Skip to content

When writing a batch to an implicit bucket upon schema conflict, I receive a 400. The logs seems to be f-printed. #693

Open
@BernardClabots

Description

@BernardClabots

Specifications

  • Client Version: 1.40.0
  • InfluxDB Version: Cloud 2.0 TSM
  • Platform: GCP

I am extracting some data from BQ and pushing it to Influx. That part is working well.
Next, I write a point in an Influx bucket to mark the task completed. This was written by someone else and it uses a different call.
In this particular case, I have a conflict for the second parameter. In the GCP logs, I can see the error as "default" on top of the log generated by loguru (the log appears twice).

Code sample to reproduce problem

in the calling function

    points = [
        Point.from_dict(
            {
                "measurement": __DATA_MEASUREMENT,
                "fields": {state: value},
                "tags": {"type": "status", **tags},
            }
        ),
        Point.from_dict(
            {
                "measurement": __DATA_MEASUREMENT,
                "fields": {state: datetime.now().isoformat()},
                "tags": {"type": "timestamp", **tags},
            }
        ),
    ]

    write_points(
        __INFLUX_URL,
        __INFLUX_TOKEN,
        __INFLUX_BUCKET,
        __INFLUX_ORGANISATION,
        points,
    )
```
in the lib to write_points:
``````python
def write_points(
    influx_url: str,
    influx_token: str,
    influx_bucket: str,
    influx_organization: str,
    points: List[Union[Point, Dict[str, Any]]],
) -> None:
    logger.info("Writing to InfluxDB, Points length={}", len(points))
    client = InfluxDBClient(url=influx_url, token=influx_token, timeout=300_000)
    try:
        write_api = client.write_api(
            success_callback=callback.success,
            error_callback=callback.error,
            retry_callback=callback.retry,
            write_options=WriteOptions(max_retries=10, max_retry_time=280_000)
        )
        
        try:
            write_api.write(
                batch_size=200,
                bucket=influx_bucket,
                org=influx_organization,
                record=points
            )
            logger.debug("Write operation completed")
            
        except InfluxDBError as e:
            logger.error("InfluxDB Asynchronous write error: {}", str(e))
            if hasattr(e, 'response'):
                logger.error("Response status: {}", e.response.status_code)
                logger.error("Response headers: {}", dict(e.response.headers))
                logger.error("Response body: {}", e.response.text)
            raise

        except Exception as ex:
            logger.error("Unexpected error during write operation: {}", str(ex))
            raise

    finally:
        client.close()

```
I know this is not correctly configured to do SYNCHRONOUS write, but that is how I see the issue.
The callback object is just implementing calls to logger.{info|error|warning} resp.

### Expected behavior

The logs are formatted as per my loguru logger settings

### Actual behavior

some logs (error 400 schema conflict) appear twice.

### Additional info

(DEBUG) 2025-04-25 15:15:06.130 CEST
Write operation completed
(DEFAULT) 2025-04-25 15:15:10.914 CEST
The batch item wasn't processed successfully because: (400)
(DEFAULT) 2025-04-25 15:15:10.914 CEST
Reason: Bad Request
(DEFAULT) 2025-04-25 15:15:10.914 CEST
HTTP response headers: HTTPHeaderDict({'Date': 'Fri, 25 Apr 2025 13:15:09 GMT', 'Content-Type': 'application/json', 'Content-Length': '247', 'Connection': 'keep-alive', 'influxdata-partial-write': 'true', 'trace-id': 'xxxxxxxxxxxxxxxx', 'trace-sampled': 'false', 'Strict-Transport-Security': 'max-age=31536000; includeSubDomains', 'X-Influxdb-Request-ID': 'zzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz', 'X-Influxdb-Build': 'Cloud'})
(DEFAULT) 2025-04-25 15:15:10.914 CEST
HTTP response body: {"code":"invalid","message":"partial write has occurred, errors encountered on line(s): line 2: batch schema conflict - line has column 'inserted_data' of type iox::column_type::field::string but type is iox::column_type::field::integer","line":2}
(ERROR) 2025-04-25 15:15:10.915 CEST
Cannot write batch: ('{bucket_name}', '{org_name}', 'ns'), data: b'GCP_Data_monitoring,app=myapp,stage=preprod,type=status,zone=eu-west-01 inserted_data=1i\nGCP_Data_monitoring,app=myapp,stage=preprod,type=timestamp,zone=eu-west-01 inserted_data="2025-04-25T13:15:06.127660"' due: (400) Reason: Bad Request HTTP response headers: HTTPHeaderDict({'Date': 'Fri, 25 Apr 2025 13:15:09 GMT', 'Content-Type': 'application/json', 'Content-Length': '247', 'Connection': 'keep-alive', 'influxdata-partial-write': 'true', 'trace-id': 'xxxxxxxxxxxxxxxx', 'trace-sampled': 'false', 'Strict-Transport-Security': 'max-age=31536000; includeSubDomains', 'X-Influxdb-Request-ID': 'zzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz', 'X-Influxdb-Build': 'Cloud'}) HTTP response body: {"code":"invalid","message":"partial write has occurred, errors encountered on line(s): line 2: batch schema conflict - line has column 'inserted_data' of type iox::column_type::field::string but type is iox::column_type::field::integer","line":2}

Metadata

Metadata

Assignees

No one assigned

    Labels

    bugSomething isn't working

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions