Collect Trace with Zipkin
Zipkin
Distributed tracing is a method used to profile and monitor applications built with microservices architecture. Distributed tracing helps pinpoint where failures occur and what causes poor performance.
Zipkin is a distributed tracing system that helps collect timing data needed to troubleshoot latency problems in service architectures. EventMesh exposes a collection of trace data that could be collected and analyzed by Zipkin. Please follow the “Zipkin Quickstart” tutorial to download and install the latest release of Zipkin.
Configuration
To enable the trace exporter of EventMesh Runtime, set the eventMesh.server.trace.enabled
field in the conf/eventmesh.properties
file to true
.
# Trace plugin
eventMesh.server.trace.enabled=true
eventMesh.trace.plugin=zipkin
To customize the behavior of the trace exporter such as timeout or export interval, edit the exporter.properties
file.
# Set the maximum batch size to use
eventmesh.trace.max.export.size=512
# Set the queue size. This must be >= the export batch size
eventmesh.trace.max.queue.size=2048
# Set the max amount of time an export can run before getting(TimeUnit=SECONDS)
eventmesh.trace.export.timeout=30
# Set time between two different exports (TimeUnit=SECONDS)
eventmesh.trace.export.interval=5
To send the exported trace data to Zipkin, edit the eventmesh.trace.zipkin.ip
and eventmesh.trace.zipkin.port
fields in the conf/zipkin.properties
file to match the configuration of the Zipkin server.
# Zipkin's IP and Port
eventmesh.trace.zipkin.ip=localhost
eventmesh.trace.zipkin.port=9411