ERR_INCOMPLETE_CHUNKED_ENCODING/curl: (18) transfer closed with outstanding read data remaining

[javier]

Steps to reproduce:
List the minimal actions needed to reproduce the behaviour.

1. Start influxDB 3 Core Alpha and start via influxdb3 serve --node-id=local01 --object-store=file --data-dir ~/.influxdb/data/node0/ --http-bind=0.0.0.0:8181
2. Load ~34 million rows (I used the TSBS benchmark to generate a day of data, as in tsbs_generate_data --use-case="cpu-only" --seed=123 --scale=4000 --timestamp-start="2016-01-01T00:00:00Z" --timestamp-end="2016-01-02T00:00:00Z" --log-interval="10s" --format="influx"), then ingested via python script.
3. Execute a query via curl as in curl -v "http://localhost:8181/api/v3/query_sql?db=sensors&q=select%20*%20from%20cpu%20limit%205" or directly on chrome as in http://localhost:8181/api/v3/query_sql?db=sensors&q=select%20*%20from%20cpu%20limit%205

Expected behaviour:
Query should return

Actual behaviour:
Curl errors with curl: (18) transfer closed with outstanding read data remaining
Chrome errors with ERR_INCOMPLETE_CHUNKED_ENCODING
A python script using the influx client or using the influxdb3 query command return just fine
Describe What actually happened.

Environment info:

• Installation as per the docs at https://docs.influxdata.com/influxdb3/core/get-started/
• System info: Linux 6.8.0-1021-aws x86_64
• AWS m6a.4xlarge instance

[hiltontj]

Hey @javier, thanks for opening the issue. Would it be possible to provide an excerpt from the server logs at the time of querying?

Specifically, I would like to see if the query handler is throwing a panic and interrupting the response stream. Or, any other related errors may be informative.