FabricInsight – Data Center Network Analyzer
FabricInsight is an intelligent data center network analyzer. Based on big data analytics, FabricInsight provides ubiquitous network application analysis and visualization functions to streamline applications and networks.
FabricInsight collects masses of real service packets through telemetry, provides correlation analysis between the internal applications and networks of the data center, and displays the application map and network quality in real time, helping customers quickly identify faults and proactively identify risks before services are affected.
Based on Huawei’s best IT practices, FabricInsight meets the intelligent O&M requirements of financial and large enterprise customers’ data centers.
Rapid Fault Identification
Second-level display of service flows and network-wide KPIs through telemetry;
Intelligent identification of more than 10 types of issues and one-click diagnosis of network issues
Precise Historical Fault Playback and Rapid Data Retrieval
Retrieval of tens of billions of records and processing of millions of packets in seconds;
Millisecond-level detection of historical microbursts, and accurate fault playback
AI-based Predictive Maintenance
ML/AI-based dynamic baseline construction and issue identification at the device, queue, and port levels;
Proactive prediction of optical module issues;
Effective network issue rate reduction
Feature Description Dashboard
- Supports visualized hosts statistics and analysis.
- Supports global filtering by multiple dimensions.
- Allows users to view historical statistics in a user-defined period.
- Allows users to add, delete, and modify templates, as well as modify and save the layout of the current view.
- Supports multi-dimensional retrieval of flow data.
- Allows users to view the number of SYN events, traffic, and delay in a specified period.
- Compares and analyzes the average and maximum delays of TCP events on the network within a specified period.
- Displays the fabric network topology, marks abnormal links, and collects statistics on the number of active IP addresses of leaf switches.
- Displays abnormal TCP events on the network within a specified period, including TCP RST, TCP retransmission, TCP flag packet exception, and TTL exception.
- Supports link-based flow tracing.
- Searches and displays data from multiple dimensions.
- Analyzes application health status.
- Analyzes interaction between applications and displays application interaction relationships in the topology.
- Supports intra-application interaction analysis and mutual visibility between applications and networks.
- Defines applications based on IP addresses or ports.
- Searches and filters data from multiple dimensions.
- Allows users to view the TCP event time axis and event details of a single flow in a specified period.
- Displays detailed information about packet loss events.
- Combines packets for analysis in the 1:1 NAT mapping scenario.
Network device performance monitoring
- Displays the metric trend of devices, boards, interfaces, queues, and optical modules in a chart.
- Displays dynamic baselines and detects exceptions of metrics of devices, boards, and interfaces.
- Supports millisecond-level queue congestion and packet loss detection.
- Predicts optical module issues.
- Detects and analyzes application quality issues, including: continuous service interruption, intermittent service interruption, unreachable host port, and abnormal sessions matched based on rules.
- Detects and analyzes network service issues, including: insufficient TCAM resources, insufficiency or sharp change of FIB entry resources, and insufficiency or sharp change of ARP entry resources.
- Detects and analyzes security compliance issues, including: non-compliant traffic interaction, suspicious SYN flood attack, and suspicious port scanning attack.
For inquiries, email us at email@example.com