# KafkaJSDeleteTopicRecordsErrorPartition

**Kind:** Interface

**Source:** [`packages/microservices/external/kafka.interface.ts`](https://github.com/nestjs/nest/blob/master/packages/microservices/external/kafka.interface.ts#L1276)

**Part of:** [Microservices](subsystem-packages-microservices)

`KafkaJSDeleteTopicRecordsErrorPartition` describes a failed partition result returned while deleting Kafka topic records. It identifies the affected `partition`, the requested deletion `offset`, and the underlying `KafkaJSError` so callers can inspect and handle partition-specific failures.

## Properties

| Property | Type |
|---|---|
| `partition` | `number` |
| `offset` | `string` |
| `error` | `KafkaJSError` |

## Diagram

```mermaid
graph LR
  Request[Delete topic records request] --> PartitionResult[KafkaJSDeleteTopicRecordsErrorPartition]
  PartitionResult --> Partition[partition: number]
  PartitionResult --> Offset[offset: string]
  PartitionResult --> Error[error: KafkaJSError]
  Error --> Handling[Retry, log, or surface failure]
```

## Usage

```ts
import type { KafkaJSError } from 'kafkajs';
import type { KafkaJSDeleteTopicRecordsErrorPartition } from './kafka.interface';

function handleDeleteRecordsFailure(
  failure: KafkaJSDeleteTopicRecordsErrorPartition,
) {
  console.error(
    `Unable to delete records through offset ${failure.offset} ` +
      `for partition ${failure.partition}: ${failure.error.message}`,
  );

  // Use the partition and offset to retry or report the failed operation.
  return {
    partition: failure.partition,
    offset: failure.offset,
    retryable: failure.error.retriable,
  };
}

const failure: KafkaJSDeleteTopicRecordsErrorPartition = {
  partition: 2,
  offset: '1840',
  error: new Error('Broker unavailable') as KafkaJSError,
};

handleDeleteRecordsFailure(failure);
```

## AI Coding Instructions

- Treat `offset` as a string; Kafka offsets may exceed JavaScript's safe integer range.
- Preserve the original `KafkaJSError` rather than replacing it with a generic error, since it contains Kafka-specific diagnostics and retry metadata.
- Handle failures per partition instead of assuming a delete-records operation succeeds or fails for the entire topic.
- Include the partition and offset in logs, metrics, and retry context to make partial failures traceable.