If you place a balancer in front of KrakenD, such as an ELB, you can check KrakenD health using a TCP port check. If, on the other hand, you need to have an HTTP endpoint like /health or /ping in systems like Kubernetes, you can do it in different ways.
Although there is no default health check implementation, the result can be achieved using different strategies. For instance:
- Add a
/healthendpoint in the configuration with stub data (see example below) - Enable the
/__debug/endpoint with your desired log level and use it as the health check. - Use a static Martian modifier and serve any static file as
/health. - Pass the
/healthto another backend (although you are checking the backend, not KrakenD)
Adding a /health endpoint with stub data
The easiest option is to use the static proxy functionality. With this technique, we are going to add an endpoint that uses an unreachable (and fake) backend. As the data can’t be fetched from KrakenD, the strategy always is going to make sure we return the desired static data that we want.
{
"version": 2,
"port": 8080,
"endpoints": [
{
"endpoint": "/health",
"extra_config": {
"github.com/devopsfaith/krakend/proxy": {
"static": {
"data": {
"status": "OK"
},
"strategy": "always"
}
}
},
"backend": [
{
"url_pattern": "/",
"host": [
"http://fake-backend"
]
}
]
}
]
}
In the previous example we are returning a { "status": "OK" } every time the /health endpoint is called:
- curl http://localhost:8080/health
- {"status":"OK"}