An insurance provider is implementing Anypoint platform to manage its application infrastructure and is using the customer hosted runtime for its business due to certain financial requirements it must meet. It has built a number of synchronous API's and is currently hosting these on a mule runtime on one server

These applications make use of a number of components including heavy use of object stores and VM queues.

Business has grown rapidly in the last year and the insurance provider is starting to receive reports of reliability issues from its applications.

The DevOps team indicates that the API's are currently handling too many requests and this is over loading the server. The team has also mentioned that there is a significant downtime when the server is down for maintenance.

As an integration architect, which option would you suggest to mitigate these issues?