Capturing traces from multiple applications to a single TempoStack instance, then while querying those traces for a longer duration (say 6 hours) with trace count of 20 or more, it results in 504 Gateway Timeout

Tempo Operator 0.15.4-1 

100%

1. Install and configure multi tentant TempoStack
2. Setup OpenTelemetryCollector to collect the traces.
3. Create multiple namespaces (I created 25)
4. Deploy Instrumentation and Applications in above created namespaces.
5 Wait for sometime and let TempoStack capture some traces (say 3-4 hours)
6. Try to query traces (for 1 service) for 6 or more hours of duration and it results in 504 gateway timeout

504 Gateway Timeout is seen when querying traces for a longer duration and limit results to 20.

Tempo should present the traces without getting timed out.

The test was carried out with AWS S3 as object storage.
Configured 2 vCPU and 5 Gi of memory for querier and query-frontend pods.
Neither of querier and query-frontend pods got OOMKilled.

---
Gateway streams below errors when query fails:

level=warn name=observatorium ts=2025-05-29T19:57:37.387386306Z caller=stdlib.go:105 caller=reverseproxy.go:661 msg="http: proxy error: context deadline exceeded"
level=warn name=observatorium ts=2025-05-29T19:57:37.387439501Z caller=instrumentation.go:33 request=tempo-simplest-gateway-7d4c694d98-l4ft4/lKPOKWFDP3-000164 proto=HTTP/1.1 method=GET status=502 content= path=/api/traces/v1/dev/api/traces duration=30.000367897s bytes=0

---
The issue seems to be either at query-frontend or querier pod side.