http

Http runner service

Http runner sends one or more HTTP request to the specified endpoint; it manages cookie within SendRequest.

Usage

• Basic and conditional requests
• Repeating request
• User defined function transformation
• Sequential request with data extraction
• Request with validation
• Testing http request from cli
• Sending http request from inline workflow
• Stress testing
• Data organization

Basic and conditional requests execution

	import (
		"log"
		"net/http"
		"github.com/viant/endly"
		"github.com/viant/toolbox"
		runner "github.com/viant/endly/service/testing/runner/http"
	)
	
	
	func main() {
        response := &runner.SendResponse{}
        
        err := endly.Run(context, &runner.SendRequest{
            Options:[]*toolbox.HttpOptions{
                 {
                    Key:"RequestTimeoutMs",
                    Value:12000,
                 },	
            },
            Requests: []*runner.Request{
                {
                    URL:    "http://127.0.0.1:8111/send1",
                    Method: "POST",
                    Body:   "some body",
                    Header:http.Header{
                        "User-Agent":[]string{"myUa"},
                    },
                },
                {   //Only run the second request in previous response body contains 'content1-2' fragment
                    When:   "${httpTrips.Response[0].Body}:/content1-2/",
                    URL:    "http://127.0.0.1:8111/send2",
                    Method: "POST",
                    Body:   "other body",
                },
            },
        }, response)
        if err != nil {
            log.Fatal(err)
        }
        
    }

A send request represents a group of http requests, Individual request run can be optionally conditioned with When criteria Each run publishes 'httpTrips' journal to the context.State with .Response and .Request keys representing collection about the active execution.

Repeating request

     response := &runner.SendResponse{}
     err := endly.Run(context, &runner.SendRequest{
            Requests: []*runner.Request{
    			{
    				URL:    "http://127.0.0.1:8111/send1",
    				Method: "POST",
    				Body:   "0123456789",
    				Repeater: &model.Repeater{
    					Repeat: 10,
    					
    				},
    			}
  			}
	})

Simply repeat 10 times post request

@http.json

{
    "Requests": [
        {
            "URL": "http://testHost/ready",
            "Method": "GET",
            "Extraction": [
                {
                    "Key": "testHostStatus",
                    "RegExpr": "Current Server Status: ([A-Z]*)"
                }
            ],
            "Repeat": 150,
            "SleepTimeMs": 2000,
            "Exit": "$testHostStatus: READY"
        }
    ]
}

Repeat till test host status is 'READY', keep testing for status no more than 150 times with 2 second sleep.

User defined function transformation

    registerUDF, err := udf.NewRegisterRequestFromURL("udf.json")
    if err != nil {
      	log.Fatal(err)
    }
    err = endly.Run(context, registerUDF, nil)
    if err != nil {
    	log.Fatal(err)
    }
    request, err := runner.NewSendRequestFromURL("http.json")
    if err != nil {
    	log.Fatal(err)
    }
	var response = &runner.SendResponse{}
	err = endly.Run(context, request, response)
	if err != nil {
		log.Fatal(err)
		return
	}

@http.json

{
  "Requests": [
    {
      "Method": "post",
      "URL": "http://127.0.0.1:8987/xxx?access_key=abc",
      "RequestUdf": "UserAvroWriter",
      "ResponseUdf": "AvroReader",
      "JSONBody": {
        "ID":1,
        "Desc":"abc"
      }
    }
  ] 
}

@udf.json

{
 "Udfs": [
    {
      "Id": "UserAvroWriter",
      "Provider": "AvroWriter",
      "Params": [
        "{\"type\": \"record\", \"name\": \"user\", \"fields\": [{\"name\": \"ID\",\"type\":\"int\"},{\"name\": \"Desc\",\"type\":\"string\"}]}"
      ]
    }
  ]
}

See more how to register common codec UDF (avro, protobuf) with custom schema

Sequential request with data extraction

@http.json

{
	"Requests": [
		{
			"Method": "POST",
			"URL": "http://${bidderHost}/bidder",
			"Body": "$bid0",
			"Extraction": [
            				{
            					"Key": "winURI",
            					"RegExpr": "(/pixel/won[^\\']+)",
            					"Reset": true
            				},
            				{
            					"Key": "clickURI",
            					"RegExpr": "(/pixel/click[^\\']+)",
            					"Reset": true
            				}
            ],
			"Variables": [
				{
					"Name": "AUCTION_PRICE",
					"From": "seatbid[0].bid[0].price"
				},
                {
                    "Name": "winURL",
                    "Value": "http://${loggerHost}/logger${winURI}"
                },
                {
                    "Name": "clickURL",
                    "Value": "http://${loggerHost}/logger${clickURI}"
                }
			]
		},
		{
			"When": "${httpTrips.Response[0].Body}://pixel/won//",
			"URL": "${httpTrips.Data.winURL}",
			"Method": "GET"
		},
		{
			"When": "${httpTrips.Response[0].Body}://pixel/click//",
			"URL": "${httpTrips.Data.clickURL}",
			"Method": "GET"
		}
	]
}

Where $bidX is defined in separated file stichted with Neatly

@bids.json

{
	"bid0": {
		"app": {
			"cat": [
				"IAB14"
			],
			"domain": "http://www.wildisthewind.com/",
			"id": "yyyyy",
			"name": "RTB TEST",
			"publisher": {
				"id": "xxxxxx",
				"name": "Match Media Group"
			}
		},
		"at": 2,
		"badv": [
			"go-text.me/"
		],
		"bcat": [
			"IAB7-39"
		],
		"device": {
			"carrier": "310-410",
			"devicetype": 1,
			"dpidsha1": "${userProfile.uuid}",
			"ext": {
				"idfa": "${userProfile.uuid}"
			},
			"geo": {
				"city": "Whitewright",
				"country": "USA"
			},
			"language": "en",
			"make": "Apple",
			"model": "iPhone",
			"os": "iOS",
			"osv": "6.1",
			"ua": "Mozilla/5.0 (iPhone; U; CPU iPhone 6_1 like Mac OS X; en-us) AppleWebKit/528.18 (KHTML, like Gecko) Mobile/7E18 Grindr/1.8.5 (iPhone3,1/6.1)"
		},
		"imp": [
			{
				"banner": {
					"api": [
						3
					],
					"btype": [
						4
					],
					"h": 300,
					"pos": 1,
					"w": 250
				}
			}
		]
	},
	"bid1": {
		
	}
}

The following example extracts from POST Body tracking URL with regular expression matches and from structured POST Body AUCTION_PRICE and variables that define subsequent URL.

Request with validation

@send.json

{
  "Requests": [
    {
      "Method": "POST",
      "URL": "http://127.0.0.1:8080/v1/api/dummy",
      "Body": "{}"
    }
  ],
  "Expect": {
    "Responses": [
      {
        "Code": 200,
        "JSONBody": {
          "Status": "error",
          "Error": "data was empty"
        }
      }
    ]
  }
}

Testing http request from cli

endly -w=action service='http/runner' action=send request='@send.json'

@send.json

{
  "Requests": [
    {
      "Method": "POST",
      "URL": "http://127.0.0.1:8080/uri",
      "JSONBody": {
        "key1":1
      }
    }
  ]
}

Sending http request from inline workflow

endly -r=http

@http.yaml

pipeline:
  task1:
    action: http/runner:send
    request: "@send.json" 

Sending http request and validation

endly -r=http

@http.yaml

pipeline:
  task1:
    action: http/runner:send
    requests:
      - url: http://www.wp.pl
        expect:
          Code: 200

Sending http request with custom http client option

endly -r=http_with_options

@http_with_options.yaml

pipeline:
  task1:
    action: http/runner:send
    options:
      FollowRedirects: false
      TimeoutMs: 3000 
        
    requests:
      - url: http://www.wp.pl
        expect:
          Code: 301

Supported options with defaults:

• RequestTimeoutMs = 30 * time.Second
• KeepAliveTimeMs = 30 * time.Second
• TLSHandshakeTimeoutMs = 10 * time.Second
• ExpectContinueTimeout = 1 * time.Second
• IdleConnTimeout = 90 * time.Second
• DualStack = true
• MaxIdleConnsPerHost = http.DefaultMaxIdleConnsPerHost
• MaxIdleConns = 100
• FollowRedirects = true
• ResponseHeaderTimeoutMs time.Duration
• TimeoutMs time.Duration

Stress testing

HTTP runner provides stress testing capabilities to generate a HTTP endpoint load and to validate desired responses.

endly -r=load_test

@load_test.yaml

init:
  testEndpoint: 127.0.0.1:8988
pipeline:
  startEndpoint:
    action: http/endpoint:listen
    port: 8988
    rotate: true
    baseDirectory: test/stress
  init:
    action: print
    message: starting load testing
  loadTest:
    action: 'http/runner:load'
    '@repeat': 100000
    assertMod: 16
    threadCount: 10
    options:
      TimeoutMs: 500
    requests:
      - Body: '000'
        Method: POST
        URL: http://${testEndpoint}/send0
        Expect:
          Body: '1000'
          Code: 200

      - Body: '111'
        Method: POST
        URL: http://${testEndpoint}/send1
        Expect:
          Body: '1111'
          Code: 200

  summary:
    action: print
    message: 'Count: $loadTest.RequestCount, QPS: $loadTest.QPS: Response: min: $loadTest.MinResponseTimeInMs ms, avg: $loadTest.AvgResponseTimeInMs ms max: $loadTest.MaxResponseTimeInMs ms, errors: $loadTest.ErrorCount, timeouts: $loadTest.TimeoutCount'

Bulk requests loading for stress testing

The following workflow provide example how to bulk load request and desired response for stress testing

@regression.yaml

init:
  testEndpoint: x.vindicosuite.com
pipeline:
  test:
    tag: Test
    subPath: use_cases/${index}*
    data:
      ${tagId}.[]Requests: '@data/*request.json'
      ${tagId}.[]Responses: '@data/*response.json'
    range: '1..002'
    template:
      info:
        action: print
        message: load testing $subPath
      load:
        action: 'http/runner:load'
        request: '@req/load'
        init:
          requests: ${data.${tagId}.Requests}
          expect:   ${data.${tagId}.Responses}
      load-info:
        action: print
        message: '$load.QPS: Response: min: $load.MinResponseTimeInMs ms, avg: $load.AvgResponseTimeInMs ms max: $load.MaxResponseTimeInMs ms'

Where

• @load.json

• regression is a folder with the following structure

endly -r=regression

Data organization

Imagine a case where Payload body can be shared across various HTTP requests within the same group.

@http_send.json

{
	"Requests": [
		{
			"Method": "POST",
			"URL": "http://${testHost}/uri",
			"Body": "$payload1"
		},
		{
            "Method": "POST",
            "URL": "http://${testHost}/uri",
            "Body": "$payload2"
        },
        {
            "Method": "POST",
            "URL": "http://${testHost}/uri",
            "Body": "$payload1"
        }
	]
}

@payloads.json

{
  "payload1": {
    "k1":"some data here"
  },
  "payload2": {
      "k100":"some data here"
  }
}

You can use the multi resource loading:

• inline workflow format

@test.yaml

pipeline:
  task1:
    action: http/runner:send
    request: "@send.json @payloads"