Skip to content

Latest commit

 

History

History
52 lines (36 loc) · 2.35 KB

README.md

File metadata and controls

52 lines (36 loc) · 2.35 KB
Raw

Logging

Logging support is a one of the most important parts for observability platforms. But logging has a long history in the history of application, it came out much earlier than distributed tracing and metrics, there have been a lot of logging solutions.

To better cooperate with existing logging systems, Bithon does not collect the application log to save it in a central place so that they can be searched by giving conditions. Instead, Bithon only injects the trace id for a request to the log. Existing logging system can still work without any change, but by the injection, users can search the log by given trace id.

This document describes the logging support.

Rationale

Bithon injects trace id to logs automatically, it requires the target application running with log4j2 or logback or slf4j(without any Maven Shading). When target application starts, once Bithon detects these logging libraries are integrated, it will first insert the trace id pattern to the existing user define logging pattern. When an HTTP request/Kafka message arrives, Bithon will generate a trace id for such request(whenever this request is sampled or not by the samplingRate configuration). Such trace id will be put into the MDC context, and when any log is output, the logging library will cooperate the trace id into the log message.

For example,

2023-04-22 23:39:54.876  INFO 59110 --- [nio-9897-exec-4] o.b.s.s.jdbc.metric.MetricJdbcReader     : [bTxId: bcde72610a4742d78841757fa55a8636, bSpanId: e6e6f12700000061, bMode: L] Executing SQL...

in above output, the message [bTxId: bcde72610a4742d78841757fa55a8636, bSpanId: e6e6f12700000061, bMode: L] is automatically injected by Bithon, indicating that

  • the trace id is bcde72610a4742d78841757fa55a8636
  • the span id is e6e6f12700000061
  • the mode L means this trace id is a logging-only trace id, it's not recorded by the distributed tracing
    • if the mode is T, it indicates that the trace id is also an id of distributed tracing, it can be used to search the tracing log

Features

Supported logging framework

  • log4j2
  • logback
  • slf4j

Supported Components

Following entries are supported to initialize a trace id for logging.

  • Spring Web Server, including
    • Tomcat Web Server
    • Undertow Web Server
    • Jetty Web Server
    • Web Flux
  • Spring Kafka Listener
  • Spring Scheduler
  • Quartz2