Application Monitoring( AIAPM)

After successfully configuring AI Application Performance Monitoring (AI APM), Ceburu provides a comprehensive Application Monitoring Dashboard under AI Monitoring.
This dashboard offers real-time visibility into application health, service performance, transactions, latency, errors, and AI-detected anomalies.

The AI APM dashboard helps teams:

• Monitor application behavior in real time

• Identify performance bottlenecks

• Detect anomalies automatically

• Troubleshoot errors faster

• Understand service and route-level performance

Navigation Path

Applications → AI Monitoring → Application Monitoring

Once navigated, the Application Monitoring Dashboard will be displayed.

Application Monitoring Dashboard Overview:

The dashboard provides a high-level summary of application activity for the selected time range.

Request Rate Over Time: Displays request volume trends over the selected time range, helping identify spikes or drops in traffic.

Status Distribution: Breaks down HTTP response codes (2xx, 3xx, 4xx, 5xx) to quickly assess error patterns.

Request Rate by Routes: Shows traffic volume per API route or endpoint.

P95 Latency by Routes: Identifies slow-performing routes using 95th percentile latency.

This allows teams to:

• Pinpoint slow APIs

• Compare route performance

• Detect latency regressions

Top Services: Displays services ranked by transaction count.

Top Transactions: Shows the most frequently executed transactions or endpoints.

Recent Anomalies: AI automatically detects anomalies based on historical behavior.

Anomaly Details Include:

• Timestamp

• Service name

• Host

• Operation

• Anomaly type (CPU, Memory, Transaction, System, etc.)

Services Overview

Applications → Services Overview

This view lists all monitored services with aggregated performance metrics.

Metrics Shown

• Failure rate

• Average latency

• Maximum latency

• Application association

Viewing Service-Level Details

How to View Service Details

1. Navigate to Services Overview

2. Click on a specific Service Name (e.g., PetClinic, Pitstop_WebApp)

3. The Service Details Page opens

Service Details – Overview Tab

The service details page provides deep insights into a selected service.

Key Metrics

• Total Requests

• Hosts

• Transactions & Spans

• Average & Max Latency

• Error Count & Error Rate

• Anomaly Count

Host Metrics

Shows per-host performance metrics including:

• CPU usage (avg & max)

• Memory usage (avg & max)

• Latency per host

• Error distribution per host

Transactions Per Minute (TPM)

Tracks service load over time.

Span Duration

Visualizes execution duration of spans.

Resource Metrics

• CPU Usage

• Process CPU Usage

• Memory Usage

These insights help correlate performance issues with resource constraints.

Service-Level Anomalies

Lists AI-detected anomalies specific to the selected service, allowing faster root cause analysis.

• Identify slow or failing services

• Detect abnormal CPU or memory usage

• Troubleshoot API latency issues

• Analyze transaction performance

• Monitor application health proactively using AI

Transactions:

Once inside a service (e.g., Pitstop_WebApp), the Transactions tab displays multiple visual panels summarizing request behavior.

Key Panels

1. Latency

Shows Average and Maximum latency over time for transactions.

• Helps identify latency spikes

• Red markers indicate potential anomaly points

2. Transactions Per Minute (TPM)

Displays transaction throughput trends.

• Useful for traffic pattern analysis

• Correlates load with latency and errors

Failed Transactions & Counts

Displays the failed rate (%) or failed count, depending on selection.

• Helps track persistent or sudden failure patterns

Response Status Codes

Breakdown of HTTP responses such as:

• 200 (Success)

• 502 (Bad Gateway)

This enables quick visibility into error-heavy periods.

Transactions List View:

Below the charts, the Transactions table lists all monitored endpoints for the selected service.

Viewing Transaction Details

How to Drill Down

1. Click on a Transaction name (e.g., GET VehicleManagement/Index)

2. The Transaction Details view opens

Transaction Details View

This view shows individual request executions for the selected transaction.

Information Displayed

• Timestamp

• Trace ID

• Transaction ID

• Latency per request

• URL

• Transaction result (HTTP status)

Trace ID & Request Breakdown

Clicking a Trace ID expands the request and reveals:

Additional Details

• Source IP

• Node name

• Service name

• Host name

• Transaction type

• Span count

This allows you to trace exactly where time is spent during request execution.

Trace Timeline (Distributed Tracing)

The Trace Timeline visualizes the full request path across services.

• Parent and child spans

• Service-to-service calls

• Database queries

• Exact execution time per span

Selecting a transaction / Span opens a detailed side panel. which provides a detailed view.

Dependencies:

The Dependencies tab provides a visual service dependency map that shows how an application or service communicates with other internal services and external resources. 

Service Dependency Map

The dependency map visually represents:

• Upstream services (callers)

• Downstream services (dependencies)

• External dependencies (HTTP, databases, APIs)

Example (Pitstop_WebApp)

• Calls CustomerManagementAPI

• Calls VehicleManagementAPI

• Connects to an external HTTP endpoint (IP:Port)

Dependency Edge Details

Clicking on a connection edge between services opens detailed metrics.

• Average latency

• Transactions per minute (TPM)

• Total calls

• Success count & success rate

• Error rate

This helps identify slow or unreliable dependencies.

Service Node Details

Clicking on a service node opens a contextual panel.

Service Node Information

• Service type (e.g., .NET)

• Connected nodes

• Connected edges

• Quick navigation to:

  • Service details

  • Service metrics

Services & Dependencies Tables

Below the map, tabular views summarize all detected relationships.

Services Table

Shows:

• Service name

• Type

• Hosts

• Transaction counts

Dependencies Table

Shows:

• Dependency endpoint

• Type (HTTP / Database)

• Span type

• Port and IP details

Database Dependencies

For API services, database dependencies are automatically detected.

Example:

• VehicleManagementAPI → MSSQL (VehicleManagement)

This enables visibility into backend bottlenecks and database-driven latency.

Metrics Overview:

The Metrics tab provides real-time and historical resource utilization insights.

CPU Usage

• Average CPU usage

• Maximum CPU usage

• Anomaly markers shown as red indicators

Process CPU Usage

• CPU usage by application process

• Helps detect inefficient code or background loops

Memory Usage (%)

• Memory consumption as a percentage

• Detects gradual memory leaks

Memory Usage (Bytes)

• Absolute memory usage values

• Useful for capacity planning

Anomaly Indicators in Metrics

Red dots on graphs indicate AI-detected anomalies.
Clicking these opens detailed anomaly explanations.

AI Insights – Anomalies:

The AI Insights tab automatically detects abnormal behavior using machine learning.

Anomalies List View

Each anomaly includes:

• Timestamp

• Service

• Host

• Operation (CPU / Memory)

• Type

• Action to View details

Anomaly Details & Root Cause Analysis

Click View on any anomaly to see detailed AI analysis.

Information Provided

• Detection time

• Metric involved

• Duration

• Service and node

• Performance metrics at detection time

AI RCA (Root Cause Analysis)

AI explains:

• What happened

• Why it likely occurred

• Whether it was load-related or internal

• Recommended next steps

Example:

• Memory spike without traffic

• Low CPU & zero TPM

• Potential background process or memory leak

AI Insights – Alerts

The Alerts section lists threshold-based or AI-generated alerts.

Alerts List

Each alert includes:

• Timestamp

• Service

• Host

• Alert type (CPU / Memory)

• Operation

• Observed value

• Action to View details

Alert Details

Clicking View opens:

• Alert type and metric

• Threshold breach value

• Timestamp

• Associated service and host

Alerts enable proactive remediation before service degradation occurs.

Service Map Overview

The Service Map provides a global, real-time visual representation of all monitored applications, services, APIs, databases, and external dependencies within the environment.

It helps teams quickly understand:

• Application-to-application flow

• Upstream and downstream dependencies

• Cross-technology interactions (Java, .NET, HTTP, Databases)

• End-to-end architecture visibility

Application Flow Map

The Application Flow Map visually displays how requests flow across services.

Example Flow

• PetClinic (Java)
  → PostgreSQL Database
  → External HTTP Endpoint (IP:Port)

• Pitstop_WebApp (.NET)
  → CustomerManagementAPI
  → VehicleManagementAPI
  → MSSQL Databases

Visual Indicators

• Green arrows – Healthy / successful traffic

• Red arrows – Error-prone or failing dependencies

• Node labels – Technology stack (.NET, Java, SQL Server, PostgreSQL)

This allows teams to instantly identify problematic paths and critical dependencies.

Interacting with the Service Map

You can:

• Click on any service node to navigate to service details

• Click on connections to analyze dependency health

• Zoom, pan, and reposition nodes for clarity

• View both internal services and external dependencies

AI Insights from Applications View:

From the Applications section, the AI Insights tab provides centralized visibility into anomalies and alerts across all applications.

Go to RUM & Synthetics:

From the Applications Monitoring  users can navigate directly to RUM & Synthetic Monitoring.

Navigation Option

Clicking Go to RUM & Synthetics redirects users to:

AI Monitoring → AI APM → Application Monitoring

Available Monitoring Types

• RUM (Real User Monitoring)
  Tracks real end-user experience such as page load time, response time, and client-side performance.

• Synthetic Monitoring
  Simulates user transactions and availability checks for proactive monitoring.