Product Introduction¶
What is Guance?¶
Guance provides a one-stop observability solution for cloud, cloud-native, applications, and business. It includes infrastructure monitoring, log analysis, application performance monitoring, real user monitoring, service availability monitoring, security monitoring, LLM monitoring, CI environment insights, and other features, helping to achieve comprehensive observability for cloud, cloud-native, applications, and business.
Guance started development in 2018. After five years of meticulous refinement by a professional team, it was officially released on April 28, 2022. Since its release, we have maintained a regular iteration and update schedule, actively collecting user feedback, continuously improving product technology stack coverage, user experience, and platform stability. We provide enterprises with intelligent monitoring cloud services for the AI era, helping them quickly achieve system observability and supporting stable business growth.
Core Features¶
Infrastructure monitoring for hosts, cloud environments, containers, processes, and networks, covering various IT facilities such as IoT, multi-cloud environments, virtualization, and microservices. Supports custom infrastructure objects, allowing you to observe any infrastructure.
Comprehensive performance insights, covering end-to-end service monitoring. Built-in flame graphs and Profiling tools allow developers to intuitively see slow queries and resource consumption hotspots, helping to optimize business experience.
Real User Monitoring collects data related to the real experience and behavior of users when interacting with website applications, identifying the real needs behind each request to optimize product performance.
Guance supports various Real User Monitoring types including Web, mobile (Android & iOS), and Mini Programs.
All logs are unified and aggregated on the Guance platform, enabling cross-system, one-stop search and rapid pinpointing. Whether it's a single issue or a widespread anomaly, investigate holistically with context to quickly discover and resolve problems.
By integrating the two core capabilities of CSPM and SIEM, we build an integrated security monitoring system of "Assets ➛ Configuration ➛ Behavior" for you.
Agent Monitoring associates Agent requests with the entire application chain, tracing the complete process of each conversation and accurately measuring the number of Tokens consumed by each generation task.
Guance supports custom monitor detection rules and comes with 700+ out-of-the-box monitor templates. By configuring detection rules, trigger conditions, and event notifications, you can receive alert notifications promptly, identifying and resolving issues in a timely manner.
The Event Center integrates event data from various sources such as monitors, intelligent monitoring, security monitoring, and SLO. Through a centralized event explorer, you can quickly locate anomalies, track system changes, and improve incident response efficiency.
Guance comes with 20+ standardized visualization charts and 200+ data monitoring view templates. You can easily and quickly build dashboards according to your business needs.
Obsy AI AI is built on advanced AI large language models. Through natural language processing and machine learning technologies, it can efficiently process data and quickly locate and resolve problems.
Key Technologies¶
1. Data Storage Technology Based on Time-Series Database and Columnar Data
Time-series databases and columnar databases offer high compression ratios and superior write and query performance. On the data writing side, they can handle massive, high-frequency data write requests, while on the data reading side, they enable flexible multi-dimensional queries and correlation analysis.
2. Compatibility with Multiple Text Analysis Engines
Supports text analysis engine databases like OpenSearch and Elasticsearch, achieving unified storage for data structures such as logs, objects, traces, and events. Data is written consistently, efficiently, and with low latency through the unified Dataway interface using the Line Protocol, and is queried and analyzed uniformly using the self-developed query language Debug Query Language (DQL).
Support for the real-time analytical data warehouse-oriented SelectDB text analysis engine can significantly improve data read/write performance and substantially reduce overall storage costs.
Additionally, the latest self-developed GuanceDB 3.0 technology runs fully cloud-native on any commercial object storage. It is designed specifically for handling massive data writes and features a flexible, completely read-write separated architecture.
3. Low-Intrusion Side-Channel Data Collection Technology
Data collection is the initial step for data analysis in a big data platform. Most of Guance's data collectors implement data collection functionality based on side-channel technology, enabling the completion of data collection tasks with minimal impact on business systems.
4. Data Consistency Assurance and High System Reliability
Throughout the entire chain from data collection to data cleansing to data processing, Guance ensures data consistency based on message queue technology and multiple retry mechanisms, compensating for the weaknesses of time-series and columnar databases.
5. Cloud-Native
The entire platform's selection of underlying technology modules and overall architecture is implemented based on cloud-native products, achieving high cost-effectiveness while ensuring functional completeness and technical architecture reliability.
Cloud Vendor Partnerships¶
Guance supports data ingestion from cloud vendors such as AWS, Azure, Alibaba Cloud, Huawei Cloud, Tencent Cloud, Volcengine, and Google Cloud. Regardless of which cloud vendor you are using, you can quickly report data to Guance for unified management and analysis, including Cloud Resource Management and Cloud Billing Analysis.
Furthermore, Guance supports billing through cloud vendors. Simply subscribe to Guance with one click in the cloud vendor's marketplace to start using Guance, paying as you go and using on-demand.