This topic describes the features of Log Service.

Fully managed services

  • Log Service can connect data sources within five minutes and use the Logtail agent to collect data over all networks.
  • LogHub provides all features of Kafka, such as data monitoring and alerts. It also scales automatically to handle petabytes of data each day, reducing more than 50% of costs compared with self-built systems.
  • The log search and analytics feature provides dashboards, saved searches, and alerts that can reduce your costs by over 80% compared with self-built systems.
  • Log Service allows you import data from over 30 data sources such as OSS, E-MapReduce, MaxCompute, Tablestore, Message Service, CDN, ARMS, and open source software such as Storm and Spark.

Comprehensive ecosystem

  • LogHub supports more than 30 log collectors, such as Logstash and Fluent. This allows you to collect logs from embedded devices, web applications, servers, and programs. LogHub can also be integrated with CloudMonitor, Storm, Spark Streaming, and other services for data consumption.
  • LogShipper supports user-defined partitions and various data formats such as text files, sequence files, and parquet files. Data can be processed directly by Presto, Hive, Spark, Hadoop, E-MapReduce, MaxCompute, HybridDB for MySQL, and Data Lake Analytics.
  • The log search and analytics feature provides complete query and analysis syntax. It is compatible with the SQL-92 standard and can integrate with Grafana through the JDBC protocol.


  • LogHub allows real-time consumption upon collection. Logtail collects and uploads data to Log Service within a second, with an availability of 99.9%.
  • Log search and analytics: Data can be queried and analyzed in real time upon collection. Billions of log entries can be queried per second with multiple query conditions. Hundreds of millions of log entries can be analyzed per second with multiple aggregate conditions.

Complete API and SDK implementations

  • Log Service supports custom management and secondary development.
  • All Log Service features can be implemented by using APIs and SDKs. SDKs developed in multiple programming languages enable you to easily manage services and millions of devices.