4x sqlalchemy pool size

[yunfeng-scale]

Getting errors like

  File "/workspace/model-engine/model_engine_server/infra/repositories/db_model_endpoint_record_repository.py", line 236, in get_model_endpoint_record
    model_endpoint_orm = await OrmModelEndpoint.select_by_id(
  File "/workspace/model-engine/model_engine_server/db/models/hosted_model_inference.py", line 589, in select_by_id
    endpoint = await session.execute(
.......
sqlalchemy.exc.TimeoutError: QueuePool limit of size 5 overflow 5 reached, connection timed out, timeout 30.00 (Background on this error at: https://sqlalche.me/e/20/3o7r)

apparently there's some traffic spike(?) that caused too many connections to the postgres connection pool

Pull Request Summary

What is this PR changing? Why is this change being made? Any caveats you'd like to highlight? Link any relevant documents, links, or screenshots here if applicable.

Test Plan and Usage Guide

How did you validate that your PR works correctly? How do you run or demo the code? Provide enough detail so a reviewer can reasonably reproduce the testing procedure. Paste example command line invocations if applicable.