AppDynamics CAAA common mistakes
AppDynamics CAAA · Understand the traps and how to avoid them.
Treating every business transaction as automatically appearing once an agent is installed, rather than defining and configuring the entry points (servlets, POJOs, web services) that should start a transaction.
Candidates conflate automatic discovery with automatic business transaction definition, so they answer questions about BT detection as if no configuration is needed.
What to do instead
Study the distinction between automatic entry points (e.g., servlets, JAX-RS resources) and custom entry points/POJOs, and practice identifying which tier and method must be configured to produce a named business transaction.
Assuming all snapshots are always captured and stored indefinitely, ignoring transaction snapshot collection policies, sampling, and the difference between slow, very slow, and error snapshots.
Snapshot collection is a frequently tested configuration area, but candidates skim the policy settings and default to 'all transactions are captured'.
What to do instead
Learn the snapshot collection policy options (slow, very slow, error, stall), how thresholds interact with BT configuration, and how the number of snapshots per minute is bounded.
Confusing health rules with policies and actions when describing an alerting workflow.
The three-part alerting pipeline is easy to memorize superficially, so candidates mix up which object holds the condition versus which object defines the notification.
What to do instead
Map the pipeline explicitly: health rule defines the condition, policy binds the health rule to an action, and action defines the notification channel; then trace an example end to end.
Answering about security enforcement as if it is enabled by default or fully agent-side, rather than recognizing the controller-driven policy and the agent's role in blocking.
Application Security spans controller configuration, agent behavior, and policy modes, and candidates learn only one layer.
What to do instead
Study the security architecture: where policies are defined, how agents receive them, the difference between detection and enforcement modes, and what happens when an agent cannot reach the controller.
Treating security vulnerabilities and business risks as the same thing or assuming all detected issues are blocked automatically.
The terminology overlaps with general AppSec concepts, so candidates import assumptions from other tools.
What to do instead
Learn the specific categories the platform reports (e.g., attack detection versus code-level vulnerability findings), how each is surfaced, and which require remediation versus runtime blocking.
Forgetting that SAP monitoring requires specific agent deployment and configuration distinct from standard Java/.NET agents, and answering as if a single generic agent covers SAP ABAP and SAP web applications.
Candidates generalize from the mainstream APM agent workflow and overlook SAP-specific collectors, transports, and configuration steps.
What to do instead
Separate the SAP monitoring components in your notes: what monitors ABAP code, what monitors the SAP web/Java stack, and what configuration each requires before data appears in the controller.
Mixing up synthetic monitoring and real user monitoring when asked how to detect availability problems versus how to measure actual end-user experience.
Both live under digital experience monitoring, so candidates blur which one measures real users and which measures scripted checks.
What to do instead
Build a comparison table: synthetic (scripted, scheduled, location-based, availability and baseline) versus RUM (real sessions, browser and device data, JavaScript agent injection), and match each to the question's intent.
Assuming the JavaScript agent is injected automatically with no configuration, or ignoring session and page-level configuration needed for RUM data to appear.
Candidates treat browser monitoring as plug-and-play and skip the injection and configuration details.
What to do instead
Learn how the JavaScript agent is injected (automatic versus manual), what configuration controls sampling and session capture, and how to verify data is flowing.
Confusing machine agent, infrastructure agent, and server monitoring capabilities, and expecting the same metrics from every agent type.
Infrastructure monitoring has several agent and extension paths, and candidates collapse them into one mental model.
What to do instead
Distinguish what the machine agent collects versus what the infrastructure visibility/extension approach collects, and note which metrics (CPU, memory, disk, network) come from which source.
Treating business analytics as a passive dashboard rather than understanding that it requires defining business transactions, data collectors, and analytics configuration to capture and query business-relevant fields.
Candidates assume business metrics appear automatically from APM data and skip the configuration chain that makes fields searchable.
What to do instead
Trace the path: business transaction definition, data collectors to capture custom fields, and analytics search/ADQL queries; practice identifying which configuration step is missing for a given symptom.