Zipkin Tracing

The Zipkin tracing sandbox demonstrates Envoy’s request tracing capabilities using Zipkin as the tracing provider. This sandbox is very similar to the front proxy architecture described above, with one difference: service1 makes an API call to service2 before returning a response. The three containers will be deployed inside a virtual network called envoymesh.

All incoming requests are routed via the front envoy, which is acting as a reverse proxy sitting on the edge of the envoymesh network. Port 80 is mapped to port 8000 by docker compose (see /examples/zipkin-tracing/docker-compose.yml). Notice that all envoys are configured to collect request traces (e.g., http_connection_manager/config/tracing setup in /examples/zipkin-tracing/front-envoy-zipkin.json) and setup to propagate the spans generated by the Zipkin tracer to a Zipkin cluster (trace driver setup in /examples/zipkin-tracing/front-envoy-zipkin.json).

Before routing a request to the appropriate service envoy or the application, Envoy will take care of generating the appropriate spans for tracing (parent/child/shared context spans). At a high-level, each span records the latency of upstream API calls as well as information needed to correlate the span with other related spans (e.g., the trace ID).

One of the most important benefits of tracing from Envoy is that it will take care of propagating the traces to the Zipkin service cluster. However, in order to fully take advantage of tracing, the application has to propagate trace headers that Envoy generates, while making calls to other services. In the sandbox we have provided, the simple flask app (see trace function in /examples/front-proxy/service.py) acting as service1 propagates the trace headers while making an outbound call to service2.

Running the Sandbox

The following documentation runs through the setup of an envoy cluster organized as is described in the image above.

Step 1: Build the sandbox

To build this sandbox example, and start the example apps run the following commands:

  1. $ pwd
  2. envoy/examples/zipkin-tracing
  3. $ docker-compose up --build -d
  4. $ docker-compose ps
  5.         Name                       Command               State      Ports
  6. -------------------------------------------------------------------------------------------------------------
  7. zipkintracing_service1_1      /bin/sh -c /usr/local/bin/ ...    Up       80/tcp
  8. zipkintracing_service2_1      /bin/sh -c /usr/local/bin/ ...    Up       80/tcp
  9. zipkintracing_front-envoy_1   /bin/sh -c /usr/local/bin/ ...    Up       0.0.0.0:8000->80/tcp, 0.0.0.0:8001->8001/tcp

Step 2: Generate some load

You can now send a request to service1 via the front-envoy as follows:

  1. $ curl -v $(docker-machine ip default):8000/trace/1
  2. *   Trying 192.168.99.100...
  3. * Connected to 192.168.99.100 (192.168.99.100) port 8000 (#0)
  4. > GET /trace/1 HTTP/1.1
  5. > Host: 192.168.99.100:8000
  6. > User-Agent: curl/7.43.0
  7. > Accept: */*
  8. >
  9. < HTTP/1.1 200 OK
  10. < content-type: text/html; charset=utf-8
  11. < content-length: 89
  12. < x-envoy-upstream-service-time: 1
  13. < server: envoy
  14. < date: Fri, 26 Aug 2016 19:39:19 GMT
  15. < x-envoy-protocol-version: HTTP/1.1
  16. <
  17. Hello from behind Envoy (service 1)! hostname: f26027f1ce28 resolvedhostname: 172.19.0.6
  18. * Connection #0 to host 192.168.99.100 left intact

Step 3: View the traces in Zipkin UI

Point your browser to http://192.168.99.100:9411 . You should see the Zipkin dashboard. If this ip address is incorrect, you can find the correct one by running: $ docker-machine ip default. Set the service to “front-proxy” and set the start time to a few minutes before the start of the test (step 2) and hit enter. You should see traces from the front-proxy. Click on a trace to explore the path taken by the request from front-proxy to service1 to service2, as well as the latency incurred at each hop.