Configure Managed Cloud

Version: 10.1

This topic describes how to scale, size, and tune your Managed Cloud container solutions. It describes how to find and work with the configuration files in the infrastructure and application repositories.

Manage the size of Azure components

You might want to customize your Azure Kubernetes Service (AKS) cluster size, SQL database, and/or Azure Container Registry (ACR) tier size.

To manage the size of Azure components:

In the Infrastructure repository, go to: .\config\resources\resources.json

Manage the AKS cluster size, SQL database, and/or ACR tier size by updating the following parameters:

Parameter	Description
windows_vm_size	The Windows VM AKS size.
windows_scaled_vm_size	The Windows Scaled VM Size. Used for more efficient scaling of Content Delivery and XDBCollection pods.
windows_node_count	The amount of Windows nodes. A minimum of two windows nodes is required to support in-place node upgrades.
windows_scaled_node_count	The amount of Windows nodes for scaled resources. Could be set to 0 for some SKU tiers. If the value is 0, scaled pods (cd and xdb collection) are deployed to regular Windows VMs.
linux_node_count	The amount of Linux nodes.
windows_node_priority	The node priority for Windows. For production scenarios always use `regular`.
size_gb	SQL pool size in Gb.
sku_capacity	SQL pool SKU capacity.
sku_name	SQL pool SKU name.
sku_family	SQL pool SKU family.
sku_tier	SQL pool SKU tier.
sku	ACR tier size.

Schedule maintenance windows for AKS

You can use Planned Maintenance to schedule maintenance windows for your AKS cluster. Planned Maintenance lets you schedule weekly maintenance windows and minimizes workload impact. Once scheduled, all your maintenance will occur during the window you selected.

To add a maintenance window:

In Azure CLI, add the az aks maintenanceconfiguration command:

RequestResponse

az aks maintenanceconfiguration add -g MyResourceGroup --cluster-name myAKSCluster --name default --weekday Monday  --start-hour 1

Tune limits for AKS pods

You can manage the limits of every Sitecore deployment pod in the AKS cluster.

To tune limits for AKS pods:

In the Application repository, open .\config\resources\resources.json. You can see the list of all pods running in the AKS cluster. For example:

Tune the pods by updating the following parameters:

Parameter	Description
replicas	The number of Sitecore replicas.
request_memory	The default pod memory.
request_cpu	The default pod CPU.
limit_memory	The maximum pod memory limit.
limit_cpu	The maximum pod CPU limit.

Use regular Virtual Machines for nonproduction environments

By default, for nonproduction environments, Azure Kubernetes Service uses the Regular instance type of Virtual Machines. If you think the Regular instance is not appropriate for your nonproduction setup, you must switch the Windows node type to Spot. They are cheaper than regular Virtual Machines and can be evicted at any time. This makes Spot instances ideal for nonproduction test environments, but not so much for production.

To switch the Windows node type from Regular to Spot:

Go to config/resources/resources.json and change the windows_node_priority to Spot.

Add a custom image

To store custom images, Sitecore uses Private Azure Container Registry (ACR). Private ACR is created as part of Sitecore Managed Cloud Infrastructure.

To add an image to Private ACR:

Build a new version of an image based on an existing Sitecore image.
Push your image to Private ACR.
Update the docker-images.json file with the new image.

Important

To change docker-images.json, you must create a pull request to main.

Update external images

To update external images:

Update docker-images.json with a new image version.

The application pipeline runs automatically and applies new images from the file as soon as the main branch is updated.

Deploy new images by running the Application pipeline

The Application pipeline uses a set of images to deploy the Sitecore environment. All image names and locations are specified in the Application repository .\config\docker-images\docker-images.json file. Sitecore uses Ansible to deploy your Docker container images. To deploy, Ansible gathers the images from the docker-images.json file located here:

application\config\docker-images.json

This file contains a list of images that are required to deploy the Sitecore application.

The docker-images.json file has the following structure:

RequestResponse

{
  ...
  "sitecore": {
    "sitecore role name": "{image repository}/{role name}:{image tag}"
  }
  ...
  "entity name": {
     "image name": "{image repository}/{role name}:{image tag}"
   } 
  ...
}

To update Sitecore images:

Go to application\config\ and open the docker-images.json file. You can replace existing roles with your custom images based on existing base Sitecore images.

Add a new infrastructure module

You can create new resources in Azure by adding custom infrastructure modules. To add an infrastructure module, you must create a new Terraform module - a set of terraform configuration files in a single directory.

To add a new infrastructure module:

This example deploys the Azure Log Analytics workspace module. You can use a similar procedure to deploy any Azure resource.

Go to the Infrastructure repository and create a new folder for the module, for example: .\modules\log-analytics-workspace.

In the new folder, create main.tf with a resource description and the module name:

RequestResponse

resource "azurerm_log_analytics_workspace" "alaw" {
  name                = var.alaw_name
  location            = var.location
  resource_group_name = var.resource_group_name
  sku                 = var.sku
  retention_in_days   = var.retention_in_days
}

In the new folder, create variables.tf with a variables list:

RequestResponse

variable "location" {
  type = string
}

variable "resource_group_name" {
  type = string
}

variable "alaw_name" {
  type = string
}

variable "sku" {
  type = string
  default = "PerGB2018"
}

variable "retention_in_days" {
  type = string
  default = 30
}

In the root folder, edit the original .\main.tf and add the new module and pass the desired variable values:

RequestResponse

Module "alaw" {
  count               = 1
  source              = "./modules/log-analytics-workspace"
  location            = local.location
  resource_group_name = var.resource_group_name
  alaw_name           = "${local.infrastructure_id}alaw"
}

Complete the pull request and the Infrastructure pipeline is executed automatically and the changes are applied.

Alternatively, you can run the Infrastructure pipeline from the branch - this will not apply the changes and will only show the planned changes.

Change the Sitecore topology size

By default, all parameters are configured based on the original Sitecore topology size. In the .\sku\ folder, there are sets of predefined Sitecore topology sizes. You can use these configuration files to scale topology sizes.

As soon as you commit changes to the main branch, the Infrastructure or Application pipeline is triggered automatically and applies changes from the file. If you must apply a new topology size, make the relevant changes in the Infrastructure repository first. Wait until the Infrastructure pipeline is completed successfully and then apply the necessary changes to the Application repository..\config\resources\resources.json

To change the topology size:

In the Infrastructure repository, copy .\sku\resources.{size}.json and rename the file to resources.json.

Change Sitecore administrator credentials

Before you deploy your Sitecore installation, you must change the administrator password to a strong password. After you change these credentials, you must update the following secrets in the Azure Key Vault:

Sitecore administrator username: sitecore-admin-username
Sitecore administrator password: sitecore-admin-password

Do you have some feedback for us?

If you have suggestions for improving this article,