500-430 Practice Questions

For centralized agent upgrades, the new agent software (AppServerAgent.zip) must be placed in a specific directory on the Controller host: /appserver/glassfish/domains/domain1/files/. When a node with an older agent version connects, the Controller will instruct it to download and apply the new version upon the next restart of the application's JVM. This allows for a centrally managed, phased rollout of agent updates.

To capture slow queries and their execution plans, the element must be set to true within the collector's configuration in the Database Agent's config.xml. This tells the agent to actively monitor for long-running queries and, when detected, to request the execution plan (EXPLAIN) from the database. The other options enable different features but are not the primary switch for capturing slow query data.

All AppDynamics REST API calls require authentication. The client must authenticate as a user with sufficient permissions. This is typically done using Basic Authentication, where the username is formatted as user@account. A bearer token acquired from the Controller is also a common method. A missing or incorrect Authorization header is the most frequent cause of API call failures.

Standard Browser RUM is designed for traditional multi-page websites. For Single-Page Applications (SPAs), which don't perform full page reloads, you must enable the specific SPA2 monitoring feature within the Browser RUM configuration in the Controller UI. This enables the JavaScript agent to monitor virtual pages, route changes, and the AJAX calls associated with them, providing the detailed timing data needed to diagnose issues like a slow-loading checkout view within the SPA.

In an HA pair, the secondary Controller's database acts as a replica (slave) of the primary Controller's database. The 'Replication Slave Host' setting is used during the secondary Controller's installation to specify the address of the primary database server. This allows the secondary database to connect to the primary and begin replicating data, which is the core mechanism for keeping the passive node ready for a failover.

The AppDynamics Controller runs on a GlassFish application server. The network listeners for GlassFish are configured in the domain.xml file. To restrict the Controller to a specific IP address, you must modify the address attribute of the relevant element (e.g., for http-listener-1) from the default 0.0.0.0 to the desired IP address. db.cnf is for the MySQL database, controller.sh is for startup scripts, and app-agent-config.xml is for agent configuration.

The AppDynamics REST API for metrics requires a specific, pipe-delimited path. For business transaction metrics, the path starts with 'Business Transaction Performance', followed by 'Business Transactions', the tier name (or application name if it's an overall BT), the BT name, and finally the specific metric name. The other options use incorrect delimiters or path structures.

For globally distributed users, the SaaS EUM service is the optimal choice. It provides geographically distributed collection points for EUM beacons, ensuring low latency for end-users regardless of their location. This removes the complexity of managing a distributed on-premises EUM infrastructure. A single on-prem server would introduce high latency for distant users. A geo-distributed on-prem deployment is possible but much more complex to manage than using the SaaS offering.

By default, the .NET Agent instruments IIS applications. To monitor standalone applications or Windows Services, you must explicitly enable them using the AppDynamics Agent Configuration utility. After specifying the services to instrument, those services must be restarted for the .NET profiler to attach to their processes. The installer typically sets the environment variables, and restarting the Coordinator service alone is insufficient.

This statement is true. A rolling upgrade of the Events Service is the recommended procedure to avoid downtime. However, running a mixed-version cluster for an extended period is not supported. The entire upgrade process for all nodes in the cluster should be completed within a single, defined maintenance window to ensure cluster stability and data consistency.