Databend in EKS -- Need help for no autoscaling & unclear errors

[isyanap94-funnel]

We’re self-hosting Databend Community Edition on AWS EKS, with the cluster set to auto-scale. Since yesterday at 6 PM, we've been running pipelines multiple times a day through Dagster. However, we've been experiencing query execution failures specifically between 6 PM - 1 AM. After 1 AM, all pipelines have been running successfully without any change

From the metrics, the cluster remained intact and stable during this period, with multiple nodes active at all times. The failures eventually resolved after retries, but the root cause is unclear because the error messages lack detailed information. Some examples:

raise RemoteDisconnected("Remote end closed connection without"
http.client.RemoteDisconnected: Remote end closed connection without response

Query failed: QueryErrors{code=1038, message=h2 protocol error: error reading a body from connection, source: Some(hyper::Error(Body, Error { kind: Io(Custom { kind: BrokenPipe, error: "stream closed because of a broken pipe" }) }))} cause: null

QueryErrors{code=404, message=query id 07760f59-a989-4af6-873c-c3ff99f160e9 not found on fOyX46MDahceCcordWb027} cause: null

1. Is there any way for us to get more detailed error messages? There’s nothing useful in the server logs either
2. If the issue were due to insufficient nodes, we would expect an auto-scaling request -- but it didn’t happen between 6 PM - 1 AM (please see the attached screenshot). Could there be some timeout or connection setting on Databend's side that we should take a closer look at?

TIA!