Typical scenarios of Log Service include data collection, real-time computing, data warehousing and offline analysis, product operation and analysis, and Operation & Maintenance (O&M) and management. This document introduces some typical scenarios. For more scenarios, see Best practices. 

Data collection and consumption

The LogHub function of Log Service enables access to massive real-time log data (including Metric, Event, BinLog, TextLog, and Click data) at the lower costs. 

Advantages of the solution:

  • Easy to use: Over 30 real-time data collection methods are provided for you to quickly build your platform. The powerful configuration and management capabilities can ease O&M workload. Nodes are available across China and the rest of the world.
  • Auto scaling: It helps easily cope with traffic peaks and business growth. 
Figure 1. Data collection and consumption 


ETL/Stream Processing 

LogHub can interconnect with various real-time computing and services, provides complete progress monitoring and alarm notification functions, and supports SDK/API-based custom consumption. 

  • Easy to operate: It provides various SDKs and programming frameworks and can interconnect with various stream computing engines seamlessly. 
  • Comprehensive functions: Rich monitoring data and delay alarm functions are provided.
  • Auto scaling: PB-grade elasticity and zero latency. 
Figure 2. Data cleaning and Flow Calculation 


Data warehouse 

LogShipper ships LogHub data to storage services and supports various storage formats such as compression, user-defined partitions, row storage, and column storage. 

  • Massive data: No upper limit is configured for the amount of data.
  • Rich storage formats: Various storage formats are supported, such as row storage, column storage, and TextFile. 
  • Flexible configuration: Configurations such as user-defined partitions are supported. 
Figure 3. Data Warehouse docking 


Real-time query and analysis of logs

LogAnalytics supports indexing LogHub data in real time and provides rich query methods such as keywords, fuzzy match, context, range, and SQL aggregation. 

  • Strong real-timeliness: Data can be queried after being written. 
  • Massive amount and low cost: Supports PB/day indexing capabilities, and the cost is 15% of the self-built solution. 
  • Strong analysis capabilities: Supports multiple query methods. Supports SQL aggregation and analysis. Visualization and alarm notification functions are provided. 
Figure 4. Real-time query and analysis of logs