Eclipse Ditto：Eclipse IoT 的数字孪生框架 - 主存储库

The benchmark-tool is used to run benchmark tests on ditto. It is implemented using the load-testing tool [k6](https://github.com/grafana/k6) with the [xk6-kafka extension](https://github.com/mostafa/xk6-kafka). [MMock](https://github.com/jmartin82/mmock) is used to mock http responses, it exposes 2 endpoints that are configurable via mmock/default.yaml and mmock/live_messages.yaml. Default values are /:thingId and /live_messages and they are used to publish [modified twin events](https://eclipse.dev/ditto/protocol-specification-things-create-or-modify.html#event-1) and [device live messages](https://eclipse.dev/ditto/protocol-twinlive.html#live) sent via HTTP POST request. ###### The benchmark test consists of 4 tests available to run, called scenarios: - **READ_THINGS** - read things via HTTP ( get things by id ) - **SEARCH_THINGS** - search things via HTTP ( get things by applying search filter ) - **MODIFY_THINGS** - Modify things by sending ditto protocol [modify messages](https://eclipse.dev/ditto/protocol-specification-things-create-or-modify.html#create-or-modify-a-thing) to specfic kafka topic. Ditto kafka connection is reading from this topic and creates [modify commands](https://eclipse.dev/ditto/basic-signals-command.html#modify-commands). [Ditto HTTP push connection](https://eclipse.dev/ditto/connectivity-protocol-bindings-http.html) is configured in ditto, which forwards the [modified twin events](https://eclipse.dev/ditto/protocol-specification-things-create-or-modify.html#event-1) from topic **/things/twin/events?filter=eq(topic:action,'modified')** to a monster mock endpoint, which replies with HTTP status code 204. - **DEVICE_LIVE_MESSAGES** - Send [live messages](https://eclipse.dev/ditto/protocol-twinlive.html#live) to things via HTTP. [Ditto HTTP push connection](https://eclipse.dev/ditto/connectivity-protocol-bindings-http.html) connection is configured in ditto, which sends events from topic **/things/live/messages** to a monster mock endpoint, which replies with predefined ditto protocol message. Also, there is a special scenario called **WARMUP**, which is used to warmup the system, by executing a single read request for each thing, which will cause them to get cached. # Getting started: ## K6 is configurable via environment variables and the following must be set, in order to run the test(sample variables in [test-local.env](https://github.com/eclipse-ditto/ditto/blob/master/benchmark-tool/test-local.env) file): ## K6 test related | Name | Description | | ----------------------------- | ----------------------------------------------------------------------------------------------------------------------------------------------------------- | | SETUP_TIMEOUT | Specify how long the k6 setup() function is allow to run before it's terminated | | TEARDOWN_TIMEOUT | Specify how long the k6 teardown() function is allowed to run before it's terminated | | THINGS_NAMESPACE | Namepsace to use for created ditto things | | THINGS_ID_TEMPLATE | Things id template, f.e. 'my-thing-{0}' | | THINGS_START_INDEX | Things start index, used in template, f.e.  if start index is 1, created things have name of: 'my-thing-1', 'my-thing-2', etc. | | CREATE_THINGS_BATCH_SIZE | Creating things is done via implicitThingCreationMapper via kafka connection. The kafka messages are sent in batches and this variable sets the batch size. | | CREATE_THINGS | If the test should create things, before executing the scenarios (0/1) | | DELETE_THINGS | If the test should delete the things, after executing the scenarios (0/1) | | KAFKA_PRODUCER_LOGGER_ENABLED | K6 kafka producer logger enabled (0/1) | | KAFKA_CONSUMER_LOGGER_ENABLED | K6 kafka consumer logger enabled (0/1) | | CREATE_DITTO_CONNECTIONS | If the test should create the needed for scenarios ditto connections, before executing the scenarios | | DELETE_DITTO_CONNECTIONS | If the test should delete the needed for scenarios ditto connections, after executing the scenarios | | CONNECTION_OPEN_MAX_RETRIES | Maximum times connection status is fetched to check if its opened. If connection is still not opened after, the test gets aborted.| | SCENARIOS_TO_RUN | Array of scenarios names that should run, available options is: WARMUP, DEVICE_LIVE_MESSAGES, SERACH_THINGS, READ_THINGS, MODIFY_THINGS | | CREATE_THINGS_LOG_REMAINING | Log the remaining things that need to be created. Useful for debugging purposes | | THINGS_WARMUP_BATCH_SIZE | Max number of simultaneous connections of a k6 http.batch() call, which is used for warming up things | ## Ditto related | Name | Description | | ------------------------------- | ---------------------------------- | | DITO_API_URI | Ditto api url | | DITTO_DEVOPS_AUTH_HEADER | Devops user authorization header name | | DITTO_DEVOPS_AUTH_HEADER_VALUE | Devops user authorization header value | | DITTO_PRE_AUTHENTICATED_HEADER_VALUE | Ditto x-ditto-pre-authenticated header value. https://eclipse.dev/ditto/installation-operating.html#pre-authentication | ## Kafka related | Name | Description | | ----------------------- | -------------------------------- | | KAFKA_BOOTSTRAP_SERVERS | Array of kafka bootstrap servers | ## Scenarios related ###### WARMUP | Name | Description | | ------------------- | -------------------------------------------------------------------------------------------- | | WARMUP_MAX_DURATION | The maximum duration of warmup scenario. After, the scenario will be forcefully stopped | | WARMUP_START_TIME | Time offset since the start of the test, at which point this scenario should begin execution | ###### Every other scenario has the same config variables, created by suffixing the variable name with the name of the scenario, f.e. SEARCH_THINGS_DURATION | Name | Description | | ------------------ | -------------------------------------------------------------------------------------------------- | | _DURATION | Total scenario duration. | | _PER_SECOND | Number of requests to execute per second. For kafka scenarios number of kafka messages per second. | | _PRE_ALLOCATED_VUS | Number of VUs to pre-allocate before scenario start to preserve runtime reso