Workers

The stresstest uses workers to execute the queries, which are supposed to simulate users. Each worker has its own set of queries and executes them parallel to the other workers.

Iguana supports multiple worker types, but currently only the SPARQLProtocolWorker is implemented. Workers have the common type property which defines the type of the worker.

tasks:
  - type: "stresstest"
    workers:
    - type: "SPARQLProtocolWorker"
      # properties of the worker
      # ...
    - type: "SPARQLProtocolWorker"
      # properties of the worker
      # ...

SPARQLProtocolWorker

The SPARQLProtocolWorker is a worker that sends SPARQL queries to an endpoint using the SPARQL protocol. The worker can be configured with the following properties:

property	required	default	description
number	no	`1`	The number of workers that should be initiated with that same configuration.
queries	yes		The configuration of the query handler these workers should use. (see here)
completionTarget	yes		Either defines how many queries the worker should send, or how long it should send them.
connection	yes		The name of the connection that the worker should use. (needs to reference an already defined connection)
timeout	yes		The duration for the query timeout.
acceptHeader	no		The accept header that the worker should use for the HTTP requests.
requestType	no	`get query`	The request type that the worker should use.
parseResults	no	`true`	Whether the worker should parse the results.

Each property is explained in more detail below.

Example

connection:
  - name: "fuseki"
    dataset: "sp2b"
    endpoint: "http://localhost:3030/sp2b"

tasks:
  - type: "stresstest"
    workers:
    - type: "SPARQLProtocolWorker"
      number: 2                       # two workers with the same configuration will be initiated
      queries:                        # the query handler configuration, both workers will use the same query handler
        path: "./example/suite/queries/"
        format: "folder"
      completionTarget:
        number: 1                     # each query will be executed once
      connection: "fuseki"            # the worker will use the connection with the name "fuseki", which is defined above
      timeout: "2S"
      acceptHeader: "application/sparql-results+json"
      requestType: "get query"
      parseResults: true

Number

The number property defines the number of workers that should be initiated with the same configuration. Workers with the same configuration will use the same query handler instance.

Queries

The queries property is the configuration of the query handler that the worker should use. The query handler is responsible for loading and selecting the queries that the worker should execute. The query handler configuration is explained in more detail here.

Completion Target

The completionTarget property defines when the worker should stop executing queries. The property takes an object as its value that contains either one of the following properties: - number: The number of times the worker should execute each query. - duration: The duration during which the worker should iterate and execute every query.

Example:

tasks:
  - type: "stresstest"
    workers:
    - type: "SPARQLProtocolWorker"
      number: 1
      completionTarget:
        number: 100 # execute each query 100 times
      # ...
    - type: "SPARQLProtocolWorker"
      number: 1
      completionTarget:
        duration: "10s" # execute queries for 10 seconds
      # ...

Timeout

The timeout property defines the maximum time a query execution should take, this includes the time it takes to send the request and to receive the response. If the timeout is reached, the worker will mark it as failed, cancel the HTTP request and continue with the execution of the next query.

The system that's being tested should make sure that it's able to abort the further execution of the query if the timeout has been reached (e.g., by using a timeout parameter for the system, if available). Otherwise, problems like high resource usage or other issues might occur.

Request Type

The requestType property defines the type of the HTTP request that the worker should use. It consists of a string that can be one of the following values:

request type	HTTP method	Content-Type header value	description
`"get query"`	`GET`		The worker will send a `GET` request with a `query` parameter that contains the query.
`"post query"`	`POST`	`application/sparq-query`	The body will contain the query.
`"post update"`	`POST`	`application/sparq-update`	The body will contain the update query.
`"post url-enc query"`	`POST`	`application/x-www-form-urlencoded`	The body will contain the a url-encoded key-value pair with the key being `query` and the query as the value.
`"post url-enc update"`	`POST`	`application/x-www-form-urlencoded`	The body will contain the a url-encoded key-value pair with the key being `update` and the query as the value.

Accept Header

The acceptHeader property defines the value for the Accept header of the HTTP requests that a worker sends to the defined endpoint. This property also affects the Response-Body-Processors that are used to process the response bodies.

Parse Results

The parseResults property defines whether the worker should parse the results of the queries. If the property is set to true, the worker will send the response body to the Response-Body-Processors for processing and calculate hash values for the response bodies. If the property is set to false, the worker will not parse the response bodies and will not calculate hash values for the response bodies.

Setting the property to false can improve the performance of the worker. This means that the worker is able to measure the performance more accurately. If the property is set to true, the worker will temporarily store the whole response bodies in memory for processing. If the property is set to false, the worker will discard any received bytes from the response.