Application Monitoring

Application monitoring gives you request-level visibility: every transaction is traced end to end, service dependencies are mapped automatically, and slow methods are surfaced down to the line of code.

Instrumentation#

Java and .NET apps are instrumented with zero code changes by attaching the Applicare agent. For other runtimes, use OpenTelemetry — Applicare ingests OTLP natively (see OpenTelemetry).

# Java — add the javaagent flag
java -javaagent:/opt/applicare/arc-agent.jar \
     -Darc.service=checkout-api -jar checkout.jar

Service map & transactions#

Applicare builds a live topology of services, databases, queues, and external calls from trace data. Each edge shows throughput, latency percentiles, and error rate. Click any transaction to open a flame graph with method-level timings and the exact SQL or HTTP call responsible for the time spent.

Finding the slow path#

Sort transactions by slowest or most impactful (latency × throughput). The trace view highlights the critical path and flags N+1 query patterns and remote-call fan-out. From there, Root Cause Analysis links the slowdown to the upstream change or resource that caused it.