fix: add read_timeout to connection pool config to respect STARROCKS_QUERY_TIMEOUT env var

[Rupesh-kumar2]

Summary

Fixes a bug where STARROCKS_QUERY_TIMEOUT environment variable was silently
ignored, causing long-running queries to fail with a ~30s timeout regardless
of what was configured.

Root Cause

The MySQL connector distinguishes between two types of timeouts:

Timeout	Purpose	Was Configurable?
connect_timeout	Time to establish a connection	✅ Yes (via STARROCKS_CONNECTION_TIMEOUT)
read_timeout	Time to wait for query results	❌ No — this PR fixes it

The STARROCKS_QUERY_TIMEOUT env var existed in documentation/configs
but was never wired into the connection pool, so the MySQL connector
fell back to its internal default (~30s), causing timeouts on analytical
queries over large datasets.

Fix

Add read_timeout to the connection pool config, reading from the
STARROCKS_QUERY_TIMEOUT environment variable:

'read_timeout': int(os.getenv('STARROCKS_QUERY_TIMEOUT', '600')),