konflikted / mssql-helm

Readme.md

This HELM chart is a sample "as-is" chart provided for reference to help guide with SQL Server deployment on Kubernetes cluster.

Prerequisites:

1. This chart is built on helm v3. It requires a kubernetes cluster to be running for you to deploy SQL container using this chart.
2. Ensure you have the helm installed on the client from where you will connect to the kubernetes cluster to deploy using the helm chart.
3. For minimum hardware requirement for the host to run SQL Server containers please refer to the system requirements section for SQL on Linux.
4. Requires the following variables to be set or changed in the values.yaml file :
   a. Please ensure that you accept the EULA for SQL Server, by changing the value of ACCEPT_EULA.value=y in values.yaml file or set it during the helm install command --set ACCEPT_EULA.value=Y.
   b. Please do choose the right edition of SQL Server that you would like to install you can change the value of the MSSQL_PID.value in the values file to the edition that you want to install or you can also change it during the helm install command using the option --set MSSQL_PID.value=Enterprise, If you do not pass the flag and do not change it in the yaml, then by default it is going to install developer edition.
   c. Also please do provide your customized value for the sa_password, if you do not provide it then by default the sa_password will the value as shown in the below table.
   

Note: Once you deploy SQL server containers using the chart below, please log into SQL Server using sa account and change the password for sa, this ensures that as DBA you have the control of the sa user and password.

Chart usage:

On the client machine where you have the Helm tools installed, download the chart on your machine and make the required changes to the values.yaml file as per your requirement. To see the list of settings that can be changed using the values.yaml file please refer to the table below.

Configuration parameters	Description	Default_Value
Values.image.repository	The SQL image to be downloaded and used for container deployment.	mcr.microsoft.com/mssql/server
Values.image.tag	The tag of the image to be download for the specific SQL image.	2022-latest
Values.ACCEPT_EULA.value	Set the ACCEPT_EULA variable to any value to confirm your acceptance of the SQL Server EULA, please refer environment variable for more details.	Y
Values.MSSQL_PID.value	Set the SQL Server edition or product key. please refer environment variable for more details	Developer
Values.MSSQL_AGENT_ENABLED.value	Enable SQL Server Agent. For example, 'true' is enabled and 'false' is disabled. By default, agent is disabled. please refer environment variable for more details.	TRUE
Values.containers.ports.containerPort	Port on which the SQL Server is listening inside the container.	1433
Values.podSecurityContext.fsgroup	Security context at the pod level.	10001
Values.service.port	The service port number.	1433
Values.replicas	This value controls the number of SQL Server deployments that would be done, consider this as the number of SQL Server instances that will run.	3

Deployment details:

Note

Here are my deployment details, please make changes to the values.yaml or other files as per your requirement.

In this scenario, I am deploying three SQL Server containers on a Azure Kubernetes Service (AKS) as statefulset deployments. You can follow Setup and connect to AKS to read instructions on setting up AKS and connecting to it. Also the storage class that I am using here is "Azure-disk". Please do find details below for each of the yaml file used in the template folder of this chart.

File Name	Description
_helpers.tpl	Template file with all the template definitions that will be used in this chart.
deployment.yaml	A manifest file to describing the deployment details for SQL Server.
mssqlconfig.yaml	SQL server mssql.conf file and its content that you would like to mount to the SQL Server container. For parameters that you can pass in this file please refer mssql.conf documentation. To modify the mssql.conf settings please modify this file.
sc.yaml	A manifest file that describes the storage class (SC) to be deployed. To make any changes to the sc please modify this file accordingly.
service.yaml	A manifest file that defines the kubernetes service type and port. Because this is a statefulset deployment, this manifest files helps in creating the headless service. Please modify this for any service modification that is needed.

With this information, and probably after you have modified the required files you are now ready to deploy SQL Server using this chart. From the client machine where you have the helm chart installed, change the directory of the CLI to the directory where you have the chart downloaded and to deploy SQL Server using this chart run the command:

helm install mssql . --set ACCEPT_EULA.value=Y --set MSSQL_PID.value=Developer

After a few minutes this should deploy the SQL Server containers and you can see all the artifacts using the command :

D:\helm-charts\sql-statefull-deploy>kubectl get all

The output should look as shown below:

NAME                               READY   STATUS    RESTARTS   AGE
pod/mssql-sql-statefull-deploy-0   1/1     Running   0          12m
pod/mssql-sql-statefull-deploy-1   1/1     Running   0          12m
pod/mssql-sql-statefull-deploy-2   1/1     Running   0          12m

NAME                                   TYPE           CLUSTER-IP     EXTERNAL-IP       PORT(S)          AGE
service/mssql-sql-statefull-deploy     ClusterIP      None           <none>            1433/TCP         16m

NAME                                          READY   AGE
statefulset.apps/mssql-sql-statefull-deploy   3/3     16m

This chart also includes an extra folder called "services" this folder has two more manifest files as described below:

Name	Description
ex_service.yaml	This containes sample code to create the external load balancer service for each of the pods created above, so SQL Server can be accessed outside the cluster
ag_endpoint.yaml	This containes sample code to expose the AG endpoint ports within the cluster, so each pod can talk to one another on the AG port. This will be needed if you are setting up AG between the pods.

Once you deploy the above files as well, using the commands shown below, you should have an external load balancer service created for each of the pods and another cluster IP service for each of the pod exposing the AG (alwayson) port for each pod within the cluster.

D:\helm-charts\sql-statefull-deploy>kubectl apply -f "D:\helm-charts\sql-statefull-deploy\services\ex_service.yaml"
D:\helm-charts\sql-statefull-deploy>kubectl apply -f "D:\helm-charts\sql-statefull-deploy\services\ag_endpoint.yaml"

Finally, after all the deployments here are the resources that you should see:

D:\>kubectl get all
NAME                               READY   STATUS    RESTARTS   AGE
pod/mssql-sql-statefull-deploy-0   1/1     Running   0          127m
pod/mssql-sql-statefull-deploy-1   1/1     Running   0          126m
pod/mssql-sql-statefull-deploy-2   1/1     Running   0          125m

NAME                                   TYPE           CLUSTER-IP     EXTERNAL-IP       PORT(S)          AGE
service/kubernetes                     ClusterIP      10.0.0.1       <none>            443/TCP          220d
service/mssql-mirror-0                 ClusterIP      10.0.148.0     <none>            5022/TCP         124m
service/mssql-mirror-1                 ClusterIP      10.0.254.58    <none>            5022/TCP         124m
service/mssql-mirror-2                 ClusterIP      10.0.196.129   <none>            5022/TCP         124m
service/mssql-sql-statefull-deploy     ClusterIP      None           <none>            1433/TCP         127m
service/mssql-sql-statefull-deploy-0   LoadBalancer   10.0.238.203   104.211.231.206   1433:30923/TCP   124m
service/mssql-sql-statefull-deploy-1   LoadBalancer   10.0.96.108    104.211.203.78    1433:32695/TCP   124m
service/mssql-sql-statefull-deploy-2   LoadBalancer   10.0.78.10     104.211.203.159   1433:31042/TCP   124m

NAME                                          READY   AGE
statefulset.apps/mssql-sql-statefull-deploy   3/3     127m

Connect to SQL Server

Now you are ready to connect to the SQL Server using any of the familiar tools that you work with, like the SSMS (SQL Server Management Studio) or SQLCMD or ADS (Azure Data Studio), etc. The IP address that you will use to connect is the External-IP address for the pod service which in this case one such example is: to connect to mssql-sql-statefull-deploy-0 SQL Server, the IP address 104.211.231.206 will be used in ssms or any other client.

For more details on the SQL Server deployment on AKS using manual method please refer Deploy a SQL Server container in Kubernetes with Azure Kubernetes Services (AKS).