Observability

Observability feature offers a comprehensive suite of tools designed to enhance visibility and insight across your application's performance and usage. With a focus on user-friendliness, the standout aspect of our observability suite is its ease of setup—every feature can be enabled with just a click of a button, streamlining the integration of advanced monitoring capabilities into your workflow.

The ComUnity Platform provides comprehensive observability tools that help you understand your application's health, performance, and user experience. With integrated monitoring, logging, and analytics, you can quickly identify issues, optimise performance, and understand how users interact with your applications.

What makes our observability different: Everything is integrated out-of-the-box. Enable observability once, and you immediately get metrics dashboards, log search, distributed tracing, and user analytics—all connected and correlated for faster troubleshooting.

The Three Pillars of Observability

Observability in the ComUnity Platform is built on three complementary data sources that work together to give you complete visibility:

Metrics: System Performance

Monitor your application's health with real-time performance data.

What you'll track:

• Request rates, error rates, and latency (P99, P95)

• Resource usage (CPU, memory, database connections)

• Service health and availability

• Custom business metrics

When to use: Daily health checks, performance optimisation, capacity planning

Explore Metrics →

Logs: Detailed Event Records

Search and analyse detailed logs to understand what happened and why.

What you'll track:

• Error messages and stack traces

• User actions and system events

• API requests and responses

• Application behaviour and debugging info

When to use: Troubleshooting errors, debugging issues, audit trails

Explore Logs →

Traces: Request Flow Visualisation

Follow individual requests through your entire system to identify bottlenecks and failures.

What you'll track:

• End-to-end request flows across services

• Duration of each operation

• Service dependencies and call chains

• Performance bottlenecks in distributed systems

When to use: Debugging slow requests, understanding service interactions, optimizing workflows

Explore Traces →

Client Analytics: User Behaviour

Understand how users interact with your application through privacy-first analytics.

What you'll track:

• User engagement and session duration

• Most visited pages and features

• Geographic distribution and device types

• User flows and drop-off points

When to use: Feature adoption analysis, UX optimisation, understanding user behaviour

Explore Client Analytics →

How These Tools Work Together

The power of the ComUnity Platform's observability comes from how these tools integrate:

Scenario: Users report slow page loads

1. Start with Metrics → Notice P99 latency spike at 14:30

2. Check Logs → Find error messages during that time period

3. View Traces → See complete request flow and identify slow database query

4. Correlate → All three tools share trace IDs for seamless navigation

Every log entry includes a trace ID. Every trace links to logs. Metrics dashboards link to both. This correlation eliminates the need to manually piece together data from different systems.

Quick Start Guide

Step 1: Access Observability

Observability is segmented per environment, with Metrics, Traces, and Logs supported by default in each active environment. Client Analytics requires manual enablement. Follow the steps below to enable Client Analytics for each deployment environment.

Enabling Client Analytics

Prerequisites:

• Access to the ComUnity Developer Toolkit

• Project with at least one active environment

Steps:

1. Log into the Toolkit with your credentials

2. Open your project from the project list

3. Navigate to Observability in the main menu

4. Go to Project Settings > Environment tab> Observability tab:

   
5. Click "Activate Page Analytics" and wait for the background process to complete

6. Access your dashboards from the Observability menu

Client Analytics is enabled for the selected environment. Your observability dashboards are now accessible from the Observability menu.

You'll need to enable Client Analytics separately for Development, QA, and Production environments.

Time to enable: Approximately 2-3 minutes per environment

Step 2: Access Your Dashboards

Once enabled, you can access four integrated dashboards:

Metrics Dashboard

Logs Dashboard

Traces Dashboard

Metrics Dashboard

• View real-time performance data

• Monitor service health

• Track custom metrics

• Create alerts for issues

Logs Search

• Search error messages

• Filter by time and service

• Find trace IDs for correlation

• Debug production issues

Traces Viewer

• Visualize request flows

• Identify bottlenecks

• Debug slow operations

• Understand service dependencies

Client Analytics

• Track user engagement

• Understand feature adoption

• Analyse user flows

• Monitor traffic sources

Step 3: Instrument Your Application (Optional)

The platform automatically collects infrastructure metrics, logs, and traces. For deeper insights, you can add custom instrumentation:

Add custom metrics for business logic (e.g., payment success rate, user signups) Add structured logging for better searchability Add custom trace spans for specific operations

Learn about instrumentation → (coming soon)

Common Use Cases

Use Case 1: Troubleshooting Production Errors

Problem: Users reporting "payment failed" errors

Investigation workflow:

1. Logs: Search for "payment failed" errors

2. Find trace ID in the error log entry

3. Traces: Open the trace to see full request flow

4. Identify: Payment gateway timeout after 30 seconds

5. Action: Increase timeout or add retry logic

Time to resolution: Minutes instead of hours

Use Case 2: Optimising Slow Endpoints

Problem: API endpoint taking 5 seconds (should be under 500ms)

Investigation workflow:

1. Metrics: Notice P99 latency spike in dashboard

2. Traces: Filter to slow requests (>4 seconds)

3. Identify: Database query taking 4.8 out of 5 seconds

4. Logs: Find the actual SQL query in log details

5. Action: Add database index or optimise query

Result: 10x performance improvement

Use Case 3: Understanding Feature Adoption

Problem: New feature launched but unsure if users are using it

Investigation workflow:

1. Client Analytics: Check page visits to feature screen

2. Compare: Feature page visits vs total visits = adoption rate

3. Analyze: Check bounce rate and time on page

4. Result: 15% adoption, high bounce rate = users trying but not engaging

5. Action: Improve feature onboarding

Insight: Data-driven feature development

Use Case 4: Capacity Planning

Problem: Need to prepare for traffic surge during campaign

Investigation workflow:

1. Metrics: Review historical peak traffic patterns

2. Identify: Current capacity handles 1,000 req/sec

3. Calculate: Expected campaign traffic is 3,000 req/sec

4. Traces: Check if any services have bottlenecks under load

5. Action: Scale infrastructure proactively

Result: Zero downtime during campaign

Getting Help

Documentation

• Metrics Guide - Understanding dashboards and creating alerts

• Logs Guide - Searching logs and debugging with LogQL

• Traces Guide - Reading trace visualisations and finding bottlenecks

• Client Analytics Guide - Understanding user behavior and analytics

• Troubleshooting Guide - Common issues and solutions (coming soon)

• Quick Reference - Query cheat sheets and glossary (coming soon)

Best Practices

For Daily Monitoring

Check metrics dashboards daily

• Review error rates and latency

• Look for unusual patterns

• Verify no alerts are firing

Use logs for investigation

• Start with time period when issue occurred

• Search for errors or specific events

• Follow trace IDs to see full context

Review analytics weekly

• Track user engagement trends

• Identify popular features

• Monitor mobile vs desktop usage

For Troubleshooting

Follow the investigation pattern:

1. Metrics → Identify when problem started

2. Logs → Find specific error messages

3. Traces → See complete request flow

4. Correlate → Use trace IDs to connect data

Ask the right questions:

• What happened? (Logs)

• When did it happen? (Metrics)

• Where in the system? (Traces)

• Who was affected? (Analytics)

For Performance Optimisation

Focus on user-facing metrics first:

• P99 latency (worst-case user experience)

• Error rate (user frustration)

• Page load time (user engagement)

Use traces to find bottlenecks:

• Identify longest operations

• Optimise database queries

• Cache expensive operations

• Consider async processing

Measure the impact:

• Compare before/after metrics

• Check if user engagement improved

• Verify error rates decreased

Security and Privacy

Data Ownership

All observability data stays on your infrastructure. No data is sent to third-party services.

Privacy Compliance

• GDPR compliant: IP anonymisation, consent management

• CCPA compliant: User opt-out and data deletion

• HIPAA compatible: Can be configured for healthcare applications

Access Control

Observability data access is controlled through ComUnity Platform permissions. Users only see data for projects and environments they have access to.

Technical Details

Technology Stack

• Metrics: Prometheus + Grafana + Thanos

• Logs: Loki with LogQL query language

• Traces: Jaeger + OpenTelemetry + Tempo

• Analytics: Matomo (open-source, privacy-first)

Data Retention

• Metrics: High resolution for 30 days, downsampled for 1 year

• Logs: 30 days by default (configurable)

• Traces: Sampled storage for 30 days

• Analytics: Unlimited retention

Performance Impact

• Minimal overhead on applications (< 1% CPU, < 50MB memory)

• Automatic sampling for traces reduces data volume

• Asynchronous logging prevents blocking

Next Steps

Just Enabled Observability?

1. Understand Metrics → Start with your service health dashboard

2. Learn Log Searching → Find and debug errors quickly

3. Explore Traces → Visualise request flows

Ready for Advanced Features?

1. Set Up Alerts → Get notified when issues occur (coming soon)

2. Add Custom Instrumentation → Track business metrics (coming soon)

Need Help?

• Review the troubleshooting guides for common issues

• Contact support through [e-mail]

• Check the quick reference for query syntax