Deploying Elastic Path Commerce

After the infrastructure is setup, you are ready to start deploying Elastic Path Commerce.

Prerequisites

Ensure that you complete the following tasks before proceeding to deploy an Elastic Path Commerce environment using CloudOps for Kubernetes:

Complete the CloudOps for Kubernetes bootstrap process and post-bootstrap tasks
Build the required Elastic Path Commerce artifacts
Deploy the Apache ActiveMQ service
Create and populate the Elastic Path Commerce database

For more information about these prerequisites, including any links to the required documentation, see the Deploy Commerce Overview section.

Run the Jenkins job deploy-or-delete-commerce-stack with the following parameters:

`plan_mode`

When selected, this runs Terraform in plan mode and prompts you to approve changes before you continue. This works in both the deploy and delete modes.

`deleteStack`

When selected, this deletes the stack instead of creating it.

`epStackResourcingProfile`

This parameter is the resourcing profile of your ep-stack. By default, CloudOps for Kubernetes provides prod-small and dev profiles.

For more information about adding custom resourcing profiles, see the Deploy with different resourcing profiles section.

`restartApps`

By default, when you update an Elastic Path Commerce deployment with the same image tag, deployments are not restarted. To restart the Elastic Path Commerce deployments, set this parameter to true.

`cloudOpsForKubernetesRepoURL`

The Git repository URL of the cloudops-for-kubernetes code.

`cloudOpsForKubernetesBranch`

The branch of cloudops-for-kubernetes to use.

`epEnvironment`

The type of environment and configuration of your ep-stack, such as CI, production or staging.

`dockerImageTag`

The tag of the ep-app Docker images to use.

`includeMockService`

Create, update, or delete the Wiremock (mock) service. Wiremock is a tool used in Elastic Path Commerce development and test environments. For more information about the Wiremock service, see the Elastic Path Commerce Git repository, under <ep-commerce>/extensions/http-mock-server.

`kubernetesNickname`

This job uses the value of kubernetesNickname as the Kubernetes namespace where the Elastic Path Commerce stack is deployed, and in the domain name of the Elastic Path Commerce stack. This nickname must match the nickname used by the populated MySQL database and ActiveMQ services that you previously deployed and want the Elastic Path Commerce stack to use.

If you specify a nickname and namespace value where an Elastic Path Commerce stack is already deployed, then that stack will be updated.

`includeIngresses`

Creates or deletes Ingresses to allow connections from the network CIDR values in allowedCIDRs.

`allowedCIDRs`

The network CIDR allowed to access the Commerce stack. Ignore this parameter if createIngresses is false.

The network CIDR the Ingresses allow to access the cortex studio app service. Ignore this parameter if createIngresses is false.

`TF_VAR_enable_firewall_rules`

You can attach the Commerce stack to the AWS Web Application Firewall for the Elastic Path Commerce and Account Management services. Deploy CloudOps for Kubernetes with the parameter TF_VAR_aws_enable_commerce_am_waf set to true in the Docker Compose configuration file, docker-compose.yml.

If TF_VAR_enable_firewall_rules is set, the web application firewall service is applied to the Commerce stack services. The firewall is also updated to allow the CIDRs in allowedCIDRs access to the Commerce stack services.

`includeHorizontalPodAutoscalers`

Creates or deletes horizontal pod autoscalers. It automatically scales the number of pods in the deployment based on CPU usage. The default setting is false.

`includeDeploymentInfoPage`

Creates or deletes a page displaying the cluster details, database details and links to the deployed applications. The default setting is false.

`enableUITests`

If set to true, UI tests are enabled for the deployment. The default setting is false.

`epChangesetsEnabled`

If it is set to true, it enables changeset for deployment. The default setting is false.

`deployDstWebapp`

Deploys or deletes the Data Sync tool webapp. The Data Sync Tool (DST) can only be deployed if epChangesetsEnabled is true and the stack is Elastic Path Commerce 7.5 or later.

`kubernetesTargetNickname`

The Kubernetes namespace of the target database that the DST pushes changes to. Leave empty if deployDstWebapp is set to false.

`dnsZoneName`

The domainName set in the docker-compose.yml file used during bootstrap. If the DNS settings for the domain are manually configured, the dnsZoneName can be overridden.

`clusterName`

The name of the Kubernetes cluster to deploy the ep-stack into. The clusterName is also the name of the A record and part of the full DNS name of the ep-stack.

`includeAWSSES`

If the cloud platform is AWS, when checked, the job deploys the Elastic Path Commerce stack with AWS Simple Email Service (SES) email settings. This job only can run if setup-AWS-SES is run before with the same kubernetesNickname value.

`smtpHost`

If includeAWSSES is false and you want to use an external SMTP server, provide the address of the external SMTP server. For example, if you are using SendGrid as an SMTP service provider, the value is smtp.sendgrid.net.

`smtpPort`

If includeAWSSES is false and you want to use an external SMTP server, provide the port number of the external SMTP server. For example, if you are using SendGrid as an SMTP service provider, you could enter a value of 587.

`smtpScheme`

If includeAWSSES is false and you want to use an external SMTP server, provide the scheme of the connection to the external SMTP server.

Use one of the following schemes when connecting to the SMTP server:

smtp: for a plaintext SMTP connection
smtps: for SMTP wrapped in SSL

`smtpUser`

If includeAWSSES is false and you want to use an external SMTP server, provide the username for the connection to the external SMTP server.

`smtpPass`

If includeAWSSES is false and you want to use an external SMTP server, provide the password for the connection to the external SMTP server.

`enableJmx`

This enables connecting to the app JMX ports. The default setting is false. For more information about connecting to JMX see the troubleshooting section.

`enableJmxAuth`

This enables JMX user and password authentication and generates a unique user and password combination that is added to the ConfigMap for the Elastic Path Commerce deployment. The default setting is false. For more information about connecting to JMX see the troubleshooting section.

`enableDebug`

This enables connecting to the application Java Virtual Machine (JVM) debug ports with the remote debugger of your IDE. The default setting is false.

Deployment Options

Creating Ingresses

This option deploys Elastic Path Commerce with Ingresses that whitelist a CIDR in the parameter allowedCIDRs. The default value 127.0.0.1/32 in allowedCIDRs blocks all public access. The Elastic Path stack won’t allow incoming connections. You cannot access your applications through the DNS endpoints unless Ingresses are created and your CIDR is in the job parameter allowedCIDRs.

warning

Attempting to create Ingresses with no value in allowedCIDRs results in a failed deployment. We recommend testing that the annotation is working correctly.

Create Horizontal Pod Autoscalers

This option deploys an Elastic Path stack with Horizontal Pod Autoscalers. For more information, see Horizontal Pod Autoscalers.

Create a deployment information page

This option deploys a deployment information page. The deployment information page display Kubernetes cluster details, database details, and links to the deployed applications. For more information, see the Example Information Page image.

warning

The deployment information page is not secured beyond IP whitelisting. It is intended for use with test environments. Do not deploy alongside production environments.

Choosing to create the Deployment Information Page exposes the ActiveMQ admin console to the public internet according to the whitelist-source-range set during configuration, with no additional security.

Example Information Page

The path to the information page from root DNS for your deployment is defined as ${kubernetesNickname}.commerce${clusterName}.${dnsZoneName}/.

Deploy with UI tests enabled

This option deploys Elastic Path Commerce with UI Tests enabled.

Deploy with change sets enabled

This option deploys Elastic Path Commerce with change sets enabled.

Deploy the Data Sync Tool

This option deploys the Data Sync Tool (DST) as a command-line tool and as a Java application. The DST can only deploy as a Java application on Elastic Path version 7.5 or later. Deploying the DST requires values for the kubernetesTargetNickname parameter.

For more information about the Data Sync Tool, see Elastic Path Production Tools.

Deploy with JMX enabled

This option allows connections to JMX for the Elastic Path Commerce applications with a Java profiler. For more information, see Connecting to the Applications JVMs.

Deploy with debug enabled

This option allows connections to the JVM in debug mode from your IDE. For more information, see Connecting to the Applications JVMs.

Deploy with different resourcing profiles

This option deploys an Elastic Path stack under a specific resourcing profile. CloudOps for Kubernetes provides the default resourcing profiles, dev, and prod-small. Details on the resources assigned for each application are under each resourcing profile in cloudops-for-kubernetes/terraform/commerce-stack/env-file. When creating your custom resourcing profile, use the existing dev.tfvars and prod-small.fvars files as templates.

There are two ways to add custom resourcing profiles when deploying an Elastic Path stack.

Add a custom tfvars resourcing profile file in your CloudOps for Kubernetes Git repository:
1. Add your custom resourcing profile to the cloudops-for-kubernetes/terraform/commerce-stack/env-file directory.
2. Push your changes to the CloudOps for Kubernetes repository and branch used by the Jenkins job deploy-or-delete-commerce-stack.
3. Run the Jenkins job deploy-or-delete-commmerce-stack with the Jenkins parameter epStackResourcingProfile set to the name of your custom resourcing profile. Do not include the file extension for this parameter.
Use a custom tfvars resourcing profile from an external Git repository:
1. Run the docker-compose process in setup mode with the parameter TF_VAR_bootstrap_extras_properties in the docker-compose override file using the same format as the example below. Replace the example values with the appropriate values.
  
  note
  
  Do not include the root Git repository directory in env_file_path.
```
TF_VAR_bootstrap_extras_properties: >
{
  "git_repos": {
    "extras-repo": {
      "repo_url": "<GIT_REPOSIORY_SSH_URL>",
      "default_branch": "<GIT_REPOSITORY_BRANCH>",
      "env_file_path": "<PATH_TO_THE_RESOURCING_PROFILE_DIRECTORY>"
    }
  }
}
```
2. Run the Jenkins job deploy-or-delete-commmerce-stack with the Jenkins parameter epStackResourcingProfile set to the name of your custom resourcing profile. Do not include the file extension for this parameter.

Deploy with different Ambassador gateway timeouts

By default, an Elastic Path stack is deployed with gateway timeouts. The timeouts are set to ten seconds for the Cortex service and three seconds for the other services.

To adjust the timeout values:

Edit the timeout values in the cloudops-for-kubernetes/terraform/commerce-stack/commerce-mappings.yaml.tmpl file.

For more information on the different timeout values, see Timeouts.
Commit and push these changes to your Git repository.
Deploy or update an Elastic Path stack with the cloudOpsForKubernetesBranch parameter set to the Git branch with the gateway timeout changes that you want.

Validating Deployment

The deploy-or-delete-commerce-stack job can take approximately 15-30 minutes to complete as it waits for the stack to become available.

Procedure

Open the Kubernetes dashboard and select the namespace you used when you ran the deploy-or-delete-commerce-stack Jenkins job.

tip

If you require, use the directions to login to the dashboard: Accessing Kubernetes Dashboard.

note

During the stack startup, Deployments, Services, and Pods in the Kubernetes dashboard show an error that the readiness probes are failing. These error messages are normal during startup and go away as the services become available.
Validate that the following Deployments are present:
- ep-activemq-deployment
- ep-cm-deployment
- ep-integration-deployment
- ep-searchmaster-deployment
- ep-searchslave-deployment
- ep-cortex-deployment
- ep-batch-deployment
- ep-data-sync-deployment (if deployDstWebapp is true)
Validate that the following Services are present:
- ep-activemq-service
- ep-cm-service
- ep-integration-service
- ep-searchmaster-service
- ep-searchslave-service
- ep-cortex-service
- ep-batch-service
- ep-data-sync-service (if deployDstWebapp is true)
If deployDstWebapp is true, validate that a secret named target-database-secrets exists.
Run the run-cortex-system-tests Jenkins job on the stack to validate that the Elastic Path stack is operational.

For more information about the job, see the run-cortex-system-tests job section.

URL reference

The Elastic Path stack URL is created from the kubernetesNickname, clusterName, and dnsZoneName parameters. This is used to create a fully qualified name for each service. For example, if clusterName = jDoeCluster, dnsZoneName = epcloud.mycompany.com and kubernetesNickname = dev the fully qualified domain name is dev.commercejDoeCluster.epcloud.mycompany.com.

The following table is a sample of the generated URL using the kubernetesNickname = dev, clustername = jDoeCluster and dnsZoneName = epcloud.mycompany.com:

Service	Example URL
Info Page	dev.commercejDoeCluster.epcloud.mycompany.com
Cortex	dev.commercejDoeCluster.epcloud.mycompany.com/cortex
Commerce Manager	dev.commercejDoeCluster.epcloud.mycompany.com/cm/
Integration Server	dev.commercejDoeCluster.epcloud.mycompany.com/integration

Prerequisites