Automated Content Deployment - Commingled Data

This article describes how to perform an automated deployment of a baseline shared data application and a baseline content application between two Qrvey instances in separate environments. This article assumes a commingled tenant data model. For more information on tenant data models, see Multi-tenant Dataset Architecture.

Before You Begin

• All of the API endpoints referenced in this document require you to authenticate with the Qrvey Admin Center API using an encrypted JWT token. When making a call to a Qrvey Admin Center endpoint, you must pass the JWT token in the request header or the request will be denied. For more information, see Generating Security Tokens.
• Verify that you have at least one user account created in the Qrvey Admin Center with “Composer” role permissions. For more information, see Managing Users of Qrvey Composer.

Prepare Source App(s) for Deployment

Perform the following steps using the Content Deployment feature in the Qrvey Admin Center. Be sure that you are in the Dev environment:

1. Create a target server that points to the Dev instance with the Dev API key. This step will not need to be repeated for all future content deployments.
2. Create a release package with a single version that points to the Master Data App.
3. Create a release package with a single version that points to the Master Content App.
4. Create a deployment definition sourced from the Master Data App release package with the following configuration:
   • Mode = Create New Application
   • Content = Baseline (i.e. all content)
   • New Application Name = Master Data Application
5. Use the following settings for each Connection in your deployment definition:
   • Host URL = {{connection.host_url}}
   • Port = {{connection.port}}
   • Username = {{connection.user_name}}
   • Password = {{connection.password}}
   • Database = {{connection.database}}
6. Use the following settings for each Dataset in your deployment definition:
   • Select the Share Data with Everyone checkbox
   • Select the Load Data checkbox
7. Create a deployment definition sourced from the Master Content App release package with the following configuration:
   • Mode = Create New Application
   • Content = Baseline (i.e. all content)
   • New Application Name = Master Content Application

Note: You will see the shared dataset from the Master Data App added as a dependency to the deployment definition for the Master Content App, but that doesn’t mean that the shared dataset will be deployed again. Rather, you will map the source of the dataset views to the correct shared datasets in the target environment at the time the deployment job is executed.

Execute the following steps as part of a programmatic routine from the Dev environment:

1. Call the GetAllDeploymentDefinitions() endpoint.
2. Parse the items object array in the response body to find the object with the name property that matches Master Data App, as well as the object where name matches Master Content App.
3. Extract the definitionId property value from both objects returned in the response, because you will need them for the next API call.
4. Call the ExportDeploymentDefinition() endpoint, passing in the definition ID for the Master Data App deployment definition, making sure to capture the jobTrackerId value from the response.
5. Call the ExportDeploymentDefinition() endpoint, passing in the definition ID for the Master Content App deployment definition, making sure to capture the jobTrackerId value from the response.
6. Make a call to the GetJobStatus() endpoint for each of the job tracker IDs.
7. Repeat the previous step until the response returns a URL path to collect both of the ZIP files for the exported deployment definitions.

Note: The time it takes to export a deployment definition is heavily dependent on the number of content objects selected for deployment.

You should now have two separate ZIP files, one for the Master Data App and the other for the Master Content App. You should copy the ZIP files to a location that is accessible from the Prod environment before beginning the next set of steps.

Prepare Target App(s) for Deployment

Execute the following steps as part of a programmatic routine from the Prod environment:

1. Call the CreateServer() endpoint and pass in the following request parameters:
   • name = any name you want
   • description = any description you want
   • host = fully qualified URL to this Qrvey instance
   • apiKey = your Prod API key
2. Parse the response, extract the value from the adminserverid property and save it off somewhere so it can be recalled for future deployments. Perform steps 1 and 2 once…and only once.
3. Call the GetUploadURL() endpoint for deployment definitions and save the url and key properties from the response.
4. Use the URL provided to make a separate POST request to upload the ZIP file for the Master Data App to the target S3 bucket.
5. Wait for the ZIP file to finish uploading to the target S3 bucket, and then call the UploadDeploymentDefinition() endpoint, setting the following request parameters:
   • key = “key” value from the GetUploadURL() response
   • definitionName = any name you want
   • description = any description you want that describes the content you are deploying
6. Repeat steps 3 - 5 for the ZIP file associated with the Master Content App.

Note: Before you continue, please make sure you have at least one user account created in the Qrvey Admin Center with “Composer” role permissions.

Deploy the Master Data App

Execute the following steps as part of a programmatic routine from the Prod environment:

1. Call the CreateDeploymentJob() endpoint, passing in any name and description that you want.
2. Extract the value from the deploymentJobId property in the response.
3. Call the GetAllDeploymentDefinitions() endpoint.
4. Parse the items object array in the response body to find the object with the name property that matches Master Data App.
5. Extract the definitionId property value returned in the response, because you will need it for the next API call.
6. Call the CreateDeploymentJobBlock() endpoint, passing in the following request parameters:
   • definitionId = ID of the Master Data App deployment definition
   • adminServerId = ID of the target server to deploy the content to (Prod)
   • selectAllUsers = false
7. Extract the deploymentJobBlockId value from the response.
8. Call the GetUserList() endpoint, parse the items array to find the user metadata for the account that will become the owner of this app, and then extract the value from the corresponding userid property.
9. Call the AddRecipientsToDeploymentJobBlock() endpoint, using the deploymentJobId and deploymentJobBlockId as path parameters for calling the endpoint. Use the following request body JSON:

{ 
 "users": [
   {
     "updateDate": NOW,
     "createDate": NOW,
     "userid": APP_OWNER_USER_ID,
     "parameters": [
       {
         "token": HOST_URL_PARAM_NAME,
         "value": HOST_URL_PARAM_VALUE
       },
       {
         "token": DB_PORT_PARAM_NAME,
         "value": DB_PORT_PARAM_VALUE
       },
       {
         "token": DB_USERNAME_PARAM_NAME,
         "value": DB_USERNAME_PARAM_VALUE
       },
       {
         "token": DB_PASSWORD_PARAM_NAME,
         "value": DB_PASSWORD_PARAM_VALUE
       }, 
       {
         "token": DB_NAME_PARAM_NAME,
         "value": DB_NAME_PARAM_VALUE
       }
     ],
     "status": "PUBLISHED",
     "email": APP_OWNER_USER_EMAIL,
     "deploymentJobBlockId": DEPLOYMENT_JOB_BLOCK_ID
   }
 ]
}

10. Make sure you pass values for each of the parameters you created for the shared dataset’s connection information. You should have at least one parameter for each connection’s host URL, which will need to change when you deploy this app and load the shared datasets.
11. Call the ExecuteDeploymentJob() endpoint, passing in the value of the deploymentJobId as a path parameter to the endpoint.
12. Parse the jobTrackerId value from the response.
13. Call the GetJobStatus() endpoint, passing in the value for the jobTrackerId and wait until the status property has a value of CREATED. This endpoint returns lots of useful information about the deployment of the content objects.

Note: You must wait until the Master Data App has finished deploying before continuing with the deployment of the Master Content App.

Deploy the Master Content App

Execute the following steps as part of a programmatic routine from the Prod environment:

1. Call the CreateDeploymentJob() endpoint, passing in any name and description that you want.
2. Extract the value from the deploymentJobId property in the response.
3. Call the GetAllDeploymentDefinitions() endpoint.
4. Parse the items object array in the response body to find the object with the name property that matches Master Content App.
5. Extract the definitionId property value returned in the response, because you will need it for the next API call.
6. Call the CreateDeploymentJobBlock() endpoint, passing in the following request parameters:
   • definitionId = ID of the Master Content App deployment definition
   • adminServerId = ID of the target server to deploy the content to (Prod)
   • selectAllUsers = false
7. Extract the deploymentJobBlockId value from the response.
8. Call the GetUserList() endpoint, parse the items array to find the user metadata for the account that will become the owner of this app, and then extract the value from the corresponding userid property.
9. Call the GetAllDatasets() endpoint to find the matching dataset IDs for all shared datasets that the dataset views should reference once deployed to the target environment.
10. Call the AddRecipientsToDeploymentJobBlock() endpoint, using the deploymentJobId and deploymentJobBlockId as path parameters for calling the endpoint. Use the following request body JSON:

{ 
 "users": [
   {
     "updateDate": NOW,
     "createDate": NOW,
     "userid": APP_OWNER_USER_ID,
     "parameters": [
       {
         "token": SHARED_DATASET_CONTENT_TOKEN_ID,
         "value": SHARED_DATASET_ID
       }
     ],
     "status": "PUBLISHED",
     "email": APP_OWNER_USER_EMAIL,
     "deploymentJobBlockId": DEPLOYMENT_JOB_BLOCK_ID
   }
 ]
}

11. Call the ExecuteDeploymentJob() endpoint, passing in the value of the deploymentJobId as a path parameter to the endpoint.
12. Parse the jobTrackerId value from the response.
13. Call the GetJobStatus() endpoint, passing in the value for the jobTrackerId and wait until the status property has a value of CREATED. This endpoint returns lots of useful information about the deployment of the content objects.

Note: The content token for a shared dataset begins with the string shared_data. Here is an example of what the content token might look like for a shared dataset called “Demo Data”:
{{shared_data.demo_data}}

You should always pass the shared dataset’s ID when setting the content token for a deployment job block, in case there are multiple datasets with the same name in the target environment.

API Reference

• Login to Qrvey Admin Center
• Create Server
• Get All Deployment Definitions
• Export Deployment Definition
• Get Job Status by jobTrackerId
• Get Upload URL (Definitions)
• Upload Deployment Definition
• Get User List
• Get All Datasets
• Create Deployment Job
• Create Deployment Job Block
• Add Recipients to Deployment Job Block