Monitoring for MongoDB cluster on Azure Kubernetes Service (AKS)

06/06/2025

In this article, you learn how to monitor a MongoDB cluster on Azure Kubernetes Service (AKS) using Percona Monitoring and Management (PMM). Monitoring is essential for ensuring the health and performance of your MongoDB cluster. By monitoring key metrics, you can identify issues early and take corrective actions to prevent downtime and data loss.

Set up Percona Monitoring and Management (PMM) for MongoDB

PMM operates as a client/server application, consisting of a PMM Server and multiple PMM Clients installed on each node of the database you want to monitor. The PMM Clients gather essential metrics and transmit this data to the PMM Server. You can then connect to the PMM Server to view database metrics across various dashboards.

Enable monitoring Percona Server for MongoDB with PMM

To enable monitoring for Percona Server for MongoDB, you need to make sure that the pmm.enabled field is set to true in the PerconaServerMongoDB custom resource. You can check this setting using the kubectl describe command.

kubectl describe PerconaServerMongoDB ${AKS_MONGODB_CLUSTER_NAME} -n ${AKS_MONGODB_NAMESPACE}

Example output:

Name:         cluster-aks-mongodb
Namespace:    mongodb
...
Spec:
  ...
  Pmm:
    Enabled:      true
    Image:        <your-acr-name>.azurecr.cn/pmm-client:2.41.2
    Server Host:  monitoring-service
  ...

If the pmm.enabled field is set to false, you can enable it using the kubectl patch command.

kubectl patch PerconaServerMongoDB ${AKS_MONGODB_CLUSTER_NAME} -n ${AKS_MONGODB_NAMESPACE} --type='merge' -p '{"spec":{"pmm":{"enabled":true}}}'

Create a node pool for PMM server

Create a dedicated node pool for the PMM server using the az aks nodepool add command.

az aks nodepool add \
--resource-group $MY_RESOURCE_GROUP_NAME \
--cluster-name $MY_CLUSTER_NAME \
--name pmmsrvpool \
--node-vm-size Standard_DS4_v2 \
--node-count 3 \
--zones 1 2 3 \
--mode User \
--output table

Create a dedicated node pool for the PMM server using terraform.

Run the following command to update the mongodb.tfvars file created earlier with the following configuration:

sed -i '/mongodbserver = {/,/}/s/^\( *}\)/  }, \
pmmserver = {\
  name       = "pmmsrvpool" \
  vm_size    = "Standard_D2ds_v4" \
  node_count = 3 \
  zones      = [1, 2, 3] \
  os_type    = "Linux" \
}/' mongodb.tfvars

Apply the terraform configuration to the target resource.

terraform fmt
terraform apply -var-file="mongodb.tfvars" -target module.default

Install PMM server

Add and update the Percona Helm repository using the helm repo add and helm repo update commands.
```
helm repo add percona https://percona.github.io/percona-helm-charts/
helm repo update
```

Install the PMM server using the helm install command.

helm install pmm --namespace mongodb \
--set nodeSelector."kubernetes\.azure\.com/agentpool"="pmmsrvpool" \
--set service.type="LoadBalancer" \
--version 1.3.13 \
percona/pmm

Get the external IP address of the PMM server using the kubectl get svc command.

export SERVICE_IP=$(kubectl get svc --namespace mongodb monitoring-service -o jsonpath="{.status.loadBalancer.ingress[0].ip}")
  echo https://$SERVICE_IP:

Get the password for the admin user using the kubectl get secret command.

  export ADMIN_PASS=$(kubectl get secret pmm-secret --namespace mongodb -o jsonpath='{.data.PMM_ADMIN_PASSWORD}' | base64 --decode)

You can now access the PMM server by using the external IP address and the "admin" user password. You can also access the PMM server on your local machine using the kubectl port-forward command.
```
kubectl port-forward svc/monitoring-service -n mongodb 8080:80
```
Once port forwarding is enabled, you can access the PMM server by navigating to http://localhost:8080.

Configure PMM server

Access the PMM server using the admin user and the password you obtained earlier.
From the side menu, select Configuration > API keys.
On the Configuration page, select Add API Key.
Enter the following information:
- Key name (you can give any desired name)
- Select the Admin role from the drop-down
- Enter a value in the Time to live text box (hover on the tooltip for more information)
Select Add.
Copy your key from the API Key Created window and export it to the environment variable as shown below.
```
export API_KEY="Paste your API Key here"
```

Set API key as secret and refresh ExternalSecret resource with the new secret

Set the API key as a secret in Azure Key Vault using the following command.

az keyvault secret set --vault-name $MY_KEYVAULT_NAME --name PMM-SERVER-API-KEY --value $API_KEY --output none

Run the following command to update the mongodb.tfvars file created earlier and apply the terraform configuration to the target resource with the following commands:

sed -i "/mongodb_kv_secrets = {/,/^ *}/s/\(PMM-SERVER-API-KEY *= *\"\)[^\"]*\"/\1$API_KEY\"/" mongodb.tfvars

terraform fmt

terraform apply -var-file="mongodb.tfvars" -target module.mongodb[0].azurerm_key_vault_secret.this

Refresh the secret in the ExternalSecret resource with the new secret key using the kubectl annotate command.

kubectl annotate es ${AKS_MONGODB_SECRETS_NAME} force-sync=$(date +%s) --overwrite -n ${AKS_MONGODB_NAMESPACE}

Test PMM server

Once you configure the PMM server, you can generate test data using Locust, as described in Validate resiliency of MongoDB cluster on Azure Kubernetes Service (AKS). After generating test data, you can monitor the performance of your MongoDB cluster on the PMM server.

Screenshot of a web page showing the PMM Mongodb dashboard.

Screenshot of another web page showing the PMM Mongodb dashboard.

Next steps

In this article, you learned how to monitor a MongoDB cluster on Azure Kubernetes Service (AKS) using Percona Monitoring and Management (PMM). To learn more about deploying stateful workloads on AKS, see the following articles: