Monitor Platform Health

Maintaining robust platform health is critical for reliable operations and optimal user experience. Our platform provides multiple tools for effective health monitoring, each catering to different needs and levels of detail. In this article, we’ll outline the three main ways to monitor platform health today and preview upcoming enhancements on our roadmap.

Installer Admin Console: Application Service State Monitoring

The installer admin console serves as your first line of insight into the platform’s operational status. Through its intuitive interface, administrators can:

• Check application service states: View the status of core services, including containers, to quickly spot any service that is stopped, failing, or restarting abnormally.

• Diagnose basic issues: The console presents clear indications of unhealthy or degraded services, allowing for rapid troubleshooting.

• Check container logs

• Analyze and generate support bundles

This method is ideal for administrators who need a quick, high-level overview of application components.

Cluster Status

Support Bundle Analysis

Platform Health User Interface: KPI-driven Insights

For a more detailed analysis, the dedicated platform health UI provides:

• Key Performance Indicators (KPIs): Real-time metrics that reflect the health of critical flows, such as data ingestion and query processing.

• Issue Highlighting: Visual cues and alerts make it easy to identify potential bottlenecks or failures in core platform operations.

• Focused troubleshooting: By surfacing relevant KPIs, the UI helps you pinpoint the area (ingestion or query) where attention is needed.

This interface is designed for users who want to proactively manage platform performance and address issues before they impact end users. It is only available to admins

Platform Health Components and Criteria

The Platform Health UI monitors specific components with defined health criteria. Understanding these criteria helps you interpret status indicators and take appropriate action.

Status Indicators

Harvester Health Criteria

The Harvester component manages data ingestion from collectors. Health is determined by:

• Collection Rate: Rate at which data is being ingested from sources

• Processing Lag: Delay between data collection and availability for queries

• Error Rate: Percentage of failed ingestion attempts

• Queue Depth: Number of pending items awaiting processing

Tip: Hover over the Harvester component in the UI to see current values for each criterion.

Query Cache Health Criteria

The Query Cache component optimizes query performance. Health is determined by:

• Cache Hit Rate: Percentage of queries served from cache vs. requiring fresh computation

• Response Time: Average time to return cached query results

• Memory Utilization: Cache memory usage relative to allocated capacity

Session Components Health

Session Components reflect the overall health of active user sessions and related services. A degraded or critical status here may indicate:

• High session load

• Authentication service issues

• Backend connectivity problems

Troubleshooting Guidance

When you encounter a Degraded or Critical status:

1. Hover over the component: The UI displays diagnostic details including current metric values and which criteria triggered the status change.

2. Check Grafana dashboards: Click the Grafana link to access detailed historical metrics and identify trends or anomalies.

3. Review recent changes: Consider whether recent configuration changes, deployments, or increased load may have contributed.

4. Contact TAC: If the issue persists or root cause is unclear, contact your Technical Assistance Center representative with the diagnostic information gathered.

Feature Under Development

Platform Health monitoring currently focuses on ingestion pipeline health. Additional components and more detailed health criteria are planned for future releases.

Local Grafana Dashboards: Deep Dive with Service KPIs

For advanced monitoring and custom analysis, each deployment includes a local Grafana instance:

• Linked from the health UI: Easy access from the platform health interface.

• Default health dashboards: Pre-configured dashboards display service KPIs for a comprehensive view of platform health.

• Powered by Prometheus: Grafana dashboards draw metrics from a local Prometheus service, enabling powerful visualization and historical analysis.

This approach is best for technical teams who need granular metrics and the ability to create custom visualizations and alerts. It is only available to admins.

Summary Table

Example Use Case:

If a user reports slow data queries, start with the platform health UI to check query flow KPIs. If issues are detected, follow links to the local Grafana dashboard for a deeper analysis of backend service performance.

Roadmap: External Notifications for Platform Health

Looking ahead, we are developing support for notifications to external systems when platform health issues are detected. This will enable:

• Automated incident response: Integrations with external monitoring, alerting, or ticketing systems.

• Faster remediation: Immediate awareness for teams outside the platform’s native interfaces.

Stay tuned for updates as we enhance our platform health monitoring with seamless external notification capabilities.