An application runs on AWS with a MySQL database on an EC2 instance in a VPC with two private subnets, and a website on Apache Tomcat in a single EC2 instance in a different VPC with one public subnet. There is a VPC peering connection between the database and website VPCs. The website recently experienced multiple outages from high traffic. Which actions should a solutions architect take to improve application reliability? (Choose three.)

To improve reliability, the Tomcat server should be placed in an Auto Scaling group with multiple instances behind an Application Load Balancer. This distributes traffic, handles increased load, and replaces unhealthy instances, preventing outages from high traffic. Migrating the MySQL database to Amazon Aurora with an Aurora Replica enhances database availability and read scalability. Aurora is highly available and fault-tolerant, and replicas provide failover and offload read traffic. Creating an additional public subnet in another Availability Zone for the website VPC allows the Auto Scaling group to deploy instances across multiple AZs, increasing fault tolerance and high availability. An additional VPC peering connection is unnecessary as one connection already exists. Provisioning two NAT gateways in the database VPC is irrelevant to improving application reliability for the website or database. Moving the Tomcat server into the database VPC would combine concerns and not inherently improve reliability; it would still require scaling and high availability measures.