I have the below queries related to the one-agent installation and its access to platforms.
1. When my application is running on multiple hosts (Let's assume multiple physical servers), how will installing a single agent on a single host help analyze the entire environment?
2. Do we need to install the agent on every host part of the entire application?
3. Installing a single agent on a host part of a single application can give me access to all other applications running on my environment?
4. If yes, how is that possible since it requires authentication and authorization each time the agent scan the environment?
1. Depends on the role of the servers. If all are e.g. load balanced application servers you only see a % of the total traffic. If you install only on application servers not the database for example, you will not miss to much at the start.
2. Again depends a bit on the role of the servers. I'd install it on all machines running application/web servers. You can install a agent in Infrastructure Only mode on severs like the database as well, to get more insights into infrastructure health or use plugins.
3. You will see only information about processes/services running on that host. You will see and get response times / error rates for external calls, but no deep insides if there is no agent on the other side.
4. See 3, agent only injects application on the server where it is running.