http-tools

HTTP Mirror

An HTTP server use for debugging and testing with access Combined log printed on stdout. Several component can be reused like MonitoredMux, AccessCombinedLogger

Setup

Build and run locally

Build from make file

make dep http-mirror

Run on default port 8080

./bin/http-mirror

Run on custom port

HTTP_SEVER_PORT=4242 ./bin/http-mirror

Use with docker

Use the latest docker image available on docker hub.

    docker run -d -p 8080:8080 gmalbrand/http-mirror:latest 

Deploy on k8s

The given manifest creates a deployment set with 2 replicas and load balancer service. The pods have prometheus and DataSet annoations for metric collection

    kubectl create namespace "your-namespace"
    kubectl apply -f ./deploy/deployment.yaml -n "your-namespace"

URIs

/metrics

Expose promhttp metrics Custom metrics :

• http_requests_in_flight : A gauge of requests currently being served by the wrapped handler.
• http_requests_total : A counter for requests to the wrapped handler.
• http_request_duration_seconds : A histogram of latencies for requests.
• http_response_size_bytes : A histogram of response sizes for requests.
• http_request_size_bytes : A histogram of requests sizes.

/mirror

Accept all method. Dump the request headers and body in the response. GET example :

curl http://localhost:8080/mirror?param1=value1&param2=value2
GET /dump?param1=value1&param2=value2 HTTP/1.1
Host: localhost:8080
Accept: */*
User-Agent: curl/7.79.1

POST example:

curl -X POST http://localhost:8080/mirror -H 'Content-Type: application/json' -d '{"login":"my_login","password":"my_password"}'
POST /dump HTTP/1.1
Host: localhost:8080
Accept: */*
Content-Length: 45
Content-Type: application/json
User-Agent: curl/7.79.1

{"login":"my_login","password":"my_password"}

/cpuLoad

Generate load on the server

• cpuLoad : generate cpu usage (parameter load, percentage, default 80%) for a period of time (parameter duration, in second, default 10s) Example:

 curl http://localhost:8080/cpuLoad?load=120&duration=60

/memLoad

• memLoad : generate memory usage (parameter size, in MB, default 80) for a period of time (parameter duration, in second, default 10s) Example:

 curl http://localhost:8080/cpuLoad?mem=1024&duration=60

/

Simple HTTP proxy. Almost untested.

Certificate generator

Generate CA cert and keys and self signed certificates restricted to a list of domain names. These certificates can then be used within k8s cluster.

make dep certgen