As businesses expand and IT infrastructure matures, enterprises have actively procured a diverse range of monitoring tools to ensure comprehensive oversight of their foundational systems. These tools broadly cover multiple layers—including operating systems, critical components, and hardware—thereby completing the initial build-out of IT infrastructure and IT O&M management tooling. However, with rapid business growth, existing O&M tools and systems have begun to reveal numerous challenges: resources are scattered across silos, and the lack of effective, unified, and standardized management has led to incomplete monitoring coverage and increasingly difficult alert management. Moreover, high monitoring configuration costs, low efficiency, and growing difficulties in team collaboration make it challenging for enterprises to respond swiftly to evolving business demands. Against this backdrop, the need to build a unified monitoring platform has gradually become the central focus of monitoring initiatives across enterprises.
Facing these challenges, a major telecom operator leveraged the CanWay BlueWhale Full-Stack Observability Center to launch an infrastructure monitoring system construction project. By reshaping its O&M framework and building a unified, integrated monitoring platform, the operator comprehensively elevated its monitoring and management capabilities, improved IT O&M efficiency, and established robust support for the safe, continuous, and uninterrupted operation of its IT systems.
01 Business Scenario
Over the years, this enterprise had self-built open-source monitoring platforms such as Zabbix and Prometheus, independently implementing monitoring for a large number of operating systems and component services. At the same time, it procured third-party hardware monitoring products to supplement hardware monitoring capabilities. However, as the enterprise grew, this management model—lacking a cohesive monitoring framework—gradually exposed critical issues: monitoring without proper governance led to low coverage rates; the absence of unified standards resulted in chaotic policy configurations; and the mixed deployment of multiple monitoring systems created excessive O&M complexity. The traditional monitoring management approach became increasingly difficult to sustain, making it urgent to build a unified monitoring platform.
02 Pain Point Analysis
The company's current monitoring infrastructure is at the siloed-tool stage, with various monitoring scenarios remaining incomplete. The customer expects to establish a mature, integrated monitoring platform while filling the gaps in monitoring capabilities. From the perspective of individual O&M scenarios, the enterprise currently faces the following pain points:
Operating System Monitoring: Zabbix and Prometheus are deployed, but monitoring is configured independently by each business system. There is no unified metric framework or threshold standards, and overall monitoring is in an ungoverned state.
Component Monitoring: The existing monitoring system provides only rudimentary component monitoring, lacking collection of core metrics. Without policy templates, teams are unclear on how to properly configure monitoring.
Container Monitoring: Container monitoring capabilities are entirely absent. Container resources and containerized component services are completely unmonitored, posing extremely high risks to system reliability.
Unified Monitoring Platform: A third-party hardware monitoring product has been procured but is managed independently. Each use requires a separate login to the hardware management platform for configuration. The authorization and management framework is complex and inconvenient. The enterprise seeks to unify management through an integrated monitoring platform.
03 Solution
Operating System Monitoring — BlueKing Agent-Based Metric Collection
The CanWay BlueWhale Full-Stack Observability Center uses the BlueKing Agent as its core, with built-in operating system collection plugins. Once the BlueKing Agent is deployed, it automatically collects OS-related metric data without manual configuration. Through the One Agent approach, the company achieved unified monitoring and data collection across all internal operating systems.
Component Monitoring — Powerful and Proven Collection Extensibility
The Full-Stack Observability Center employs an Agent + Plugins design pattern, supporting rapid extension of monitoring coverage for various objects through system scripts, SQL queries, Exporters, Datadog plugins, and more—resolving the challenge of collecting monitoring data for diverse component objects under the Agent model.
Additionally, the Observability Center supports extension via protocols and interfaces (including SNMP, IPMI, JMX, SQL, BK-Pull, etc.) for remote data collection, addressing component monitoring needs in various agentless scenarios.
Building on these approaches, the Observability Center has also accumulated a substantial library of standardized built-in plugins, covering the vast majority of mainstream databases and middleware. It also features a mature metric framework and provides best-practice configuration templates to guide users through monitoring setup.
Container Monitoring — Comprehensive Coverage of Container Resources and Service Metrics
Based on an optimized implementation of the Kubernetes-native Prometheus monitoring approach, the company achieved container monitoring across the following scenarios:
Automatic discovery of various resource objects within containers, with collection of related performance metrics including Cluster, Workload, Pod, Container, and Node.
Monitoring of component services deployed on containers, with data collection via the following methods:
-ServiceMonitor (recommended) and PodMonitor
-Sidecar approach (deploying an exporter scraper as a sidecar to expose metrics, combined with ServiceMonitor for collection)
-Centralized remote collection (for components that natively expose /metrics endpoints, combined with ServiceMonitor for collection)
Unified Monitoring — Third-Party Monitoring Data Integration to Build an Integrated Monitoring Platform
Through the CanWay BlueWhale monitoring system, the company integrated third-party monitoring data by developing monitoring source plugins to connect with, scrape, or receive data from other monitoring systems. With proper data structure cleansing, the ingested data can be associated with BlueKing CMDB instances, thereby achieving full parity with the CanWay BlueWhale Full-Stack Observability Center's native data in terms of metric management, data inspection, visualization, and other capabilities—building a truly integrated monitoring platform.
04 Results Showcase
Operating System Monitoring — BlueKing Agent-Based Metric Collection
Synchronization of CMDB operating system configuration information
BlueKing Agent-based operating system metric collection
Component Monitoring — Core Database and Middleware Monitoring Integration with Policy Configuration
MySQL data dashboard
Kafka data dashboard
Container resource discovery and display list
Container Monitoring — K8s Container Management Platform Monitoring Integration
Container resource discovery and display list
Container resource performance metrics
In-cluster ServiceMonitor and PodMonitor
Container component service metrics
Unified Monitoring — Third-Party Monitoring Data Integration to Build an Integrated Monitoring Platform
Third-party hardware monitoring system metric data integration
05 Implementation Outcomes
06 Scenario Applicability
The CanWay BlueWhale Full-Stack Observability Center has developed comprehensive monitoring solutions and best-practice guidance across all O&M layer scenarios, helping enterprises maximize monitoring coverage. It also provides mature monitoring data integration solutions that enable seamless data exchange with third-party monitoring systems while delivering a fully consistent experience across data processing, storage, and visualization. This solution is applicable to the following types of enterprises:
Enterprises with non-standardized monitoring onboarding, where individual business systems operate independently and the monitoring framework lacks unified governance.
Enterprises that have implemented basic monitoring collection but lack effective metric framework development, with non-standardized Monitoring and Alerting configurations.
Enterprises with low monitoring coverage rates, missing scenarios such as container monitoring and hardware monitoring, seeking to fill these capability gaps.
Enterprises with siloed monitoring environments where various scenario-specific monitoring systems exist but operate in complete isolation, creating management complexity and an urgent need to build a IT monitoring platform.
