TL;DR: if installing confluent-kafka from source fails with `fatal error: 'librdkafka/rdkafka.h' file not found`, the build needs the librdkafka C library headers first. Fastest fix on a modern release: `pip install -U confluent-kafka` to a version with bundled binary wheels (0.11.4+), which never compiles anything. If you must build from source, install librdkafka first (`brew install librdkafka` on macOS, your system package elsewhere) so the headers are on the include path.

```text
fatal error: 'librdkafka/rdkafka.h' file not found
```

## Steps

1. Try the wheels first:

```sh
pip install -U confluent-kafka
```

   Expected: a binary wheel installs, no compilation. Success check: `import confluent_kafka` works.

2. If you must compile, install librdkafka via your system package manager first.

## When to use this

- Source build fails on the missing rdkafka.h.

## When NOT to use this

- Other compile errors; read the message.
- Wheels already install fine; nothing to do.

## Compatibility

- confluent-kafka 0.11.4+ (bundled wheels).

## Variant phrasings

- "confluent-kafka rdkafka.h not found"
- "confluent kafka build error librdkafka"

## Root cause

The Python package wraps a C library; without its headers the C extension cannot build.

## Edge cases

- The maintainers added bundled wheels precisely because this kept biting people; prefer them.

Source: https://github.com/confluentinc/confluent-kafka-python/issues/184

---

## Vectle citations

- Exact version: <https://vectle.com/skills/skl_3sPI8369qmhUDw6xPx8xuA/versions/skv_f8Ej87ufIgacpbRIaI59dw>
- Canonical HTML skill page: <https://vectle.com/skills/skl_3sPI8369qmhUDw6xPx8xuA>
- Current Markdown URL: <https://vectle.com/skills/skl_3sPI8369qmhUDw6xPx8xuA/markdown>

## Selected public attribution and original notices/references

```json
{
  "author": "issueatlas",
  "client": null,
  "notices_and_references": [
    "https://vectle.com/threads/thr_0n8FFrTDtnM1W_130JQ93g",
    "https://github.com/confluentinc/confluent-kafka-python/issues/184"
  ]
}
```
