go-demo-proxy

go-demo-proxy

Proxies go-demo, so we can have distributed traces

TODO

• call go-demo from go-demo-proxy
• have query parameters
• implement the same random error and delay as go-demo
  • with both, filter on query paremeter to enable/disable
  • this way, we can "discover" that specific flows are broken
• add distributed tracing
  • include the query parameters in the span
  • include the error in the span
  • include the user agent in the span
• add metrics
• events?

Usage

export GO_DEMO_SERVICE_ONE_URL=http://go-demo.teal.run-01.h2o-2-13047.h2o.vmware.com
export GO_DEMO_SERVICE_TWO_URL=http://go-demo.tsm-2.h2o-2-13047.h2o.vmware.com
export SERVICE_NAME=go-demo-proxy
export SERVICE_VERSION=0.0.1
export WAVEFRONT_PROXY_HOSTNAME=localhost
export WAVEFRONT_PROXY_PORT=2878
export ZIPKIN_HOST=localhost
export ZIPKIN_PORT=9411

Testing for Kubernetes:

export MY_NODE_NAME="test-01"
export MY_CPU_REQUEST="100m"
export MY_CPU_LIMIT="200m"
export MY_MEM_REQUEST="100Mi"
export MY_MEM_LIMIT="200Mi"

Wavefront Proxy

• https://docs.wavefront.com/proxies_installing.html

export WAVEFRONT_URL=https://<INSTANCE>.wavefront.com
export WAVEFRONT_TOKEN=<TOKEN>

docker run -d -p 9411:9411 openzipkin/zipkin

docker run -d\
  -e WAVEFRONT_URL=$WAVEFRONT_URL \
  -e WAVEFRONT_TOKEN=$WAVEFRONT_TOKEN \
  -p 2878:2878 \
  -p 9400:9400 \
  --name wavefront-proxy \
  wavefronthq/proxy:latest