This repository houses StreamSets Test Framework (STF) tests for StreamSets Data Collector.
To run tests, install the StreamSets Test Framework and then, after cloning this repository, cd into it and execute stf test. By default, this will attempt to run all tests using the default SDC version. If you're just trying STF out for the first time, it might be a better idea to limit your test run to a module with no external dependencies (e.g. stage.test_dev_stages). It's also usually desirable to specify an SDC version and add -s to send test output to the console. Putting it all together:
$ git clone https://github.com/streamsets/datacollector-tests.git
Cloning into 'datacollector-tests'...
remote: Enumerating objects: 192, done.
remote: Counting objects: 100% (192/192), done.
remote: Compressing objects: 100% (185/185), done.
remote: Total 1827 (delta 109), reused 89 (delta 6), pack-reused 1635
Receiving objects: 100% (1827/1827), 741.96 KiB | 0 bytes/s, done.
Resolving deltas: 100% (1254/1254), done.
$ cd datacollector-tests/
$ stf test --sdc-version 3.4.0 -s stage/test_dev_stages.py
2018-09-24 11:47:15 AM [INFO] [streamsets.testframework.cli] Pulling Docker image streamsets/testframework:master ...
============================= test session starts ==============================
platform linux -- Python 3.6.6, pytest-3.0.4, py-1.6.0, pluggy-0.4.0
benchmark: 3.1.1 (defaults: timer=time.perf_counter disable_gc=False min_rounds=5 min_time=0.000005 max_time=1.0 calibration_precision=10 warmup=False warmup_iterations=100000)
rootdir: /root/tests, inifile:
plugins: benchmark-3.1.1
collected 6 items
stage/test_dev_stages.py 2018-09-24 11:47:39 AM [INFO] [streamsets.testframework.sdc] Starting StreamSets Data Collector 3.4.0 container ...
2018-09-24 11:47:52 AM [INFO] [streamsets.testframework.sdc] SDC is now running. SDC container (43d14e8197e4:18630) can be followed along on http://localhost:32955
2018-09-24 11:47:53 AM [WARNING] [streamsets.sdk.sdc_models] PipelineBuilder missing error stage. Will use 'Discard.'
2018-09-24 11:47:53 AM [INFO] [streamsets.sdk.sdc] Importing pipeline Pipeline35fe51c8-ca73-4a75-9ebc-8332ab49adff...
2018-09-24 11:47:53 AM [INFO] [streamsets.sdk.sdc_api] Waiting for status EDITED ...
2018-09-24 11:47:53 AM [INFO] [streamsets.sdk.sdc_api] Pipeline Pipeline35fe51c8-ca73-4a75-9ebc-8332ab49adff reached status EDITED (took 0.00 s).
2018-09-24 11:47:53 AM [INFO] [streamsets.sdk.sdc] Starting pipeline Pipeline35fe51c8-ca73-4a75-9ebc-8332ab49adff ...
2018-09-24 11:47:53 AM [INFO] [streamsets.sdk.sdc_api] Waiting for status ['RUNNING', 'FINISHED'] ...
...The folder structure of this repository (visualized below using tree) is as follows::
├── datacollector
├── fault
├── package
├── performance
├── pipeline
├── resources
│ ├── protobuf
│ └── tcp_server
├── stage
└── upgrade
└── pipelines
├── sdc_1.1.0
├── sdc_1.6.0.0
├── sdc_2.0.0.0
├── sdc_2.1.0.0
└── sdc_2.2.0.0- datacollector/: Tests that exercise DataCollector-wide functionality (e.g. classpath validation).
- fault/ (in progress): Tests that exercise product resilience in the presence of injected faults.
- package/ (in progress): Packaging tests.
- performance/ (in progress): Tests that focus on product performance using the pytest-benchmark plugin.
- pipeline/: Tests that exercise end-to-end workflows (e.g. the drift synchronization solution) or pipeline-level functionality. If the pipeline you want to test is complex, it should probably have a test here.
- resources/: Resources to be used within tests (e.g. protobuf object schema).
- stage/: Tests that attempt to isolate functionality of individual stages. This tends to be in the form of
<dev origin> >> <stage>tests for destinations,<dev origin> >> stage >> <trash>tests for processors, or<stage> >> <trash>tests for origins. - upgrade/: Legacy SDC pipeline upgrade tests. Unless there's a really good reason to do so, don't add new tests to this folder.