Sleuth + Zipkin

1. sleuth What is it? ？

• Spring Cloud Sleuth It provides a distributed tracking solution for microservices .
• Sleuth For distributed link tracing, only trace data is generated , These data are not easy to read , So the tracking data is usually uploaded to Zipkin Server, from Zipkin adopt UI The interface displays data uniformly .

2. sleuth What makes up ？

• span（ span ）： Basic work unit . Use one 64 Bit id Unique identification . except id Outside ,span There are other data , For example, describe 、 Timestamp event 、 Comments on key value pairs （ label ）,spanId、parentId etc. .

• trace（ track ）：trace It's a group. span A tree structure of composition , Use one 64 Bit id Unique identification ,trace All in span Share the trace Of id.

• traceId ： Assigned to a request ID Number , Used to identify a request link .

• spanId ： A basic unit of work , A request can contain multiple steps , Each step has its own spanId. A request contains a TraceId, Multiple SpanId.

3. Zipkin What is it? ？

• Zipkin It is a distributed real-time data tracking system . Its main function is to collect the time series data of the system , So as to track the system delay of the microservice architecture , Monitoring and tracking link calls ,Zipkin It also provides a very friendly UI Interface , To help analyze the tracking data .
• Zipkin It provides pluggable data storage mode ：In-Memory、MySql、Cassandra as well as Elasticsearch.

4. zipkin What makes up ？

• Collector： Collector as , It is mainly used to process tracking information sent from external systems , Convert this information to Zipkin Internally processed Span Format , To support subsequent storage 、 analysis 、 Display and other functions .

• Storage： Storage component , It mainly processes the trace information received by the collector , This information is stored in memory by default , We can also modify this storage policy , Store trace information into the database by using other storage components .

• RESTful API：API Components , It is mainly used to provide external access interface . For example, display tracking information to clients or access external system to realize monitoring, etc .

• Web UI：UI Components , be based on API Upper application of component implementation , adopt UI Component users can easily and intuitively query and analyze tracking information .

5. Sleuth + Zipkin Implement the distributed link tracking process ？

• Zipkin Divided into two ends , One is Zipkin Server side , One is Zipkin client , The client is also the application of microservices . The client will configure the server URL Address , Once a call between services occurs , It will be configured in microservices Sleuth Monitor monitoring of , And generate the corresponding Trace and Span Send information to the server （ There are two main ways to send , One is HTTP Message mode , There is also a way of message bus, such as ：RabbitMQ）, Final ZipKin adopt UI The interface displays the data .

• summary ：Sleuth and Zipkin It's like Spring Boot Actuator and Spring Boot Admin The relationship between , One for generating data , One for displaying data .

6. Sleuth + Zipkin Implementation of distributed link tracking details ？

• stay zuul The gateway implements a OncePerRequestFilter filter , Inject Zipkin Of Tracer take traceId and spanId Add to response header , It is convenient to view the interface call link

 String traceId = tracer.currentSpan().context().traceIdString();
 String spanId  = tracer.currentSpan().context().spanIdString();
 httpServletResponse.setHeader("traceId",traceId+","+spanId);