Operating System Console-Alibaba Cloud Linux(Alinux)-阿里云帮助中心

The Operating System Console provides features to manage operating systems, including operations and maintenance, an intelligent OS assistant, component management, and subscription management.

Category		Feature	Description
System Health	System Health		System Health provides an overall assessment of a cluster, node, or container's health based on key monitoring metrics. This helps you understand the general health status and investigate the specific factors affecting it.
System Diagnosis	Memory Diagnosis	Memory Panorama Analysis	Memory Panorama Analysis is for scenarios where memory usage is high, but the cause is unclear.
		OOM Diagnosis	OOM Diagnosis helps you identify the root cause of Out of Memory (OOM) events in the operating system.
	Storage Diagnosis	I/O Traffic Analysis	I/O Traffic Analysis helps you identify which workloads are generating I/O traffic in the system. It is typically used to troubleshoot I/O burst issues.
		One-click I/O Diagnosis	One-click I/O Diagnosis focuses on frequently occurring issues such as high I/O latency, I/O burst, and I/O wait. This feature identifies different types of I/O issues, calls the appropriate subtools to analyze I/O data, and provides conclusions and recommendations. This helps you effectively analyze and resolve I/O-related issues.
	Network Diagnosis	Packet Loss Diagnosis	Packet Loss Diagnosis helps you analyze packet loss that occurs at the operating system kernel level during network transmission and provides recommended solutions.
		Network Jitter	This feature helps you determine the cause of network jitter on your ECS instance.
	Scheduling Diagnosis	Scheduling Jitter Diagnosis	Scheduling Jitter Diagnosis helps analyze why user-space processes remain unscheduled for long periods, often because the CPU is not switching tasks. This can happen during operations like memory reclamation.
		System Load Diagnosis	System Load Diagnosis analyzes the cause of an abnormal 1-minute average system load (the load1 metric) and provides detailed information along with recommended actions.
	Crash Diagnosis		When an ECS instance experiences issues such as a kernel panic or an Out of Memory (OOM) event, or when you are notified of an OS crash by a system event, use Crash Diagnosis to analyze the root cause. The diagnostic report provides recommendations to help you fix the issue.
	One-click Diagnosis		One-click Diagnosis helps troubleshoot complex performance issues that span multiple kernel subsystems. When a single metric is abnormal but the root cause is unclear, this feature intelligently selects and runs multiple diagnostic tools. It then correlates the results to perform a comprehensive analysis and recommends a complete solution.
	Pod Diagnosis		Pod Diagnosis diagnoses Pods deployed in ACK/ACS clusters. It currently covers memory-related scenarios, including Memory Panorama Analysis, OOM Diagnosis, and Java application memory diagnostics.
System Observation	Process Hotspot Tracking		Process Hotspot Tracking analyzes hotspots on a single node over a specific period. It generates a call graph for processes, helping developers quickly identify performance bottlenecks and hotspots in their applications.
	Hotspot Comparison Analysis		When system metrics become abnormal due to factors like application performance fluctuations, Hotspot Comparison Analysis uses hotspot comparison time-series charts and differential flame graphs to help you identify hotspot differences over time for processes on an instance.
AI Performance & Diagnosis	AI Profiling		AI Profiling dynamically injects GPU Profiling to collect and analyze information during AI job execution, such as data on operators, scheduling, and kernel functions, providing observability. This enables non-intrusive, flexible, and low-overhead observation of AI jobs.
	GPU Diagnosis		GPU Diagnosis diagnoses system anomalies in AI scenarios and generates diagnostic conclusions, GPU runtime status, and AI job execution status. This helps developers quickly identify issues in their applications.
	Continuous GPU Profiling		Continuous GPU Profiling helps analyze performance hotspots of AI applications on both CPUs and GPUs. By generating flame graphs, it allows developers to visualize process call stacks and time consumption, making it easier to locate bottlenecks and optimize task execution.
System Management	System Management		System Management displays the managed status and other relevant information for instances within the Operating System Console. Only managed instances can use the features of the Operating System Console.
Component Management	Component Management		Component Management manages OS extension components. It allows you to install, upgrade, and uninstall components on nodes (ECS instances) or on ECS instances within ACK clusters.
Subscription Management	Subscription Management		A subscription grants an ECS instance permission to access private Yum repositories. With Subscription Management, you can view the current subscription status of your ECS instances and create new subscriptions as needed.
Abnormal Event Alerting	Abnormal Event Alerting		By creating an alerting policy and configuring notification subscriptions, you can receive timely notifications through channels like DingTalk, email, or SMS when system events, such as abnormal CPU usage, occur. This enables a quick response and resolution.