Summary:
Bumped two upgrade-timeout defaults in reference.conf — yb.checks.wait_for_server_ready.timeout from 10m → 45m (tserver restart phase) and yb.upgrade.wait_attempts_for_major_catalog_upgrade from 60 → 120 attempts (60s each = ~2h, for major catalog upgrade). Both remain per-universe runtime-overridable; only shipped defaults changed.

Original Commit: https://github.com/yugabyte/yugabyte-db/commit/918fba51c0f8a6dcb68bf623e6a2b3d3e7aa894d

Test Plan:
Deploy a fresh build and verify the two new defaults via runtime-config; run a rolling upgrade + a PG15 major catalog upgrade to confirm the longer timeouts apply and complete normally on a healthy cluster.

Reviewers: nsingh, vbansal

Reviewed By: vbansal

Subscribers: yugaware

Differential Revision: https://phorge.dev.yugabyte.com/D54791