Random destinations fail after a trunk change, and both HSRP routers are active

Users in VLAN 10 (10.10.10.0/24) report that some websites and internal apps fail while others work. Which ones fail seems random, and it affects users on both access switches. R1 (10.10.10.2, priority 110) and R2 (10.10.10.3, priority 100) run HSRP version 2, group 10, virtual IP 10.10.10.1, both with preempt. Each router connects to its own switch, and SW1 and SW2 are joined by a trunk. Last night a change added VLAN 30 to that trunk.