept` header.

**Why it matters:** Useful for content negotiation debugging.

### `http.request.header.accept-encoding`

**Meaning:** Request `Accept-Encoding` header.

**Why it matters:** Useful for compression-related issues.

### `http.request.body.size`

**Meaning:** Request body size.

**Why it matters:** Good for spotting unusually large inputs.

### `http.response.status_code`

**Meaning:** HTTP response status code.

**Why it matters:** Core success/failure signal.

### `http.response.body.size`

**Meaning:** Response body size.

**Why it matters:** Useful for latency and bandwidth analysis.

## Fetch handler enrichment

### `cloudflare.verified_bot_category`

**Meaning:** Verified bot category assigned by Cloudflare.

**Why it matters:** Useful for bot traffic analysis.

### `cloudflare.asn`

**Meaning:** Client ASN.

**Why it matters:** Useful for ISP-specific issues.

### `cloudflare.response.time_to_first_byte_ms`

**Meaning:** Time to first byte in milliseconds.

**Why it matters:** Very useful for end-user latency analysis.

### `geo.timezone`

**Meaning:** Client timezone.

**Why it matters:** Useful for geo slicing.

### `geo.continent.code`

**Meaning:** Client continent code.

**Why it matters:** Useful for regional traffic analysis.

### `geo.country.code`

**Meaning:** Client country code.

**Why it matters:** Very common geo filter.

### `geo.locality.name`

**Meaning:** Client locality name.

**Why it matters:** Useful for deep geo debugging.

### `geo.locality.region`

**Meaning:** Client locality region.

**Why it matters:** Useful for geo debugging.

### `user_agent.os.name`

**Meaning:** OS name parsed from the user agent.

**Why it matters:** Useful for client platform analysis.

### `user_agent.os.version`

**Meaning:** OS version.

**Why it matters:** Helpful for compatibility debugging.

### `user_agent.browser.name`

**Meaning:** Browser name.

**Why it matters:** Common browser grouping field.

### `user_agent.browser.major_version`

**Meaning:** Browser major version.

**Why it matters:** Good for coarse compatibility grouping.

### `user_agent.browser.version`

**Meaning:** Full browser version.

**Why it matters:** Good for precise compatibility debugging.

### `user_agent.engine.name`

**Meaning:** Browser engine name.

**Why it matters:** Can help explain browser behavior.

### `user_agent.engine.version`

**Meaning:** Browser engine version.

**Why it matters:** Mostly detailed compatibility context.

### `user_agent.device.type`

**Meaning:** Device type like `mobile` or `desktop`.

**Why it matters:** Very useful segmentation field.

### `user_agent.device.vendor`

**Meaning:** Device vendor.

**Why it matters:** Useful for client debugging.

### `user_agent.device.model`

**Meaning:** Device model.

**Why it matters:** Useful for rare device-specific issues.

### `http.request.header.accept-language`

**Meaning:** Request `Accept-Language` header.

**Why it matters:** Useful for localization debugging.

## Scheduled, queue, RPC, email, tail, alarm

### `faas.cron`

**Meaning:** Cron expression for scheduled invocation.

**Why it matters:** Useful for identifying scheduled jobs.

### `cloudflare.scheduled_time`

**Meaning:** Scheduled timestamp for scheduled or alarm execution.

**Why it matters:** Useful for cron and alarm debugging.

### `cloudflare.queue.name`

**Meaning:** Queue name.

**Why it matters:** Important for queue consumer analysis.

### `cloudflare.queue.batch_size`

**Meaning:** Queue batch size.

**Why it matters:** Good for throughput analysis.

### `cloudflare.jsrpc.method`

**Meaning:** RPC method name.

**Why it matters:** Very important for Durable Object RPC exploration.

### `cloudflare.email.from`

**Meaning:** Sender email.

**Why it matters:** Useful for email Worker debugging.

### `cloudflare.email.to`

**Meaning:** Recipient email.

**Why it matters:** Useful for email flow debugging.

### `cloudflare.email.size`

**Meaning:** Email size.

**Why it matters:** Useful for payload-size issues.

### `cloudflare.trace.count`

**Meaning:** Trace count on tail handler telemetry.

**Why it matters:** Mostly tail-specific context.

## Cache API

### `cache.request.url`

**Meaning:** Cache request URL.

**Why it matters:** Useful for cache key debugging.

### `cache.request.method`

**Meaning:** Cache request method.

**Why it matters:** Useful for cache behavior analysis.

### `cache.request.payload.status_code`

**Meaning:** Cached payload status code.

**Why it matters:** Helpful when debugging cache storage behavior.

### `cache.request.payload.header.cache_control`

**Meaning:** Cached response cache-control header.

**Why it matters:** Helpful for cache semantics.

### `cache.request.payload.header.cache_tag`

**Meaning:** Cached cache-tag header.

**Why it matters:** Useful for tag-based invalidation debugging.

### `cache.request.payload.header.etag`

**Meaning:** Cached ETag header.

**Why it matters:** Useful for validation flows.

### `cache.request.payload.header.expires`

**Meaning:** Cached expires header.

**Why it matters:** Useful for TTL behavior.

### `cache.request.payload.header.last_modified`

**Meaning:** Cached last-modified header.

**Why it matters:** Useful for conditional request debugging.

### `cache.request.payload.size`

**Meaning:** Cached payload size.

**Why it matters:** Helpful for storage-size debugging.

### `cache.response.success`

**Meaning:** Whether cache operation succeeded.

**Why it matters:** Good success/failure signal.

### `cache.request.ignore_method`

**Meaning:** Whether method was ignored when matching or deleting.

**Why it matters:** Useful for cache semantics.

### `cache.request.header.range`

**Meaning:** Range header used in cache match.

**Why it matters:** Useful for partial content debugging.

### `cache.request.header.if_modified_since`

**Meaning:** Conditional header for cache match.

**Why it matters:** Useful for cache validation analysis.

### `cache.request.header.if_none_match`

**Meaning:** Conditional ETag header for cache match.

**Why it matters:** Useful for cache validation analysis.

### `cache.response.body.size`

**Meaning:** Cache response body size.

**Why it matters:** Useful for payload-size debugging.

### `cache.response.cache_status`

**Meaning:** Cache status result.

**Why it matters:** Useful for hit/miss style analysis.

### `cache.response.status_code`

**Meaning:** Cache response status code.

**Why it matters:** Useful for cache debugging.

## D1

### `db.system.name`

**Meaning:** Database system identifier.

**Why it matters:** Standard database filter.

### `db.operation.name`

**Meaning:** Database operation name.

**Why it matters:** Important for query grouping.

### `db.query.text`

**Meaning:** SQL query text.

**Why it matters:** Extremely useful for diagnosing slowness.

**Caution:** Often sensitive and high-cardinality.

### `cloudflare.binding.type`

**Meaning:** Binding type.

**Why it matters:** Important for classifying spans like D1, KV, R2.

### `cloudflare.d1.response.size_after`

**Meaning:** Size after operation.

**Why it matters:** Useful for D1 mutation/debug context.

### `cloudflare.d1.response.rows_read`

**Meaning:** Number of rows read.

**Why it matters:** Very useful for expensive query analysis.

### `cloudflare.d1.response.rows_written`

**Meaning:** Number of rows written.

**Why it matters:** Useful for write cost analysis.

### `cloudflare.d1.response.last_row_id`

**Meaning:** Last inserted row id.

**Why it matters:** Mostly mutation context.

### `cloudflare.d1.response.changed_db`

**Meaning:** Whether the DB changed.

**Why it matters:** Mutation context.

### `cloudflare.d1.response.changes`

**Meaning:** Number of changes made.

**Why it matters:** Mutation analysis.

### `cloudflare.d1.response.served_by_region`

**Meaning:** Region serving the D1 response.

**Why it matters:** Useful for regional behavior.

### `cloudflare.d1.response.served_by_primary`

**Meaning:** Whether served by primary.

**Why it matters:** Useful for replication/topology debugging.

### `cloudflare.d1.response.sql_duration_ms`

**Meaning:** SQL duration in milliseconds.

**Why it matters:** One of the most important D1 attrs.

### `cloudflare.d1.response.total_attempts`

**Meaning:** Total attempts made.

**Why it matters:** Useful for retry visibility.

### `db.operation.batch.size`

**Meaning:** Batch size for `d1_batch`.

**Why it matters:** Important for batch-cost analysis.

### `cloudflare.d1.query.bookmark`

**Meaning:** Query bookmark sent with the request.

**Why it matters:** Useful for paging or replication state analysis.

### `cloudflare.d1.response.bookmark`

**Meaning:** Bookmark returned by the response.

**Why it matters:** Useful for paging or replication state analysis.

## KV

### `cloudflare.binding.name`

**Meaning:** Bound resource name.

**Why it matters:** Helpful for distinguishing namespaces or buckets.

### `cloudflare.kv.query.keys`

**Meaning:** KV key or keys used in the request.

**Why it matters:** Useful for debugging, but can be high-cardinality.

### `cloudflare.kv.query.keys.count`

**Meaning:** Number of keys queried.

**Why it matters:** Helpful for batch access analysis.

### `cloudflare.kv.query.type`

**Meaning:** KV query type.

**Why it matters:** Useful for read behavior analysis.

### `cloudflare.kv.query.cache_ttl`

**Meaning:** Cache TTL used for the query.

**Why it matters:** Useful for cache effectiveness debugging.

### `cloudflare.kv.response.size`

**Meaning:** Size of the returned value.

**Why it matters:** Helpful for payload analysis.

### `cloudflare.kv.response.returned_rows`

**Meaning:** Number of rows returned.

**Why it matters:** Useful for list/get behavior.

### `cloudflare.kv.response.metadata`

**Meaning:** Returned metadata.

**Why it matters:** Useful when metadata affects behavior.

### `cloudflare.kv.response.cache_status`

**Meaning:** Cache status.

**Why it matters:** One of the most useful KV attrs.

### `cloudflare.kv.query.value_type`

**Meaning:** Value type used for put.

**Why it matters:** Useful for write debugging.

### `cloudflare.kv.query.expiration`

**Meaning:** Absolute expiration.

**Why it matters:** TTL debugging.

### `cloudflare.kv.query.expiration_ttl`

**Meaning:** TTL duration.

**Why it matters:** TTL debugging.

### `cloudflare.kv.query.metadata`

**Meaning:** Metadata supplied on write.

**Why it matters:** Useful for write debugging.

### `cloudflare.kv.query.payload.size`

**Meaning:** Write payload size.

**Why it matters:** Useful for payload debugging.

### `cloudflare.kv.query.prefix`

**Meaning:** Prefix used in list.

**Why it matters:** Useful for list scans.

### `cloudflare.kv.query.limit`

**Meaning:** List limit.

**Why it matters:** Useful for paging analysis.

### `cloudflare.kv.query.cursor`

**Meaning:** List cursor.

**Why it matters:** Useful for paging analysis.

### `cloudflare.kv.response.list_complete`

**Meaning:** Whether list is complete.

**Why it matters:** Useful for paging logic.

### `cloudflare.kv.response.cursor`

**Meaning:** Response cursor.

**Why it matters:** Useful for paging logic.

### `cloudflare.kv.response.expiration`

**Meaning:** Returned expiration.

**Why it matters:** Useful for TTL visibility.

## R2

### `cloudflare.r2.bucket`

**Meaning:** R2 bucket name.

**Why it matters:** Core grouping field for R2.

### `cloudflare.r2.operation`

**Meaning:** R2 operation name.

**Why it matters:** One of the most useful R2 attrs.

### `cloudflare.r2.response.success`

**Meaning:** Whether the R2 operation succeeded.

**Why it matters:** Core success/failure signal.

### `cloudflare.r2.error.message`

**Meaning:** R2 error message.

**Why it matters:** Useful failure detail.

### `cloudflare.r2.error.code`

**Meaning:** R2 error code.

**Why it matters:** Useful failure detail.

### `cloudflare.r2.request.key`

**Meaning:** Object key.

**Why it matters:** Very useful for debugging, but high-cardinality.

### `cloudflare.r2.response.etag`

**Meaning:** Object ETag.

**Why it matters:** Useful for validation/debugging.

### `cloudflare.r2.response.size`

**Meaning:** Object size.

**Why it matters:** Important performance context.

### `cloudflare.r2.response.uploaded`

**Meaning:** Uploaded timestamp.

**Why it matters:** Useful object metadata context.

### `cloudflare.r2.response.checksum.value`

**Meaning:** Object checksum value.

**Why it matters:** Integrity/debug context.

### `cloudflare.r2.response.checksum.type`

**Meaning:** Checksum type.

**Why it matters:** Integrity/debug context.

### `cloudflare.r2.response.storage_class`

**Meaning:** Storage class.

**Why it matters:** Useful for storage policy analysis.

### `cloudflare.r2.response.ssec_key`

**Meaning:** SSE-C key indicator.

**Why it matters:** Encryption/debug context.

### `cloudflare.r2.response.content_type`

**Meaning:** Content type.

**Why it matters:** Useful for content debugging.

### `cloudflare.r2.response.content_encoding`

**Meaning:** Content encoding.

**Why it matters:** Useful for compression debugging.

### `cloudflare.r2.response.content_disposition`

**Meaning:** Content disposition.

**Why it matters:** Useful for download behavior debugging.

### `cloudflare.r2.response.content_language`

**Meaning:** Content language.

**Why it matters:** Useful for localization/content debugging.

### `cloudflare.r2.response.cache_control`

**Meaning:** Cache control metadata.

**Why it matters:** Useful for cache policy debugging.

### `cloudflare.r2.response.cache_expiry`

**Meaning:** Cache expiry metadata.

**Why it matters:** Useful for TTL/cache policy debugging.

### `cloudflare.r2.response.custom_metadata`

**Meaning:** Custom metadata.

**Why it matters:** Useful for app-specific object metadata.

### `cloudflare.r2.request.range.offset`

**Meaning:** Range offset.

**Why it matters:** Useful for partial-read debugging.

### `cloudflare.r2.request.range.length`

**Meaning:** Range length.

**Why it matters:** Useful for partial-read debugging.

### `cloudflare.r2.request.range.suffix`

**Meaning:** Range suffix.

**Why it matters:** Useful for partial-read debugging.

### `cloudflare.r2.request.range`

**Meaning:** Range descriptor.

**Why it matters:** Useful for partial-read debugging.

### `cloudflare.r2.request.ssec_key`

**Meaning:** SSE-C key request field.

**Why it matters:** Encryption/debug context.

### `cloudflare.r2.request.only_if.etag_matches`

**Meaning:** Conditional ETag match.

**Why it matters:** Conditional operation debugging.

### `cloudflare.r2.request.only_if.etag_does_not_match`

**Meaning:** Conditional ETag mismatch.

**Why it matters:** Conditional operation debugging.

### `cloudflare.r2.request.only_if.uploaded_before`

**Meaning:** Conditional upload time check.

**Why it matters:** Conditional operation debugging.

### `cloudflare.r2.request.only_if.uploaded_after`

**Meaning:** Conditional upload time check.

**Why it matters:** Conditional operation debugging.

### `cloudflare.r2.request.size`

**Meaning:** Request object size.

**Why it matters:** Helpful for upload analysis.

### `cloudflare.r2.request.checksum.type`

**Meaning:** Request checksum type.

**Why it matters:** Integrity/debug context.

### `cloudflare.r2.request.checksum.value`

**Meaning:** Request checksum value.

**Why it matters:** Integrity/debug context.

### `cloudflare.r2.request.custom_metadata`

**Meaning:** Custom metadata supplied on write.

**Why it matters:** Useful for application behavior.

### `cloudflare.r2.request.http_metadata.content_type`

**Meaning:** HTTP metadata content type on write.

**Why it matters:** Useful for upload debugging.

### `cloudflare.r2.request.http_metadata.content_encoding`

**Meaning:** HTTP metadata content encoding on write.

**Why it matters:** Useful for upload debugging.

### `cloudflare.r2.request.http_metadata.content_disposition`

**Meaning:** HTTP metadata content disposition on write.

**Why it matters:** Useful for upload debugging.

### `cloudflare.r2.request.http_metadata.content_language`

**Meaning:** HTTP metadata content language on write.

**Why it matters:** Useful for upload debugging.

### `cloudflare.r2.request.http_metadata.cache_control`

**Meaning:** HTTP metadata cache control on write.

**Why it matters:** Useful for cache policy debugging.

### `cloudflare.r2.request.http_metadata.cache_expiry`

**Meaning:** HTTP metadata cache expiry on write.

**Why it matters:** Useful for cache policy debugging.

### `cloudflare.r2.request.storage_class`

**Meaning:** Requested storage class.

**Why it matters:** Useful for storage policy analysis.

### `cloudflare.r2.request.limit`

**Meaning:** List limit.

**Why it matters:** Useful for paging analysis.

### `cloudflare.r2.request.prefix`

**Meaning:** List prefix.

**Why it matters:** Useful for bucket browsing analysis.

### `cloudflare.r2.request.cursor`

**Meaning:** List cursor.

**Why it matters:** Useful for paging analysis.

### `cloudflare.r2.request.delimiter`

**Meaning:** List delimiter.

**Why it matters:** Useful for pseudo-folder analysis.

### `cloudflare.r2.request.start_after`

**Meaning:** Start-after key.

**Why it matters:** Useful for paging analysis.

### `cloudflare.r2.request.include.http_metadata`

**Meaning:** Whether list includes HTTP metadata.

**Why it matters:** Useful for cost/shape understanding.

### `cloudflare.r2.request.include.custom_metadata`

**Meaning:** Whether list includes custom metadata.

**Why it matters:** Useful for cost/shape understanding.

### `cloudflare.r2.response.returned_objects`

**Meaning:** Number of returned objects.

**Why it matters:** Useful for list analysis.

### `cloudflare.r2.response.delimited_prefixes`

**Meaning:** Returned delimited prefixes.

**Why it matters:** Useful for pseudo-folder analysis.

### `cloudflare.r2.response.truncated`

**Meaning:** Whether list is truncated.

**Why it matters:** Useful for paging analysis.

### `cloudflare.r2.response.cursor`

**Meaning:** Response cursor.

**Why it matters:** Useful for paging analysis.

### `cloudflare.r2.request.keys`

**Meaning:** Keys used for delete.

**Why it matters:** Useful for delete debugging.

### `cloudflare.r2.response.upload_id`

**Meaning:** Multipart upload ID.

**Why it matters:** Important for multipart flow debugging.

### `cloudflare.r2.request.upload_id`

**Meaning:** Multipart upload ID on request.

**Why it matters:** Important for multipart flow debugging.

### `cloudflare.r2.request.part_number`

**Meaning:** Multipart part number.

**Why it matters:** Useful for multipart debugging.

### `cloudflare.r2.request.uploaded_parts`

**Meaning:** Uploaded parts summary.

**Why it matters:** Useful for multipart completion debugging.

## Durable Object storage SQL and KV

### `cloudflare.durable_object.query.bindings`

**Meaning:** Query bindings for DO SQL execution.

**Why it matters:** Useful for SQL debugging.

### `cloudflare.durable_object.response.rows_read`

**Meaning:** Rows read in DO SQL.

**Why it matters:** Useful for expensive-query analysis.

### `cloudflare.durable_object.response.rows_written`

**Meaning:** Rows written in DO SQL.

**Why it matters:** Useful for mutation analysis.

### `cloudflare.durable_object.response.db_size`

**Meaning:** DO embedded DB size.

**Why it matters:** Useful for storage growth analysis.

### `cloudflare.durable_object.kv.query.keys`

**Meaning:** DO KV keys queried.

**Why it matters:** Useful for debugging, but often high-cardinality.

### `cloudflare.durable_object.kv.query.keys.count`

**Meaning:** Count of DO KV keys.

**Why it matters:** Useful for batch analysis.

### `cloudflare.durable_object.kv.response.deleted_count`

**Meaning:** Deleted key count.

**Why it matters:** Useful for delete analysis.

### `cloudflare.durable_object.kv.query.start`

**Meaning:** Start key for list.

**Why it matters:** Useful for paging analysis.

### `cloudflare.durable_object.kv.query.startAfter`

**Meaning:** Start-after key for list.

**Why it matters:** Useful for paging analysis.

### `cloudflare.durable_object.kv.query.end`

**Meaning:** End key for list.

**Why it matters:** Useful for range scans.

### `cloudflare.durable_object.kv.query.prefix`

**Meaning:** Prefix for list.

**Why it matters:** Useful for namespaced scans.

### `cloudflare.durable_object.kv.query.reverse`

**Meaning:** Reverse-order listing flag.

**Why it matters:** Useful for paging/debugging.

### `cloudflare.durable_object.kv.query.limit`

**Meaning:** List limit.

**Why it matters:** Useful for paging analysis.

## Images

### `cloudflare.images.options.format`

**Meaning:** Requested output format.

**Why it matters:** Useful for image transform debugging.

### `cloudflare.images.options.quality`

**Meaning:** Requested quality.

**Why it matters:** Useful for performance vs quality tradeoffs.

### `cloudflare.images.options.background`

**Meaning:** Background option.

**Why it matters:** Useful for transformation debugging.

### `cloudflare.images.options.anim`

**Meaning:** Animation option.

**Why it matters:** Useful for animated image handling.

### `cloudflare.images.options.transforms`

**Meaning:** Image transforms.

**Why it matters:** Useful for understanding expensive transforms.

### `cloudflare.images.error.code`

**Meaning:** Image processing error code.

**Why it matters:** Useful failure detail.

### `cloudflare.images.options.encoding`

**Meaning:** Input encoding option.

**Why it matters:** Useful for transformation debugging.

### `cloudflare.images.result.format`

**Meaning:** Result format.

**Why it matters:** Useful for output analysis.

### `cloudflare.images.result.file_size`

**Meaning:** Output file size.

**Why it matters:** Useful for payload/performance analysis.

### `cloudflare.images.result.width`

**Meaning:** Output width.

**Why it matters:** Useful for transform verification.

### `cloudflare.images.result.height`

**Meaning:** Output height.

**Why it matters:** Useful for transform verification.

## Important gaps and caveats

## Cloudflare logs are less fully documented than traces

Cloudflare clearly documents the trace attributes above.

Cloudflare also documents that logs include:

* `console.log()` output
* system-generated logs
* uncaught exceptions in Workers logs products

Validation against a real Worker export confirmed that destination OTLP logs did **not** expose a structured uncaught-exception payload with fields like:

* `exception.type`
* `exception.message`
* `exception.stacktrace`

So agents should use this rule:

* **for failures, trust traces first** via `cloudflare.outcome`
* **for detailed error grouping, inspect logs second** for surrounding context, but do not assume `exception.*` exists

## Recommended mental model for agents

When investigating Cloudflare Worker problems in Strada:

1. **Filter to Cloudflare** with `cloud.platform = 'cloudflare.workers'`
2. **Find the root span** and inspect:
   * `cloudflare.handler_type`
   * `cloudflare.entrypoint`
   * `cloudflare.outcome`
   * `cloudflare.cpu_time_ms`
   * `cloudflare.wall_time_ms`
3. **Inspect child spans** for D1, KV, R2, `fetch`, or Durable Object work
4. **If the request failed**, search logs on the same trace for `exception.*`
5. **If the problem is slowness**, compare total duration vs CPU vs child-span durations

## Best docs to cross-check

1. **Exporting OpenTelemetry Data**\
   [https://developers.cloudflare.com/workers/observability/exporting-opentelemetry-data/](https://developers.cloudflare.com/workers/observability/exporting-opentelemetry-data/)
2. **Traces**\
   [https://developers.cloudflare.com/workers/observability/traces/](https://developers.cloudflare.com/workers/observability/traces/)
3. **Spans and attributes**\
   [https://developers.cloudflare.com/workers/observability/traces/spans-and-attributes/](https://developers.cloudflare.com/workers/observability/traces/spans-and-attributes/)
4. **Workers Logs**\
   [https://developers.cloudflare.com/workers/observability/logs/workers-logs/](https://developers.cloudflare.com/workers/observability/logs/workers-logs/)
5. **Workers Logpush dataset docs**\
   [https://developers.cloudflare.com/logs/logpush/logpush-job/datasets/account/workers\_trace\_events/](https://developers.cloudflare.com/logs/logpush/logpush-job/datasets/account/workers_trace_events/)
6. **Export to Sentry**\
   [https://developers.cloudflare.com/workers/observability/exporting-opentelemetry-data/sentry/](https://developers.cloudflare.com/workers/observability/exporting-opentelemetry-data/sentry/)
7. **Export to Grafana Cloud**\
   [https://developers.cloudflare.com/workers/observability/exporting-opentelemetry-data/grafana-cloud/](https://developers.cloudflare.com/workers/observability/exporting-opentelemetry-data/grafana-cloud/)
8. **Export to Axiom**\
   [https://developers.cloudflare.com/workers/observability/exporting-opentelemetry-data/axiom/](https://developers.cloudflare.com/workers/observability/exporting-opentelemetry-data/axiom/)


---

*Powered by [holocron.so](https://holocron.so)*