Alteryx Promote Knowledge Base

  Often, when deploying a model up to Promote, the model requires certain dependencies to run. These dependencies can be certain functions, files, etc. If your model requires them, you’ll need to create a promote.sh, which contains commands to import these dependencies. This will be one of the factors needed to ensure your model will be set up for success on Promote, because sometimes a model needs a little help.     If we go to https://github.com/alteryx/promote-python we can go into the article-summarizer example, which contains one of these promote.sh files. You’ll notice that if you open the file, you’ll see this command:   python -c "import nltk; nltk.download('punkt')" This is required because the newspaper package in the model (main.py) requires an NLP dataset. Now, when we deploy the model, the promote.sh file will run at the same time, which will ensure the dependencies live inside the model environment (docker model image). We can now properly test the model in Promote!   If we're looking at an R example (there is one on the Promote GitHub), you will have the same folder structure, except the promote.sh file will look something like this: curl https://packages.microsoft.com/keys/microsoft.asc | apt-key add - curl https://packages.microsoft.com/config/ubuntu/16.04/prod.list > /etc/apt/sources.list.d/mssql-release.list apt-get update # ACCEPT_EULA=Y ACCEPT_EULA=Y apt-get -y install msodbcsql17 apt-get -y install unixodbc-dev apt-get -y install r-cran-rodbc apt-get -y install libiodbc2-dev In this case, our model requires an ODBC driver, therefore our model container will also need it in order to run on Promote. Just as in the above Python example, when we deploy this model, the promote.sh file will run and the proper driver will be installed, enabling us to work and test this model on Promote!   Once you get these all set, you'll be good to venture on and make your model the best it can be!

How To: Backup the Promote PostgreSQL database   Promote uses a PostgreSQL database. This article explains how to create a backup of a Promote instance's database. Backing up your Promote database can save you time and headaches in the event of a system failure.   Prerequisites   Alteryx Promote  ≥ 2018.1   Procedure   The PostgreSQL database must be backed up from the Master node, and not just the Leader node (note: these can be the same node). To return a list of node IDs, run the following command: docker node ls --format='{{.ID}}' To check if a node is the Master node, run the following command on each node ID from the list returned by the above command, replacing {node_id} with the node ID. docker node inspect {node_id} --format='{{.Spec.Labels.master}}' Once you have determined the Master node (where "yes" is the returned response value from the above command), run the following commands from that node to start the PostgreSQL backup process. On the host machine  run  the following bash command : docker exec -it $(docker ps | grep promote-db | awk '{print$1}') bash Run the backup script: sh /scripts/backups.sh Change to the directory of the database backup: cd /var/backups/postgres Copy the backup file to the host machine (the node that is hosting the promote-db docker container) from the database container (promote_app). This command needs to be run outside of the database container, changing /location/path to the database path on the host machine: docker cp $(docker ps | grep promote-db | awk '{print$1}'):/var/backups/postgres/{backup db name} /location/path A backup of your database should now be saved in your specified directory. You deserve a coffee!

How To: Restore a Promote PostgreSQL Database   This article outlines the process of restoring the Promote PostgreSQL database from a backup. For instructions on creating a backup, please see How To: Create a Promote PostgreSQL Database Backup.   Use these steps only if there is no data in your database.    Only run these commands if: Your PostgreSQL database is corrupted. Your database is in a new state.   Prerequisites   Alteryx Promote  ≥  2018.2.1   Procedure   The PostgreSQL database must be restored from the Master node, and not just the Leader node (note: these can be the same node). To return a list of node IDs, run the following command: docker node ls --format='{{.ID}}' To check if a node is the Master node, run the following command on each node ID from the list returned by the above command, replacing {node_id} with the node ID. docker node inspect {node_id} --format='{{.Spec.Labels.master}}' Once you have determined the Master node (where "yes" is the returned response value from the above command), run the following commands from that node to start the PostgreSQL restoration process. Ensure the backed up PostgreSQL database is on your host machine.  Provide the Database Password with the following command: cat /var/promote/credentials/db.txt Copy the PostgreSQL database from the master node to the promote-db  container on the same node, where /location/path is the location of the backed-up database: docker cp /location/path $(docker ps | grep promote-db | awk '{print$1}'):/var/backups/postgres  On the host machine run the following bash command : docker exec -it $(docker ps | grep promote-db | awk '{print$1}') bash Restore the database within the promote-db container: pg_restore -c -U ${POSTGRES_USER} -d ${POSTGRES_DB} -v "/var/promote/postgres/{database name}" You should now be able to log into the UI and see your predictive models rebuild and go online. If not, please open a support ticket in the Case Portal.

  This article outlines the step-by-step procedure for enabling a set of backup nodes as a production cluster. This article does not describe how to capture the backup clones.  Typically, these backup clones should come from a snapshot. The specific process for capturing these clones will depend on your deployment's infastructure. 

Issue    Although the Admin page indicates that there are models deployed to the Promote instance, the models are not appearing in the Promote UI home page.     Environment   Alteryx Promote  ≥  2018.2.1   Cause   There was an issue with your web browser cache, causing an incomplete view of the Promote UI home page to load.   Solution   Clear your browser's cache for the Promote UI site address.  Follow the steps provided for your browser here. If following the steps to clear the cache for your browser does not resolve the issue, please open a support ticket using the Case Portal.

Troubleshooting steps for missing model logs.

Keras  is an  open-source  neural network API library, written in Python  designed to run on top of  TensorFlow ,  CNTK , or  Theano . In this article, we demonstrate how to deploy a Keras model to Promote.

Models deployed to Promote can be queried through a couple of different ways, one of them being a standard REST API post request. Querying a model consists of sending in the predictor variables to the model, allowing the model to process the data and make predictions. After the prediction is made from the model, the return is the score based on the predictor variables entered.

Did you know that you can use Promote to query a database (or include a database query in your Promote model)? 

An overview of user management in Promote. 

This article provides an overview of the administrative options in Promote (excluding user management, which can be found here).   Basic   Models   Within the Admin Dashboard, an Admin user can view a list of all the models deployed to any environment by any user by clicking on the Models tab.      System Overview   Within the Admin dashboard, an Admin user can monitor system health metrics for each node in the Promote cluster by clicking on the System Overview tab.     Advanced   Within the Admin dashboard, an admin user can adjust several settings that affect the performance and behavior of the system by clicking on the Advanced tab.    Base Image   An Admin user can change the base image used to deploy both R and Python models. An admin user may do this if they create a new image that has custom R or Python libraries available on it, or if they'd like to use a different version of R or Python.       Disk Bundle Limit   An admin user can change the disk bundle limit to protect the system against running out of disk space. The disk bundle limit limits the number of versions of a model that are stored on disk.            Prediction Logging   Promote can store logs for every prediction request made for up to 14 days, with a maximum of 50GB. You can toggle this logging on and off for Development/Staging and Production in this section.         We hope this gives you a good foundation for administering your Promote instance. Good luck, we're all counting on you.   

There are two tools in Alteryx Designer that connect to Promote; the   Deploy  tool and  the Score tool. The Deploy tool allows you to send trained models from Alteryx Designer to Promote. The Score tool allows you to connect to a model that has already been deployed to Promote to create predictions from a provided data set.    The first step is to have a model object from a trained model. You can use any of the standard Predictive Tools to train a model (including the R tool), as long it is not from the revoscaler package, which is not currently supported by Promote.   In this example, let's say we are interested in training a random forest model to predict forest type (classification) based on remotely sensed spectral data. The study area covers Japan, and the predictor variables include values for  visible to near-infrared wavelengths, derived from ASTER satellite imagery.    After performing some data investigation and pre-processing (this dataset is already very clean) we can create, refine, and ultimately select our model.     Once we have a model we are happy with, we can send it to Promote using the Deploy tool. You can start by adding a Deploy tool to the canvas and adding it to the O anchor of your selected model.      If you haven't already, connect your Alteryx Designer instance to Promote.   To being the process of adding a Promote connection, click the Add Connection button in the Configuration window of the Deploy tool.     After clicking the Add Connection button, a modal window will pop up on your screen. Type your Promote instance's URL in the first screen and click Next.       Now add your Username and API key.         For your API key, you may need to log in to your Promote instance and navigate to the Account page.         Once you have your username and API key correctly added to the modal window, click Connect. If all your information checks out, you will see this success message.     After clicking Finish, there will be an option in your Alteryx Promote Connection drop-down menu. You will also see a new option to Remove Connection.     To deploy a model, give it a name in the Model Name setting and run your workflow. If this is a new or updated version of a model that already exists on Promote, give it the same name as the currently deployed version, and check the Overwrite existing model option.   After running the workflow, if the model deploys successfully, you will see a message from the Deploy tool that says "your model is building, check the UI for build logs" in your results window.      To check the build logs, navigate back to the Promote UI in your web browser, click on your model, and then click on the Logs tab. You still see the messages from the model building process. If all is well, the log will end with a "model built successfully" message.        Your model now lives on Promote! 

One of the most important features of Promote is its ability to return near-real-time predictions from deployed models. Here is a list of frequently asked questions relating to Promote prediction requests.

Promote is data science model hosting and management software that allows its users to seamlessly deploy their data science models as highly available microservices that return near-real-time predictions by leveraging REST APIs. In this article, we provide an overview of Promote’s technical requirements and architecture.

If you’ve never heard of Docker, or aren’t particularly familiar with it, you are probably wondering “what’s the deal with Docker?”

If you have a model that takes longer than 10 seconds to return results, by default Promote will time out your model API query. If you would  like  Promote to wait longer than 10 seconds before timing out, you can adjust this timeout setting with an environmental variable called PREDICTION_TIMEOUT. 

When making a call to a Promote model, the input data used to make a prediction is sent in a JSON format. When working with an R model, prior to reaching the model.predict() function, the JSON string that was sent to your model is converted to an R format (either an R dataframe or an R list). By default, this conversion is performed with the fromJSON() function in the jsonlite R package.

Some Promote customers have run into an  issue  where the status of the predictive model will flicker between online and offline continuously on the Promote UI page. This article discusses the cause of the issue, as well as how to resolve it.

Welcome to part 3 of the Supporting Promote series. In this series, we will tackle some common issues and questions, and provide best practices for troubleshooting. This article will step through the process of restoring the Promote web app.