This project implements an utility mock server to be used in Load Tests. It implements some HTTP endpoints to return custom responses, and with custom delays.
This server could be quite useful for testing load balancers or reverse proxies such as Nginx, Varnish, or HAProxy, where we need to configure one or many backend servers to forward the requests.
The server is implemented in Java using Vert.x, an asynchronous and non-blocking library, so it should be able to handle thousands of simultaneous connections without any problem.
Below we will describe the API provided by the Load Test Server.
Basic Health Check endpoint that just returns 200 OK, with the text "OK".
Example request:
curl -v http://localhost:8080/health
Example response:
< HTTP/1.1 200 OK
< content-type: text/plain
< content-length: 2
<
OK
Basic endpoint with a "ping-pong" functionality. It just returns 200 OK with the same
body and Content-Type received in the request.
Example request:
curl http://localhost:8080/ping -v \
-H 'Content-Type: text/plain' \
-d 'Hello world!'
Example response:
< HTTP/1.1 200 OK
< transfer-encoding: chunked
< content-type: text/plain
< delay: 0
<
Hello world!
As you can see in the example, by default the server returns the response with no
delays (delay: 0). We can change this default behaviour passing optional headers
in the request:
delay: <delay>: This header will force the server to delay/sleep for<delay>milliseconds before returning the response.random-delay: [<minDelay>,]<maxDelay>: This header will force the server to delay/sleep for a random period of time, between<minDelay>and<maxDelay>milliseconds, both inclusive.<minDelay>is optional, and it defaults to 0.Accept-Encoding: gzip | deflate: This header will force the server to return the response compressed with gzip or DEFLATE respectively.response-<header_name>: <header_value>: This header in the request will force a header of type<header_name>: <header_value>in the response.
Example request, asking for a compressed response, with a Cache-Control: public header, and a random
delay between 500ms and 2000ms:
curl http://localhost:8080/ping -v \
-H 'Content-Type: application/json' \
-H 'random-delay: 500,2000' \
-H 'response-cache-control: public' \
-H 'Accept-Encoding: gzip' \
-d '{"message": "Hello world!"}' \
--compressed
Example response:
< HTTP/1.1 200 OK
< delay: 703
< content-type: application/json
< cache-control: public
< content-encoding: gzip
< transfer-encoding: chunked
<
{"message": "Hello world!"}
GET:/data and POST:/data are generic endpoints that return random responses of
a given size/length. In the POST version, the body passed in the request is ignored.
Example request:
curl http://localhost:8080/data -v
Example response:
< HTTP/1.1 200 OK
< content-type: text/plain
< delay: 0
< data-length: 1024
< content-length: 1024
<
f4~MXvvyGw+u(*?CE**,+;}r1&k#fJCUyifH...
As you can see in the example, by default the server returns 1024 random characters
(data-length: 1024) with no delays (delay: 0). We can change this default behaviour
passing the following headers in the request:
delay: <delay>: This header will force the server to delay/sleep for<delay>milliseconds before returning the response.random-delay: [<minDelay>,]<maxDelay>: This header will force the server to delay/sleep for a random period of time, between<minDelay>and<maxDelay>milliseconds, both inclusive.<minDelay>is optional, and it defaults to 0.Accept-Encoding: gzip | deflate: This header will force the server to return the response compressed with gzip or DEFLATE respectively.data-length: <length>: This header indicates the desired length of the response's body, so the server will reply with<length>random characters.random-data-length: [<minLength>,]<maxLength>: This header indicates that we want a respose with a random length between<minLength>and<maxLength>, both inclusive.<minLength>is optional, and it defaults to 0.response-<header_name>: <header_value>: This header in the request will force a header of type<header_name>: <header_value>in the response.
Example request, asking for a compressed response, with a Cache-Control: private header, with a length
between 2000 and 5000 random characters, and with a delay of maximum 5000ms:
curl http://localhost:8080/data -v \
-H 'random-data-length: 2000,5000' \
-H 'random-delay: 5000' \
-H 'response-cache-control: private' \
-H 'Accept-Encoding: gzip' \
--compressed
Example response:
< HTTP/1.1 200 OK
< delay: 3890
< content-type: text/plain
< data-length: 3240
< cache-control: private
< content-encoding: gzip
< transfer-encoding: chunked
<
@l?}|:"|](I]AY1T>{G_...
Build with Java 11 and Maven 3+.
mvn clean install
This will generate a "fat JAR" file, containing the application and all its dependencies.
This will also build a docker image locally.
Once the application is built, we just need to execute the generated fat JAR file:
Example:
java -jar target/load-test-server-1.0-SNAPSHOT-app.jar
We should see something like this:
2020-05-25T20:14:50,301+0200 [main] INFO Main - Operating System: Mac OS X (10.15.4)
2020-05-25T20:14:50,304+0200 [main] INFO Main - Number of Processors: 12
2020-05-25T20:14:50,304+0200 [main] INFO Main - Max memory for application: 3,641.00 MB
2020-05-25T20:14:50,355+0200 [main] INFO Main - Max open files: 10240 (soft) , unlimited (hard)
2020-05-25T20:14:50,356+0200 [main] INFO Main - Number of event loop threads: 24
2020-05-25T20:14:50,357+0200 [main] INFO Main - Deploying 48 ServerVerticle instances
2020-05-25T20:14:50,618+0200 [vert.x-eventloop-thread-6] INFO ServerVerticle - Registered endpoint GET:/health
2020-05-25T20:14:50,619+0200 [vert.x-eventloop-thread-6] INFO ServerVerticle - Registered endpoint POST:/ping
2020-05-25T20:14:50,619+0200 [vert.x-eventloop-thread-6] INFO ServerVerticle - Registered endpoint GET:/data
2020-05-25T20:14:50,619+0200 [vert.x-eventloop-thread-6] INFO ServerVerticle - Registered endpoint POST:/data
2020-05-25T20:14:50,641+0200 [vert.x-eventloop-thread-6] INFO ServerVerticle - Listening on 0.0.0.0:8080 ...
Optionally, you can pass the following parameters when starting the application:
-p <port>: Listening port for the server. Default8080.