Log Service provides the real-time log consumption feature. This feature reads and writes full data in the first-in, first-out (FIFO) order, which is similar to Kafka. This topic describes the types of real-time log consumption.
After log data is sent to Log Service, you can process the data in the following three
methods.
| Method | Scenario | Timeliness | Retention period |
|---|---|---|---|
| Real-time log consumption | Stream computing and real-time computing | Real time | You can set a retention period based on your business requirements. |
| Log query | Online query of recent hot data | Near real time (delayed for 1 second in 99.9% cases, and up to 3 seconds) | You can set a retention period based on your business requirements. |
| Log shipping | Full log storage for offline analysis | Delayed for 5 to 30 minutes | The retention period is subject to the storage system. |
Note For more information about the differences between reading and querying logs, see
What are the differences between LogHub and LogSearch?.
Real-time log consumption
- Use an SDK to consume log data
Log Service provides SDKs in various programming languages, such as Java, Python, and Go. You can use an SDK to consume log data. For more information about the SDKs, see SDK reference.
- Use consumer groups to consume log data
- Use Function Compute to consume log data
- Consume logs by using cloud services
- Consume logs by using open source services