PlantD: (Performance, Latency ANalysis and Testing for Data pipelines) is a harness for measuring the performance of data pipelines during and after development. PlantD collects a standard suite of metrics and visualizations, for use when developing or deciding among data pipeline architectures, configurations, and business use cases.
To use PlantD, you configure it with the following information:
- How to reach your pipeline-under-test: a description of the pipeline you want to measure, including at least an IP address and port number to send data in, and tags that uniquely identify your pipeline's resources on your cloud provider.
- The data schema that your pipeline requires as input, that is, what data items are fed into the pipeline, as well as their data format and allowable values.
- From this, PlantD will generate a dataset: a quantity of generated fake data that meets that schema, for use in testing
- A load pattern describing a variable rate of load generation, for example: 100 records per second steadily for 5 minutes, then ramping up over 1 minute to 200 records per second, staying steady for 10 minutes, then ramping down to 0 over a 2 minute span.
- PlantD's load generator will send data to your pipeline following this pattern
- A description of the experiment you want to run: a timed session where the load generator sends a dataset to a pipeline-under-test using a load pattern, and collects metrics during and after the load generation.
You will need:
- A pipeline to measure
- A Kubernetes Cluster (Managed or Standalone)
- kubectl with access to the cluster
If you aren't ready to test your own pipeline, we supply a toy pipeline for demonstration
If you don't have a kubernetes cluster handy, you can use a small test cluster using Minikube. Make sure the cluster has at least 10GB of memory assigned.
Note that this will not scale up well to measuring dataflow of large pipelines, but it's enough to experiment and find out how PlantD works.
Type kubectl cluster-info to check that it's running
The easiest way to setup oeprator is to use the bundle.yaml deployments.
### Instal the K6 Operator
curl https://raw.githubusercontent.com/grafana/k6-operator/main/bundle.yaml | kubectl create -f -
### Install the Prometheus Operator
curl https://raw.githubusercontent.com/prometheus-operator/prometheus-operator/main/bundle.yaml | kubectl create -f -
### Install the PlantD Operator
curl https://raw.githubusercontent.com/CarnegieMellon-PlantD/PlantD-operator/main/bundle.yaml | kubectl create -f -
### Get the Studio service hostname
kubectl get svc plantd-studio-service -n plantd-operator-system -o jsonpath='{.status.loadBalancer.ingress[0].hostname}{"\n"}'
Note that it may take upto 2-3 minutes for the PlantD Studio to be available at the above hostname.
We welcome contributions from the open-source community, from bug fixes to new features and improvements. See CONTRIBUTING.md for more information on how to contribute.
PlantD is funded by Honda's 99P labs, with implementation and ongoing support provided by the TEEL team at Carnegie Mellon University.
PlantD is licensed under the GPLv2 License. See LICENSE for more details.
For more detailed information about how to use PlantD, see our full documentation.
API documentation can be found in the docs
For more information about the PlantD project, please contact us:
- Honda 99P labs: support@99plabs.com
- TEEL Labs: teel@andrew.cmu.edu
We are always open to collaboration, questions, and suggestions!