Resource definitions

Resource definitions describe Models, Applications, and Deployment Configurations in the YAML format. You can learn more about them in the How to write resource definitions section.

Serving

Models & Model Versions

A Model is a machine learning model or a processing function that consumes provided inputs and produces predictions or transformations.

Within the Hydrosphere platform, we break down a model into its versions. Each Model version represents a single Docker image containing all the artifacts that you have uploaded to the platform. Consequently, Model is a group of Model versions with the same name.

Runtimes

A Runtime is a Docker image with the predefined gRPC interface which loads and serves your model.

We have implemented a few runtimes, which you can use in your own projects.

Servable

Servable is a deployed instance of a Model version combined with a Runtime. It exposes a gRPC endpoint that can be used to send requests.

Users should not use Servables as-is, since they are designed to be building blocks, rather than inference endpoints. Hydrosphere provides a better alternative to deploy a Model version — Application.

Applications

An Application is a pipeline of one or more stages, each consisting of one or multiple Model Versions. Data sent to an application stage is shadowed to all of its model versions. The output of a stage is picked randomly with respect to weights.

When a user creates an Application, the Manager service automatically deploys appropriate Servables. The Application handles monitoring of your models and can perform A/B traffic splits.

Each Application has publicly available HTTP and gRPC endpoints that you can send requests to.

Deployment Configurations

A Deployment Configuration is a collection of Kubernetes settings that you can set for your Servables and Model Versions used inside of Application stages.

Deployment Configuration covers:

• Horizontal Pod Autoscaler specs

• Container Specs

  • Resource requirements: limits and requests

• Pod Specs

  • Node Selectors

  • Affinity

  • Tolerations

• Deployment Specs

  • Replicas count

Model's Signature

A Model's Signature is a specification of your model computation which identifies the name of a function with its inputs and outputs, including their names, shapes, and data types.

Example of a signature defined in a YAML file:

contract:
  name: predict
  inputs:
    x:
      shape: [-1, 2]
      type: double
      profile: numerical
  outputs:
    y:
      shape: scalar
      type: int
      profile: categorical

Field

A Field is a basic element of a Model's signature. It has a name, shape, data type, and profile.

Example of a model's signature field defined in a YAML file:

x:
  shape: [-1, 2]
  type: double
  profile: numerical

Field`s Profile

A Profile is a special tag that tells how Hydrosphere should interpret the field's data.

There are multiple available tags: Numerical, Categorical, Image, Text, Audio, Video, etc.

Monitoring

Metrics

Data coming through deployed Model Versions can be monitored with metrics.

Metric is a Model Version that takes a combination of inputs & outputs from another monitored Model Version, receives every request and response from the monitored model, produces a single value, and compares it with a threshold to determine whether this request was healthy or not.

Every request is evaluated against all metrics assigned to the model.

Auto OD Metric is an automatically generated Outlier Detection metric. More details are described here.

Checks

A check is a boolean condition associated with a field of a Model Version signature which shows for every request whether the field value is acceptable or not.

For example, Min/Max checks ensure that a field value is in an acceptable range which is inferred from training data values.

Automatic Outlier Detection

Sonar

Sonar service is responsible for managing metrics, training and production data storage, calculating profiles, and shadowing data to the Model Versions which are used as an outlier detection metrics.

Drift Report

Gateway

Gateway is a service responsible for routing requests to/from or between Servables and Applications and validating these requests for matching a Model's/Application signature.

The Gateway maps a model’s name to a corresponding container. Whenever it receives a request via HTTP API, GRPC, or Kafka Streams, it communicates with that container via the gRPC protocol.

Manager

Manager is responsible for:

• Building a Docker Image from your ML model for future deployment

• Storing these images inside a Docker Registry deployed alongside with

  manager service

• Versioning these images as Model Versions

• Creating running instances of these Model Versions called Servables

  inside Kubernetes cluster

• Combining multiple Model Versions into a linear graph with a single

  endpoint called Application

Interpretability provides EDA (Exploratory Data Analysis) and explanations for predictions made by your models to make predictions understandable and actionable. It also produces explanations for monitoring metrics to let you know why a particular request was marked as an outlier. The component consists of 2 services:

• Explanations

• Data Projections

Both services are built with Celery to run asynchronous tasks from apps and consists of a client, a worker, and a broker that mediates in between. A client generates a task and initiates it by adding a message to a queue, а broker delivers it to a worker, then the worker executes the task.

Interpretability services use MongoDB as both a Celery broker and backend storage to save task results. To save and retrieve model training and production data, the Interpretability component uses S3 storage.

When Explanation or Data Projection receives a task they create a new temporary Servable specifically for the model they need to make an explanation for. They use this Servable to run data through it in order to make new predictions and delete it after.

Prediction Explanations

Prediction Explanations generate explanations of model predictions to help you understand them. Depending on the type of data your model uses, it provides an explanation as either a set of logical predicates if your data is in a tabular format or a saliency map if your data is in the image format. Saliency Map is a heat map that highlights parts of a picture that a prediction was based on.

Data Projections

Data Projection visualizes high-dimensional data in a 2D scatter plot with an automatically trained UMAP transformer to let you evaluate data structure and spot clusters, outliers, novel data, or any other patterns. It is especially helpful if your model works with high-dimensional data, such as images or text embeddings.

Hydrosphere is a platform for deploying, versioning, and monitoring your machine learning models in production. It is language-agnostic and framework-agnostic, with support for all major programming languages and frameworks - Python, Java, Tensorflow, Pytorch, etc.

What to do next?

Serving

• Model Registry

• Inference Pipelines

• A/B Model Version Deployment

• Traffic Shadowing

• Language-Agnostic Deployment

Monitoring

• Automatic Outlier Detection

• Data Drift Report

• Monitoring Dashboard and Data Health Metrics

• Alerts

Interpretability

• Prediction Explanations

• Data Projection

Third-Party Integrations

• Kubeflow Components

• AWS Sagemaker

Hydrosphere is composed of several microservices, united to efficiently serve and monitor machine learning models in production. Hydrosphere features are divided between multiple services. You can learn more about each of them in this section.

UI / nginx

A/B Deployment

Hydrosphere users can use multiple inside of the same stage. Hydrosphere shadows traffic to all model versions inside of an application stage.

Users can specify the likelihood that a model output will be selected as an application stage output by using the weight argument.

Traffic Shadowing

Hydrosphere shadows traffic to all model versions inside of an application stage.

If you want to shadow your traffic between model versions without producing output from them simply set weight parameter to 0. This way your model version will receive all incoming traffic, but its output will never be chosen as an output of an application stage.

Hydrosphere allows you to A/B test your ML models in production.

A/B testing is a great way of measuring how well your models perform or which of your model versions is more effective and taking data-driven decisions upon this knowledge.

Production ML applications always have specific goals, for example driving as many users as possible to perform some action. To achieve these goals, it’s necessary to run online experiments and compare model versions using metrics in order to measure your progress against them. This approach allows to track whether your development efforts lead to desired outcomes.

To perform a basic A/B experiment on an application consisting of 2 variants of a model, you need to train and upload both versions to Hydrosphere, create an application with a single execution stage from them, invoke it by simulating production data flow, then analyze production data using metrics of your choice.

Learn how to set up an A/B application:

Hydrosphere is a language-agnostic platform. You can use it with models written in any language and trained in any framework. Your ML models can come from any background, without restrictions of your choices regarding ML model development tools.

In Hydrosphere you operate ML models as Runtimes, which are Docker containers packed with predefined dependencies and gRPC interfaces for loading and serving them on the platform with a model inside. All models that you upload to Hydrosphere must have the corresponding runtimes.

Runtimes are created by building a Docker container with dependencies required for the language that matches your model. You can either use our pre-made runtimes or create your own runtime.

The Hydrosphere component responsible for building Docker images from models for deployment, storing them in the registry, versioning, and more is Manager.

What is Hydrosphere?

Hydrosphere is an open-source MLOps platform for deploying, managing, and monitoring ML models in production with Kubernetes.

Hydrosphere supports all major machine learning frameworks, including Tensorflow, Keras, PyTorch, XGBoost, scikit-learn, fastai, etc. The platform is designed to effectively measure performance and health metrics of your production models, making it possible to spot early signs of performance drops and data drifts, get insights into why they happen.

Hydrosphere offers immediate value to ML-based products:

• Сovers all aspects of the production ML lifecycle - model versioning & deployment, traffic & contract management, data monitoring, gaining insights.

• Easy & fast management of production models that brings models to production in minutes by reducing time to upload, update, and roll your models into production.

• Allows to create reproducible, observable, and explainable machine learning pipelines.

• Provides understanding and control of models’ performance in production via data and target metrics analysis.

• Adds in-depth observability for your production models and data flowing through them.

• Improves business metrics of ML-based products as a result of a reduction in MTTR and MTTD incidents related to ML models due to early alerts once data drifts happen.

Why Hydrosphere?

Production ML is a dangerous place where numerous things can and usually do go wrong, making issues harder to discover and fix. Hydrosphere automates MLOps in the production part of the ML lifecycle combining best practices of CI/CD and DevOps and putting special emphasis on monitoring performance of ML models after their deployment.

MLOps problems Hydrosphere addresses:

• Non-interpretable, biased models

• Integration between the tools of each step of production ML lifecycle

• Long time to find & debug issues with production ML Models

• Monitoring for Model Degradation and Performance Loss

• Understanding the reasons behind wrong predictions

What Hydrosphere is not

Hydrosphere is not an ML model training framework. Before using Hydrosphere, you need to train your models with one of many existing frameworks for ML model training.

We suggest you use one of the orchestrators, such as Kubeflow or Airflow, to deliver your model to the Hydrosphere.

Hydrosphere Components

Hydrosphere platform includes all steps of a production ML model cycle - Versioning, Deployment, Monitoring, and Maintenance. This combination allows us to use a single tool to build an observable, reproducible, and scalable workflow, and start getting early warnings once anything goes wrong. These steps of an ML lifecycle are divided between three components that make up the Hydrosphere platform - Serving, Monitoring, and Interpretability.

Serving

Hydro Serving is responsible for framework-agnostic model deployment and management. It allows Data Scientists to upload, version, combine into linear pipelines and deploy ML models trained in any ML framework to a Docker/Kubernetes cluster and expose HTTP/gRPC/Kafka API to other parties.

Monitoring

Hydro Monitoring tracks model performance over time, raising alerts in case of detected issues. It provides a real-time updated UI, where you can monitor your models to see service health and usage. This constant monitoring of model health is crucial for any ML-based business as it’s tied to business and financial metrics.

Hydrosphere is capable of monitoring model quality with or without getting additional labeled data. Labeled data is often used in production drawing conclusions about the quality of a model’s predictions. Sometimes it is hard to get labeled data in production in a timely and cost-effective manner, especially when you deal with large volumes of complex data. Hydrosphere circumvents this issue by analyzing data that flows through a model as a proxy evaluating model quality to detect if ML models start to degrade and make unreliable predictions due to production data drifts from training data.

Interpretability

Hydrosphere Interpretability provides human-readable explanations of the predictions made by your ML models, as well as the explanations of monitoring analytics made by Hydrosphere Monitoring. It helps to evaluate and analyze models and understand what features influence their decisions. The Interpretability component demystifies your ML process, provides a new level of confidence about the reasons behind your models’ decisions and a certain level of trust business can rely on.

A Hydrosphere user can create a linear inference pipeline from multiple model versions. Such pipelines are called .

Monitoring Dashboard lets you track your performance metrics and get a high-level view of your data health.

Monitoring Dashboard plots all requests streaming through a model version which are colored in respect with how "healthy" they are. On the horizontal axis we group our data by batches and on the vertical axis we group data by signature fields. In this plot cells are determined by their batch and field. Cells are colored from green to red, depending on the average request health inside the batch.

Data Projection is a service that visualizes high-dimensional data in a 2D scatter plot with an automatically trained transformer to let you evaluate the data structure and spot clusters, outliers, novel data, or any other patterns. This is especially helpful if your model works with high-dimensional data, such as images or text embeddings.

Data Projection is an important tool, which helps to describe complex things in a simple way. One good visualization can show more than text or data. Monitoring and interpretation of machine learning models are hard tasks that require analyzing a lot of raw data: training data, production requests, as well as model outputs.

Essentially, this data is just numbers that in their original form of vectors and matrices do not have any meaning since it is hard to extract any meaning from thousands of vectors of numbers. In Hydrosphere we want to make monitoring easier and clearer that is why we created a data projection service that can visualize your data in a single plot.

Usage

To start working with Data Projection you need to create a model that has an output field with an embedding of your data. Embeddings are real-valued vectors that represent the input features in a lower dimensionality.

1. Create a model with an embedding field

   Data Projection service delegates the creation of embeddings to the user. It expects that model will create embedding from input features and pass it as output vector. Thus embedding field is required, models without this field are not supported. Data Projection also expects that output labels field is called class and model confidence is called respectively confidence. Other outputs are ignored.

2. Send data through your model

3. Check Data Projection service inside the Model Details menu

Each point in the plot presents a request. Requests with similar features are close to each other. You can select a specific request point and inspect what it consists of.

Above plot, there are several scores: global score, stability score, MSID score, etc. These scores reflect the quality of projection of multidimensional requests to 2D. To interpret scores you refer to technical documentation on Data Projection service.

In the Colorize menu, you can choose how to colorize model requests: by class, by monitoring metric or by confidence. Data Projection searchers specifically for output scalars class and confidence.

In the Accent Points menu, you can highlight the nearest in original space points to the selected one by picking the nearest variant. Counterfactuals will show you nearest points to selected but with a different predicted label.

Drift Report service creates a statistical report based on a comparison of training and production data distributions. It compares these two sets of data by a set of statistical tests and finds deviations.

Drift report uses multiple different tests with p=.95 for different features:

Numerical features:

• Levene's test with a trimmed mean

• Welch's t-test

• Mood's test

• Kolmogorov–Smirnov test

Categorical features:

• Chi-Square test

• Unseen categories

Supported Models

Right now Drift Report feature works only for Models with numerical scalar fields.

For each model with uploaded training data, Hydrosphere creates an outlier detection (Auto OD) metric, which assigns an outlier score to each request. A request is labeled as an outlier if the outlier score is greater than the 97th percentile of training data outlier scores distribution.

You can observe those models deployed as metrics in your monitoring dashboard. These metrics provide you with information about how novel/anomalous your data is.

If these values of the metric deviate significantly from the average, you can tell that you experience a data drift and need to re-evaluate your ML pipeline to check for errors.

Supported Models

Right now Auto OD feature works only for Models with numerical scalar fields and uploaded training data.

The Hydrosphere platform can be installed in the following orchestrator's:

1. Docker Compose

2. Kubernetes

Docker installation

To install Hydrosphere using docker-compose, you should have the following prerequisites installed on your machine.

• Docker 18.0+

• Docker Compose 1.23+

Install from releases

1. Download the latest 2.4.2 release from the releases page:

export HYDROSPHERE_RELEASE=2.4.2
wget -O hydro-serving-${HYDROSPHERE_RELEASE}.tar.gz https://github.com/Hydrospheredata/hydro-serving/archive/${HYDROSPHERE_RELEASE}.tar.gz

1. Unpack the tar ball:

tar -xvf hydro-serving-${HYDROSPHERE_RELEASE}.tar.gz

1. Set up an environment:

cd hydro-serving-${HYDROSPHERE_RELEASE}
docker-compose up

Install from source

1. Clone the serving repository:

   Copy

   git clone https://github.com/Hydrospheredata/hydro-serving

2. Set up an environment:

   Copy

   cd hydro-serving
   docker-compose up -d

To check the installation, open http://localhost/. By default, Hydrosphere UI is available at port 80.

Kubernetes installation

To install Hydrosphere on the Kubernetes cluster you should have the following prerequisites fulfilled.

• Helm 2.9+

• Kubernetes 1.14+ with v1 API

• PV support on the underlying infrastructure (if persistence is required)

• Docker registry with pull/push access (if the built-in one is not used)

Install from charts repository

1. Add the Hydrosphere charts repository:

   Copy

   helm repo add hydrosphere https://hydrospheredata.github.io/hydro-serving/helm

2. Install the chart from repo to the cluster:

   Copy

   helm install --name serving --namespace hydrosphere hydrosphere/serving

Install from source

1. Clone the repository:

   Copy

   git clone https://github.com/Hydrospheredata/hydro-serving.git
   cd hydro-serving/helm

2. Build dependencies:

   Copy

   helm dependency build serving

3. Install the chart:

   Copy

   helm install --namespace hydrosphere serving

After the chart has been installed, you have to expose the ui component outside of the cluster. For the sake of simplicity, we will just port-forward it locally.

kubectl port-forward -n hydrosphere svc/serving-ui 8080:9090

To check the installation, open http://localhost:8080/.

Overview

****Sonar sends data about any failed health checks of live production models and applications to Prometheus AlertManager. Once a user deploys a model to production, adds training data and starts sending production requests, these requests start getting checked by Sonar. If Sonar detects an anomaly (for example, a data check failed, or a metric value exceeded the threshold), AlertManager sends an appropriate alert.

Users can manage alerts by setting up AlertManager for Prometheus on Kubernetes. This can be helpful when you have models that you get too many alerts from and need to filter, group, or partly silence them. AlertManager can take care of grouping, inhibition, silencing of alerts, and routing them to the receiver integration of your choice. To configure alerts, modify the prometheus-am-configmap-<release_name> ConfigMap.

For more information about Prometheus AlertManager please refer to its official documentation.

Hydrosphere CLI, orhs, is a command-line interface designed to work with the Hydrosphere platform.

Source code: https://github.com/Hydrospheredata/hydro-serving-cli PyPI: https://pypi.org/project/hs/

Installation

Use pip to install hs:

pip install hs

Check the installation:

hs --version

Usage

hs cluster

This command lets you operate cluster instances. A cluster points to your Hydrosphere instance. You can use this command to work with different Hydrosphere instances.

See hs cluster --help for more information.

hs upload

This command lets you upload models to the Hydrosphere platform. During the upload, hs looks for a serving.yaml file in the current directory. This file must contain a definition of the model (example).

See hs upload --help for more information.

hs apply

This command is an extended version of the hs upload command, which also allows you to operate applications and host selector resources.

See hs apply --help for more information.

hs profile

This command lets you upload your training data to build profiles.

• $ hs profile push - upload training data to compute its profiles.

• $ hs profile status - show profiling status for a given model.

See hs profile --help for more information.

hs app

This command provides information about available applications.

• $ hs app list - list all existing applications.

• $ hs app rm - remove a certain application.

See hs app --help - for more information.

hs model

This command provides information about available models.

• $ hs model list - list all existing models.

• $ hs model rm - remove a certain model.

See hs model --help for more information.

Hydrosphere Serving Components for Kubeflow Pipelines provide integration between Hydrosphere model serving benefits and Kubeflow orchestration capabilities. This allows launching training jobs as well as serving the same models in Kubernetes in a single pipeline.

You can find examples of sample pipelines here.

Serving components

Deploy

The Deploy component allows you to upload a model, trained in a Kubeflow pipelines workflow to a Hydrosphere platform.

For more information, check Hydrosphere Deploy Kubeflow Component

Release

The Release component allows you to create an Application from a model previously uploaded to Hydrosphere platform. This application will be capable of serving prediction requests by HTTP or gRPC.

For more information, check Hydrosphere Release Kubeflow Component

Prediction Explanation service is designed to help Hydrosphere users understand the underlying causes of changes in predictions coming from their models.

Prediction Explanation generates explanations of predictions produced by your models and tells you why a model made a particular prediction. Depending on the type of data your model uses, Prediction Explanation provides an explanation as either a set of logical predicates (if your data is in a tabular format) or a saliency map (if your data is in the image format). A saliency map is a heat map that highlights parts of a picture that a prediction was based on.

Hydrosphere uses model-agnostic methods for explaining your model predictions. Such methods can be used on any machine learning model after they've been uploaded to the platform.

As of now, Hydrosphere supports explaining tabular and image data with Anchor and RISE tools correspondingly.

Overview

In this tutorial, you will learn the basics of working with Hydrosphere. We will prepare an example model for serving, deploy it to Hydrosphere, turn it into an application, invoke it locally, and use monitoring. As an example model, we will take a simple logistic regression model fit with randomly generated data, with some noise added to it.

By the end of this tutorial you will know how to:

• Prepare a model for Hydrosphere

• Serve a model on Hydrosphere

• Create an Application

• Invoke an Application

• Use basic monitoring

Prerequisites

For this tutorial, you need to have Hydrosphere Platform deployed and Hydrosphere CLI (hs) along with Python SDK (hydrosdk) _**_installed on your local machine. If you don't have them yet, please follow these guides first:

To let hs know where the Hydrosphere platform runs, configure a new cluster entity:

Before you start

In the next two sections, we will prepare a model for deployment to Hydrosphere. It is important to stick to a specific folder structure during this process to let hs parse and upload the model correctly. Make sure that the structure of your local model directory looks like this by the end of the model preparation section:

• train.py - a training script for our model

• requirements.txt - provides dependencies for our model

• model.joblib - a model artifact that we get as a result of model training

• src/func_main.py - an inference script that defines a function for making model predictions

• serving.yaml - a resource definition file to let Hydrosphere know which function to call from the func_main.py script and let the model manager understand model’s inputs and outputs.

Training a model

While Hydrosphere is a post-training platform, let's start with basic training steps to have a shared context.

First, create a directory for your model and create a new train.py inside:

Put the following code for your model in the train.py file:

Next, we need to install all the necessary libraries for our model. In your logistic_regression folder, create a requirements.txt file and provide dependencies inside:

Install all the dependencies to your local environment:

Train the model:

As soon as the script finishes, you will get the model saved to a model.joblib file.

Model preparation

Every model in the Hydrosphere cluster is deployed as an individual container. After a request is sent from the client application, it is passed to the appropriate Docker container with your model deployed on it. An important detail is that all model files are stored in the /model/files directory inside the container, so we will look there to load the model.

To run our model we will use a Python runtime that can execute any Python code you provide. Model preparation is pretty straightforward, but you have to create a specific folder structure described in the "Before you start" section.

Provide the inference script

Let's create the main file func_main.pyin the /src folder of your model directory:

To do inference you have to define a function that will be invoked every time Hydrosphere handles a request and passes it to the model. Inside that function, you have to call a predict (or similar) method of your model and return your predictions:

Inside func_main.py we initialize our model outside of the serving function infer.This process will not be triggered every time a new request comes in.

The infer function takes the actual request, unpacks it, makes a prediction, packs the answer, and returns it. There is no strict rule for naming this function, it just has to be a valid Python function name.

Provide a resource definition file

To let Hydrosphere know which function to call from the func_main.py file, we have to provide a resource definition file. This file will define a function to be called, inputs and outputs of a model, a signature function, and some other metadata required for serving.

Create a resource definition file serving.yamlin the root of your model directorylogistic_regression:

Inside serving.yaml we also providerequirements.txt andmodel.joblib as payload files to our model:

At this point make sure that the overall structure of your local model directory looks as shown in the "Before you start" section.

Serving a Model

Now we are ready to upload our model to Hydrosphere. To do so, inside the logistic_regression model directory run:

If you cannot find your newly uploaded model and it is listed on your models' page, it is probably still in the building stage. Wait until the model changes its status to Released, then you can use it.

Creating an Application

Invoking an application

Invoking applications is available via different interfaces. For this tutorial, we will cover calling the created Application by gRPC via our Python SDK.

To install SDK run:

Define a gRPC client on your side and make a call from it:

Getting Started with Monitoring

Hydrosphere Platform has multiple tools for data drift monitoring:

1. Data Drift Report

2. Automatic Outlier Detection

3. Profiling

In this tutorial, we'll look at the monitoring dashboard and Automatic Outlier Detection feature.

Provide training data

To provide training data users need to add the training-data=<path_to_csv> field to the serving.yaml file. Run the following script to save training data used in previous steps as a trainig_data.csv file:

Next, add the training data field to the model definition inside the serving.yaml file:

Upload a model

Now we are ready to upload our model. Run the following command to create a new version of the logistic_regresion model:

For each model with uploaded training data, Hydrosphere creates an outlier detection metric, which assigns an outlier score to each request. This metric labels a request as an outlier if the outlier score is greater than the 97th percentile of training data outlier scores distribution.

Update an Application

Let's send some data to our new model version. To do so, we need to update our logistic_regression application. To update it, we can go to the Application tab and click the "Update" button:

Send data to Application

After updating our Application, we can reuse our old code to send some data:

Monitor data quality

You can monitor your data quality in the Monitoring Dashboard:

The Monitoring dashboard plots all requests streaming through a model version as rectangles colored according to how "healthy" they are. On the horizontal axis, we group our data by batches and on the vertical axis, we group data by signature fields. In this plot, cells are determined by their batch and field. Cells are colored from green to red, depending on the average request health inside this batch.

Check data drift detection

To check whether our metric will be able to detect data drifts, let's simulate one and send data from another distribution. To do so, let's slightly modify our code:

You can validate that your model was able to detect data drifts on the monitoring dashboard.

Contents

• A/B Deployment and Traffic Shadowing

• Using Deployment Configurations

• Train & Deploy Census Income Classification Model

• Monitor Anomalies with a Custom Metric

• Monitor External Models

Overview

This section contains tutorials to help you get started with the Hydrosphere platform. A tutorial shows how to accomplish a goal rather than a single basic task.

Typically, a tutorial has several sections. When a tutorial section has several pieces of code to illustrate it, they can be shown as a group of tabs that you can switch between.

For guides on performing more basic technical steps, please look in the How-To section:

• How-To

Hydrosphere has an internal Model Registry as centralized storage for Model Versions. When you build a Dockerized model and upload it to Hydrosphere or create new model versions, they get uploaded/stored to the configured model registry in the form of images. This organizes and simplifies model management across the platform and production lifecycle.

Overview

In this tutorial, you will learn how to configure deployed Applications.

By the end of this tutorial you will know how to:

• Train and upload an example model version

• Create a Deployment Configuration

• Create an Application from the uploaded model version with previously created deployment configuration

• Examine settings of a Kubernetes cluster

Prerequisites

• Hydrosphere platform installed in Kubernetes cluster

• Python SDK / CLI

Upload a Model

In this section, we describe the resources required to create and upload an example model used in further sections. If you have no prior experience with uploading models to the Hydrosphere platform we suggest that you visit the Getting Started Tutorial.

Here are the resources used to train sklearn.ensemble.GradientBoostingClassifier and upload it to the Hydrosphere cluster.

numpy~=1.18
scipy==1.4.1
scikit-learn~=0.23

kind: Model
name: my-model
runtime: hydrosphere/serving-runtime-python-3.7:2.3.2
install-command: pip install -r requirements.txt
payload:
  - src/
  - requirements.txt
  - model.joblib
contract:
  name: infer
  inputs:
    x:
      shape: [30]
      type: double
  outputs:
    y:
      shape: scalar
      type: int64

import joblib
import pandas as pd
from sklearn.datasets import make_blobs
from sklearn.ensemble import GradientBoostingClassifier

# initialize data
X, y = make_blobs(n_samples=3000, n_features=30)

# create a model
model = GradientBoostingClassifier(n_estimators=200)
model.fit(X, y)

# Save training data and model
pd.DataFrame(X).to_csv("training_data.csv", index=False)
joblib.dump(model, "model.joblib")

import joblib
import numpy as np

# Load model once
model = joblib.load("/model/files/model.joblib")


def infer(x):
    # Make a prediction
    y = model.predict(x[np.newaxis])

    # Return the scalar representation of y
    return {"y": np.asscalar(y)}

Our folder structure should look like this:

dep_config_tutorial
├── model.joblib
├── train.py
├── requirements.txt
├── serving.yaml
└── src
    └── func_main.py

After we have made sure that all files are placed correctly, we can upload the model to the Hydrosphere platform by running hs upload from the command line.

hs upload

Create a Deployment Configuration

Next, we are going to create and upload an instance of Deployment Configuration to the Hydrosphere platform.

Deployment Configurations describe with which Kubernetes settings Hydrosphere should deploy servables. You can specify Pod Affinity and Tolerations, the number of desired pods in deployment, ResourceRequirements, and Environment Variables for the model container, and HorizontalPodAutoScaler settings.

Created Deployment Configurations can be attached to Servables and Model Variants inside of Application.

Deployment Configurations are immutable and cannot be changed after they've been uploaded to the Hydrosphere platform.

You can create and upload Deployment Configuration to Hydrosphere via YAML Resource definition or via Python SDK.

For this tutorial, we'll create a deployment configuration with 2 initial pods per deployment, HPA, and FOO environment variable with value bar.

kind: DeploymentConfiguration
name: my-dep-config
deployment:
  replicaCount: 2
hpa:
  minReplicas: 2
  maxReplicas: 4
  cpuUtilization: 70
container:
  env:
    FOO: bar

hs apply -f deployment_configuration.yaml

from hydrosdk import Cluster, DeploymentConfigurationBuilder

cluster = Cluster("http://localhost")

dep_config_builder = DeploymentConfigurationBuilder("my-dep-config", cluster)
dep_config = dep_config_builder. \
    with_replicas(replica_count=2). \
    with_env({"FOO":"bar"}). \
    with_hpa(max_replicas=4,
             min_replicas=2,
             target_cpu_utilization_percentage=70).build()

Create an Application

kind: Application
name: my-app-with-config
pipeline:
  - - model: my-model:1
      weight: 100
      deploymentConfiguartion: my-config

hs apply -f application.yaml

from application import ApplicationBuilder, ExecutionStageBuilder
from hydrosdk import ModelVersion, Cluster, DeploymentConfiguration

cluster = Cluster('http:\\localhost')
my_model = ModelVersion.find(cluster, "my-model", 1)
my_config = DeploymentConfiguration.find(cluster, "my-config")

stage = ExecutionStageBuilder().with_model_variant(model_version=my_model,
                                                   weight=100,
                                                   deployment_configuration=my_config).build()

app = ApplicationBuilder(cluster, "my-app-with-config").with_stage(stage).build()

Examine Kubernetes Settings

Replicas

You can check whether with_replicas was successful by calling kubectl get deployment -A -o wide and checking the READYcolumn.

HPA

To check whether with_hpa was successful you should get a list of all created Horizontal Pod Autoscaler Resources. You can do so by calling kubectl get hpa -A

The output is similar to:

NAME                        REFERENCE                                            TARGETS    MINPODS   MAXPODS   REPLICAS   AGE
my-model-1-tumbling-star    CrossVersionObjectReference/my-model-1-tumbling-star 20%/70%    2         4         2          1d

Environment Variables

To list all environment variables run kubectl exec my-model-1-tumbling-star -it /bin/bash and then execute the printenv command which prints ann system variables.

The output is similar to:

MY_MODEL_1_TUMBLING_STAR_SERVICE_PORT_GRPC=9091
...
FOO=bar

Python SDK offers a simple and convenient way of integrating a user's workflow scripts with Hydrosphere API.

Source code: https://github.com/Hydrospheredata/hydro-serving-sdk PyPI: https://pypi.org/project/hydrosdk/

You can learn more about it in its documentation here.

Installation

You can use pip to install hydrosdk

pip install hydrosdk

Usage

You can access the locally deployed Hydrosphere platform from previous steps by running the following code:

from hydrosdk import Cluster, Application 
import pandas as pd

cluster = Cluster("http://localhost", grpc_address="localhost:9090")

app = Application.find(cluster, "my-model")
predictor = app.predictor()

df = pd.read_csv("path/to/data.csv")
for row in df.itertuples(index=False):
    predictor.predict(row)

To use private pip repository you must add customized pip.conf file pointing to your custom PyPI repository.

For example, your custom pip.conf file can look like this:

[global]
timeout = 60
index-url = http://pypi.python.org/simple/

If you need to specify the certificate to use during pip install you want to specify the path to it in a pip.conf file e.g.

[global]
timeout = 60
index-url = http://pypi.python.org/simple/
cert = /model/files/cert.pem

You can tell pip to use this pip.conffile in the install-command field inside serving.yaml:

kind: Model
name: linear_regression
runtime: "hydrosphere/serving-runtime-python-3.7:$2.4.2$"
install-command: "PIP_CONFIG_FILE=pip.conf pip install -r requirements.txt"
payload:
  - "src/"
  - "requirements.txt"
  - "pip.conf"  # location of your pip.conf
  - "cert.pem"  # location of your certificate. It'll be available under /model/files/cert.pem
  - "model.h5"
contract:
  name: infer
  inputs:
    x:
      shape: [-1, 2]
      type: double
  outputs:
    y:
      shape: [-1]
      type: double

Sometimes our runtime images are not flexible enough. In that case, you might want to implement one yourself.

The key things you need to know to write your own runtime are:

• How to implement a predefined gRPC service for a dedicated language

• How to our contracts' protobufs work to describe entry points, such as inputs and outputs

• How to create your own Docker image and publish it to an open registry

Generate GRPC code

There are different approaches to generating client and server gRPC code in different languages. Let's have a look at how to do that in Python.

First, let's clone our protos repository and prepare a folder for the generated code:

$ git clone https://github.com/Hydrospheredata/hydro-serving-protos
$ mkdir runtime

To generate the gRPC code we need to install additional packages:

$ pip install grpcio-tools googleapis-common-protos

Our custom runtime will require contracts and tf protobuf messages. Let's generate them:

$ python -m grpc_tools.protoc --proto_path=./hydro-serving-protos/src/ --python_out=./runtime/ --grpc_python_out=./runtime/ $(find ./hydro-serving-protos/src/hydro_serving_grpc/contract/ -type f -name '*.proto')
$ python -m grpc_tools.protoc --proto_path=./hydro-serving-protos/src/ --python_out=./runtime/ --grpc_python_out=./runtime/ $(find ./hydro-serving-protos/src/hydro_serving_grpc/tf/ -type f -name '*.proto')
$ cd runtime
$ find ./hydro_serving_grpc -type d -exec touch {}/__init__.py \;

The structure of the runtime should now be as follows:

runtime
└── hydro_serving_grpc
    ├── __init__.py
    ├── contract
    │   ├── __init__.py
    │   ├── model_contract_pb2.py
    │   ├── model_contract_pb2_grpc.py
    │   ├── model_field_pb2.py
    │   ├── model_field_pb2_grpc.py
    │   ├── model_signature_pb2.py
    │   └── model_signature_pb2_grpc.py
    └── tf
        ├── __init__.py
        ├── api
        │   ├── __init__.py
        │   ├── model_pb2.py
        │   ├── model_pb2_grpc.py
        │   ├── predict_pb2.py
        │   ├── predict_pb2_grpc.py
        │   ├── prediction_service_pb2.py
        │   └── prediction_service_pb2_grpc.py
        ├── tensor_pb2.py
        ├── tensor_pb2_grpc.py
        ├── tensor_shape_pb2.py
        ├── tensor_shape_pb2_grpc.py
        ├── types_pb2.py
        └── types_pb2_grpc.py

Implement Service

Now that we have everything set up, let's implement a runtime. Create a runtime.py file and put in the following code:

from hydro_serving_grpc.tf.api.predict_pb2 import PredictRequest, PredictResponse
from hydro_serving_grpc.tf.api.prediction_service_pb2_grpc import PredictionServiceServicer, add_PredictionServiceServicer_to_server
from hydro_serving_grpc.tf.types_pb2 import *
from hydro_serving_grpc.tf.tensor_pb2 import TensorProto
from hydro_serving_grpc.contract.model_contract_pb2 import ModelContract
from concurrent import futures

import os
import time
import grpc
import logging
import importlib


class RuntimeService(PredictionServiceServicer):
    def __init__(self, model_path, contract):
        self.contract = contract
        self.model_path = model_path
        self.logger = logging.getLogger(self.__class__.__name__)

    def Predict(self, request, context):
        self.logger.info(f"Received inference request: {request}")

        module = importlib.import_module("func_main")
        executable = getattr(module, self.contract.predict.signature_name)
        result = executable(**request.inputs)

        if not isinstance(result, hs.PredictResponse):
            self.logger.warning(f"Type of a result ({result}) is not `PredictResponse`")
            context.set_code(grpc.StatusCode.OUT_OF_RANGE)
            context.set_details(f"Type of a result ({result}) is not `PredictResponse`")
            return PredictResponse()
        return result


class RuntimeManager:
    def __init__(self, model_path, port):
        self.logger = logging.getLogger(self.__class__.__name__)
        self.port = port
        self.model_path = model_path
        self.server = None

        with open(os.path.join(model_path, 'contract.protobin')) as file:
            contract = ModelContract.ParseFromString(file.read())
        self.servicer = RuntimeService(os.path.join(self.model_path, 'files'), contract)

    def start(self):
        self.logger.info(f"Starting PythonRuntime at {self.port}")
        self.server = grpc.server(futures.ThreadPoolExecutor(max_workers=10))
        add_PredictionServiceServicer_to_server(self.servicer, self.server)
        self.server.add_insecure_port(f'[::]:{self.port}')
        self.server.start()

    def stop(self, code=0):
        self.logger.info(f"Stopping PythonRuntime at {self.port}")
        self.server.stop(code)
© 2020 GitHub, Inc.

Let's quickly review what we have here. RuntimeManager simply manages our service, i.e. starts it, stops it, and holds all necessary data. RuntimeService is a service that actually implements thePredict(PredictRequest) RPC function.

The model will be stored inside the /model directory in the Docker container. The structure of /model is a follows:

model
├── contract.protobin
└── files
    ├── ...
    └── ...

Thecontract.protobin file will be created by the Manager service. It contains a binary representation of the ModelContract message.

files directory contains all files of your model.

To run this service let's create an another file main.py.

from runtime import RuntimeManager

import os
import time
import logging

logging.basicConfig(level=logging.INFO)

if __name__ == '__main__':
    runtime = RuntimeManager('/model', port=int(os.getenv('APP_PORT', "9090")))
    runtime.start()

    try:
        while True:
            time.sleep(60 * 60 * 24)
    except KeyboardInterrupt:
        runtime.stop()

Publish Runtime

Before we can use the runtime, we have to package it into a container.

To add requirements for installing dependencies, create a requirements.txt file and put inside:

grpcio==1.12.1 
googleapis-common-protos==1.5.3

Create a Dockerfile to build our image:

FROM python:3.6.5 

ADD . /app
RUN pip install -r /app/requirements.txt

ENV APP_PORT=9090

VOLUME /model 
WORKDIR /app

CMD ["python", "main.py"]

APP_PORT is an environment variable used by Hydrosphere. When Hydrosphere invokes Predict method, it does so via the defined port.

The structure of the runtime folder should now look like this:

runtime
├── Dockerfile
├── hydro_serving_grpc
│   ├── __init__.py
│   ├── contract
│   │   ├── __init__.py
│   │   ├── model_contract_pb2.py
│   │   ├── model_contract_pb2_grpc.py
│   │   ├── model_field_pb2.py
│   │   ├── model_field_pb2_grpc.py
│   │   ├── model_signature_pb2.py
│   │   └── model_signature_pb2_grpc.py
│   └── tf
│       ├── __init__.py
│       ├── api
│       │   ├── __init__.py
│       │   ├── model_pb2.py
│       │   ├── model_pb2_grpc.py
│       │   ├── predict_pb2.py
│       │   ├── predict_pb2_grpc.py
│       │   ├── prediction_service_pb2.py
│       │   └── prediction_service_pb2_grpc.py
│       ├── tensor_pb2.py
│       ├── tensor_pb2_grpc.py
│       ├── tensor_shape_pb2.py
│       ├── tensor_shape_pb2_grpc.py
│       ├── types_pb2.py
│       └── types_pb2_grpc.py
├── main.py
├── requirements.txt
└── runtime.py

Build and push the Docker image:

$ docker build -t {username}/python-runtime-example
$ docker push {username}/python-runtime-example

That's it. You have just created a simple runtime that you can use in your own projects. It is an almost identical version of our python runtime implementation. You can always look up details there.

Hydrosphere UI

To send a sample request using Hydrosphere UI, open the desired application, and press the Test button at the upper right corner. We will generate dummy inputs based on your model's contract and send an HTTP request to the model's endpoint.

HTTP Inference

POST /gateway/application/<application_name>

To send an HTTP request, you should send a POST request to the /gateway/application/<applicationName> endpoint with the JSON body containing your request data, composed with respect to the model's contract.

Path Parameters

Request Body

gRPC

To send a gRPC request you need to create a specific client.

import grpc 
import hydro_serving_grpc as hs  # pip install hydro-serving-grpc

# connect to your ML Lamba instance
channel = grpc.insecure_channel("<host>")
stub = hs.PredictionServiceStub(channel)

# 1. define a model, that you'll use
model_spec = hs.ModelSpec(name="model")

# 2. define tensor_shape for Tensor instance
tensor_shape = hs.TensorShapeProto(
    dim=[hs.TensorShapeProto.Dim(size=-1), hs.TensorShapeProto.Dim(size=2)])

# 3. define tensor with needed data
tensor = hs.TensorProto(dtype=hs.DT_DOUBLE, tensor_shape=tensor_shape, double_val=[1,1,1,1])

# 4. create PredictRequest instance
request = hs.PredictRequest(model_spec=model_spec, inputs={"x": tensor})

# call Predict method
result = stub.Predict(request)

import com.google.protobuf.Int64Value;
import io.grpc.ManagedChannel;
import io.grpc.ManagedChannelBuilder;
import io.hydrosphere.serving.tensorflow.DataType;
import io.hydrosphere.serving.tensorflow.TensorProto;
import io.hydrosphere.serving.tensorflow.TensorShapeProto;
import io.hydrosphere.serving.tensorflow.api.Model;
import io.hydrosphere.serving.tensorflow.api.Predict;
import io.hydrosphere.serving.tensorflow.api.PredictionServiceGrpc;

import java.util.Random;

public class HydrosphereClient {

    private final String modelName;         // Actual model name, registered within Hydrosphere platform
    private final Int64Value modelVersion;  // Model version of the registered model within Hydrosphere platform
    private final ManagedChannel channel;
    private final PredictionServiceGrpc.PredictionServiceBlockingStub blockingStub;

    public HydrosphereClient2(String target, String modelName, long modelVersion) {
        this(ManagedChannelBuilder.forTarget(target).build(), modelName, modelVersion);
    }

    HydrosphereClient2(ManagedChannel channel, String modelName, long modelVersion) {
        this.channel = channel;
        this.modelName = modelName;
        this.modelVersion = Int64Value.newBuilder().setValue(modelVersion).build();
        this.blockingStub = PredictionServiceGrpc.newBlockingStub(this.channel);
    }

    private Model.ModelSpec getModelSpec() {
        /*
        Helper method to generate ModelSpec.
         */
        return Model.ModelSpec.newBuilder()
                .setName(this.modelName)
                .setVersion(this.modelVersion)
                .build();
    }

    private TensorProto generateDoubleTensorProto() {
        /*
        Helper method generating random TensorProto object for double values.
        */
        return TensorProto.newBuilder()
                .addDoubleVal(new Random().nextDouble())
                .setDtype(DataType.DT_DOUBLE)
                .setTensorShape(TensorShapeProto.newBuilder().build())  // Empty TensorShape indicates scalar shape
                .build();
    }

    public Predict.PredictRequest generatePredictRequest() {
        /*
        PredictRequest is used to define the data passed to the model for inference.
        */
        return Predict.PredictRequest.newBuilder()
                .putInputs("in", this.generateDoubleTensorProto())
                .setModelSpec(this.getModelSpec())
                .build();
    }


    public Predict.PredictResponse predict(Predict.PredictRequest request) {
        /*
        The actual use of RPC method Predict of the PredictionService to invoke prediction.
        */
        return this.blockingStub.predict(request);
    }

    public static void main(String[] args) throws Exception {
        HydrosphereClient client = new HydrosphereClient("<host>", "example", 2);
        Predict.PredictRequest request = client.generatePredictRequest();
        Predict.PredictResponse response = client.predict(request);
        System.out.println(response);
    }
}

Python SDK

You can learn more about our Python SDK here.

import hydrosdk as hs

hs_cluster = hs.Cluster(http_address='{HTTP_CLUSTER_ADDRESS}',
                         grpc_address='{GRPC_CLUSTER_ADDRESS}',)

app = hs.Application.find(hs_cluster, "{APP_NAME}")

predictor = adult_servable.predictor()

data  = ...  # your data
predictor.predict(data)

Resource definitions describe Hydrosphere entities.

An entity could be your model, application, or deployment configuration. Each definition is represented by a .yaml file.

Base definition

Every definition must include the following fields:

• kind: defines the type of a resource

• name: defines the name of a resource

The only valid options for kind are:

• Model

• Application

• DeploymentConfiguration

kind: Model

A model definition must contain the following fields:

• runtime: a string defining the runtime Docker image that will be used to run a model. You can learn more about runtimes here.

• contract: an object defining the inputs and outputs of a model.

A model definition can contain the following fields:

• payload: a list of files that should be added to the container.

• install-command: a string defining a command that should be executed during the container build.

• training-data: a string defining a path to the file that will be uploaded to Hydrosphere and used as a training data reference. It can be either a local file or a URI to an S3 object. At the moment we only support .csv files.

• metadata: an object defining additional user metadata that will be displayed on the Hydrosphere UI.

The example below shows how a model can be defined on the top level.

kind: "Model"
name: "sample_model"
training-data: "s3://bucket/train.csv" | "/temp/file.csv"
runtime: "hydrosphere/serving-runtime-python-3.6:$2.4.2$"
install-command: "sudo apt install jq && pip install -r requirements.txt" 
payload: 
  - "./requirements.txt"
contract:
  ...
metadata:
  ...

Contract object

contract object must contain the following fields:

• inputs: an object, defining all inputs of a model

• outputs: an object, defining all outputs of a model

contract object can contain the following fields:

• name: a string defining the signature of the model that should be used to process requests

Field object

field object must contain the following fields:

• shape: either "scalar" or a list of integers, defining the shape of your data. If a shape is defined as a list of integers, it can have -1 value at the very beginning of the list, indicating that this field has an arbitrary number of "entities". -1 cannot be put anywhere aside from the beginning of the list.

• type: a string defining the type of data.

field object can contain the following fields:

• profile: a string, defining the profile type of your data.

The only valid options for type are:

• bool — Boolean

• string — String in bytes

• half — 16-bit half-precision floating-point

• float16 — 16-bit half-precision floating-point

• float32 — 32-bit single-precision floating-point

• double — 64-bit double-precision floating-point

• float64 — 64-bit double-precision floating-point

• uint8 — 8-bit unsigned integer

• uint16 — 16-bit unsigned integer

• uint32 — 32-bit unsigned integer

• uint64 — 64-bit unsigned integer

• int8 — 8-bit signed integer

• int16 — 16-bit signed integer

• int32 — 32-bit signed integer

• int64 — 64-bit signed integer

• qint8 — Quantized 8-bit signed integer

• quint8 — Quantized 8-bit unsigned integer

• qint16 — Quantized 16-bit signed integer

• quint16 — Quantized 16-bit unsigned integer

• complex64 — 64-bit single-precision complex

• complex128 — 128-bit double-precision complex

The only valid options for profile are:

• text — monitoring such fields will be done with text-oriented algorithms.

• image — monitoring such fields will be done with image-oriented algorithms.

• numerical — monitoring such fields will be done with numerical-oriented algorithms.

• categorical — monitoring such fields will be done with categorical-oriented algorithms.

The example below shows how a contract can be defined on the top level.

name: "infer"
inputs:
  input_field_1:
    shape: [-1, 1]
    type: string
    profile: text
  input_field_2:
    shape: [200, 200]
    type: int32
    profile: categorical
outputs: 
  output_field_1:
    shape: scalar
    type: int32 
    profile: numerical

Metadata object

metadata object can represent any arbitrary information specified by the user. The structure of the object is not strictly defined. The only constraint is that the object must have a key-value structure, where a value can only be of a simple data type (string, number, boolean).

The example below shows, how metadata can be defined.

metadata:
  experiment: "demo"
  environment: "kubernetes"

The example below shows a complete definition of a sample model.

kind: "Model"
name: "sample_model"
training-data: "s3://bucket/train.csv" | "/temp/file.csv"
runtime: "hydrosphere/serving-runtime-python-3.6:$2.4.2$"
install-command: "sudo apt install jq && pip install -r requirements.txt" 
payload: 
  - "./*"
contract:
  name: "infer"
  inputs:
    input_field_1:
      shape: [-1, 1]
      type: string
      profile: text
    input_field_2:
      shape: [-1, 1]
      type: int32
      profile: numerical
  outputs: 
    output_field_1:
      shape: scalar
      type: int32 
      profile: numerical
metadata:
  experiment: "demo"
  environment: "kubernetes"

kind: Application

The application definition must contain one of the following fields:

• singular: An object, defining a single-model application;

• pipeline: A list of objects, defining an application as a pipeline of models.

Singular object

singular object represents an application consisting only of one model. The object must contain the following fields:

• model: A string, defining a model version. It is expected to be in the form model-name:model-version.

The example below shows how a singular application can be defined.

kind: "Application"
name: "sample_application"
singular:
  model: "sample_model:1"

Pipeline object

pipeline represents a list of stages, representing models.

stage object must contain the following fields:

• model: A string defining a model version. It is expected to be in the form model-name:model-version.

stage object can contain the following fields:

• weight: A number defining the weight of the model. All models' weights in a stage must add up to 100.

The example below shows how a pipeline application can be defined.

kind: Application
name: sample-claims-app
pipeline:
  - - model: "claims-preprocessing:1"
  - - model: "claims-model:1"
      weight: 80
    - model: "claims-model:2"
      weight: 20

In this application, 100% of the traffic will be forwarded to the claims-preprocessing:1 model version and the output will be fed into claims-model. 80% of the traffic will go to the claims-model:1 model version, 20% of the traffic will go to the claims-model:2 model version.

kind: DeploymentConfiguration

The DeploymentConfiguration resource definition can contain the following fields:

• hpa: An object defining HorizontalPodAutoscalerSpec

• container: An object defining settings applied on a container level

• deployment: An object defining settings applied on a deployment level

• pod: An object defining settings applied on a pod level

HPA object

The hpa object closely resembles the Kubernetes HorizontalPodAutoscalerSpec object

The hpa object must contain:

• minReplicas : minReplicas is the lower limit for the number of replicas to which the autoscaler can scale down.

• maxReplicas : integer, upper limit for the number of pods that can be set by the autoscaler; cannot be smaller than minReplicas.

• cpuUtilization : integer from 1 to 100, target average CPU utilization (represented as a percentage of requested CPU) over all the pods; if not specified the default autoscaling policy will be used.

Container object

The container object can contain:

• resources : object with limits and requests fields. Closely resembles the k8s ResourceRequirements object

• env : object with string keys and string values which is used to set environment variables.

Pod object

The hpa object is similar to the Kubernetes PodSpec object.

The pod object can contain

• nodeSelector : selector which must be true for the pod to fit on a node. Selector which must match a node's labels for the pod to be scheduled on that node. More info.

• affinity : pod's scheduling constraints. Represented by an Affinity object.

• tolerations : array of Tolerations.

Deployment object

The deployment object must contain:

• replicaCount : integer, number of desired pods. This is a pointer to distinguish between explicit zero and not specified. Defaults to 1.

Example

The example below shows how a deployment configuration can be defined.

kind: DeploymentConfiguration
name: cool-deployment-config
hpa:
  minReplicas: 2
  maxReplicas: 10
  cpuUtilization: 80
deployment:
  replicaCount: 4
container:
  resources:
    limits:
      cpu: 500m
      memory: 4G
    requests:
      cpu: 250m
      memory: 2G
  env:
    foo: bar
pod:
  nodeSelector:
    im: a map
    foo: bar
  affinity:
    nodeAffinity:
      requiredDuringSchedulingIgnoredDuringExecution:
        nodeSelectorTerms:
        - matchExpressions:
          - key: exp1
            operator: Exists
          matchFields:
          - key: fields1
            operator: Exists
      preferredDuringSchedulingIgnoredDuringExecution:
      - preference:
          matchExpressions:
          - key: exp2
            operator: NotIn
            values:
            - aaaa
            - bvzv
            - czxc
          matchFields:
          - key: fields3
            operator: NotIn
            values:
            - aaa
            - cccc
            - zxcc
        weight: 100
    podAffinity:
      requiredDuringSchedulingIgnoredDuringExecution:
      - labelSelector:
          matchExpressions:
          - key: value
            operator: Exists
          - key: key
            operator: NotIn
            values:
            - a
            - b
        namespaces:
        - namespace1
        topologyKey: top
      preferredDuringSchedulingIgnoredDuringExecution:
      - weight: 100
        podAffinityTerm:
          labelSelector:
            matchLabels:
              key: a
            matchExpressions:
            - key: key1
              operator: In
              values:
              - a
              - b
            - key: value2
              operator: NotIn
              values:
              - b
          namespaces:
          - namespace2
          topologyKey: topo_valur
    podAntiAffinity:
      requiredDuringSchedulingIgnoredDuringExecution:
      - labelSelector:
          matchExpressions:
          - key: value
            operator: Exists
          - key: key2
            operator: NotIn
            values:
            - a
            - b
          - key: key3
            operator: DoesNotExist
        namespaces:
        - namespace1
        topologyKey: top
      preferredDuringSchedulingIgnoredDuringExecution:
      - weight: 100
        podAffinityTerm:
          labelSelector:
            matchLabels:
              key: a
            matchExpressions:
            - key: key
              operator: In
              values:
              - a
              - b
            - key: key2
              operator: NotIn
              values:
              - b
          namespaces:
          - namespace2
          topologyKey: toptop
  tolerations:
  - effect: PreferNoSchedule
    key: equalToleration
    tolerationSeconds: 30
    operator: Equal
    value: kek
  - key: equalToleration
    operator: Exists
    effect: PreferNoSchedule
    tolerationSeconds: 30