11/24/2021»»Wednesday

Install Container

  1. Yum Install Container-selinux
  2. Install Containers Hyper V

Step 2: Install NVIDIA Container Toolkit ¶. After installing podman, we can proceed to install the NVIDIA Container Toolkit. For podman, we need to use the nvidia-container-toolkit package. See the architecture overview for more details on the package hierarchy. First, setup the package repository and GPG key. To run a container, you first install a base image, which provides a foundational layer of operating system services to your container. Then you create and run a container image, which is based upon the base image. For details, read on. Install a container base image. All containers are created from container images.

-->

Applies to: SQL Server (all supported versions) - Linux

Note

The examples shown below use the docker.exe but most of these commands also work with Podman. It provides the CLI similar to Docker container Engine. You can read more about podman here.

Install ContainerInstall

In this quickstart, you use Docker to pull and run the SQL Server 2017 container image, mssql-server-linux. Then connect with sqlcmd to create your first database and run queries.

Tip

If you want to run SQL Server 2019 containers, see the SQL Server 2019 version of this article.

Note

  • Starting with SQL Server 2019 CU3, Ubuntu 18.04 is supported.
  • Starting with SQL Server 2019 CU10, Ubuntu 20.04 is supported.

In this quickstart, you use Docker to pull and run the SQL Server 2019 container image, mssql-server. Then connect with sqlcmd to create your first database and run queries.

Tip

This quickstart creates SQL Server 2019 containers. If you prefer to create SQL Server 2017 containers, see the SQL Server 2017 version of this article.

This image consists of SQL Server running on Linux based on Ubuntu 20.04. It can be used with the Docker Engine 1.8+ on Linux or on Docker for Mac/Windows. This quickstart specifically focuses on using the SQL Server on Linux image. The Windows image is not covered, but you can learn more about it on the mssql-server-windows-developer Docker Hub page.

Prerequisites

  • Docker Engine 1.8+ on any supported Linux distribution or Docker for Mac/Windows. For more information, see Install Docker. For more information on hardware requirements and processor support, see SQL Server 2019: Hardware and software requirements.
  • Docker overlay2 storage driver. This is the default for most users. If you find that you are not using this storage provider and need to change, see the instructions and warnings in the docker documentation for configuring overlay2.
  • Minimum of 2 GB of disk space.
  • Minimum of 2 GB of RAM.
  • System requirements for SQL Server on Linux.

Pull and run the 2017 container image

Before starting the following steps, make sure that you have selected your preferred shell (bash, PowerShell, or cmd) at the top of this article.

  1. Pull the SQL Server 2017 Linux container image from Microsoft Container Registry.

    Tip

    If you want to run SQL Server 2019 containers, see the SQL Server 2019 version of this article.

    The previous command pulls the latest SQL Server 2017 container image. If you want to pull a specific image, you add a colon and the tag name (for example, mcr.microsoft.com/mssql/server:2017-GA-ubuntu). To see all available images, see the mssql-server Docker hub page.

    For the bash commands in this article, sudo is used. On macOS, sudo might not be required. On Linux, if you do not want to use sudo to run Docker, you can configure a docker group and add users to that group. For more information, see Post-installation steps for Linux.

  2. To run the container image with Docker, you can use the following command from a bash shell (Linux/macOS) or elevated PowerShell command prompt.

    Note

    If you are using PowerShell Core, replace the double quotes with single quotes.

    Note

    The password should follow the SQL Server default password policy, otherwise the container can not setup SQL server and will stop working. By default, the password must be at least 8 characters long and contain characters from three of the following four sets: Uppercase letters, Lowercase letters, Base 10 digits, and Symbols. You can examine the error log by executing the docker logs command.

    By default, this creates a container with the Developer edition of SQL Server 2017. The process for running production editions in containers is slightly different. For more information, see Run production container images.

    The following table provides a description of the parameters in the previous docker run example:

    ParameterDescription
    -e 'ACCEPT_EULA=Y'Set the ACCEPT_EULA variable to any value to confirm your acceptance of the End-User Licensing Agreement. Required setting for the SQL Server image.
    -e 'SA_PASSWORD=<[email protected]>'Specify your own strong password that is at least 8 characters and meets the SQL Server password requirements. Required setting for the SQL Server image.
    -p 1433:1433Map a TCP port on the host environment (first value) with a TCP port in the container (second value). In this example, SQL Server is listening on TCP 1433 in the container and this is exposed to the port, 1433, on the host.
    --name sql1Specify a custom name for the container rather than a randomly generated one. If you run more than one container, you cannot reuse this same name.
    -h sql1Used to explicitly set the container hostname, if you don't specify it, it defaults to the container ID which is a randomly generated system GUID.
    -dRun the container in the background (daemon)
    mcr.microsoft.com/mssql/server:2017-latestThe SQL Server 2017 Linux container image.
  3. To view your Docker containers, use the docker ps command.

    You should see output similar to the following screenshot:

  4. If the STATUS column shows a status of Up, then SQL Server is running in the container and listening on the port specified in the PORTS column. If the STATUS column for your SQL Server container shows Exited, see the Troubleshooting section of the configuration guide. It will be ready for connection, once the SQL Server error logs display the message: SQL Server is now ready for client connections. This is an informational message; no user action is required. You can review the SQL Server error log inside the container using the command:

The -h (host name) parameter as discussed above, changes the internal name of the container to a custom value. This is the name you'll see returned in the following Transact-SQL query:

Setting -h and --name to the same value is a good way to easily identify the target container.

  1. As a final step, change your SA password because the SA_PASSWORD is visible in ps -eax output and stored in the environment variable of the same name. See steps below.

Pull and run the 2019 container image

Before starting the following steps, make sure that you have selected your preferred shell (bash, PowerShell, or cmd) at the top of this article.

  1. Pull the SQL Server 2019 Linux container image from Microsoft Container Registry.

    Note

    If you are using PowerShell Core, replace the double quotes with single quotes.

    Tip

    This quickstart uses the SQL Server 2019 Docker image. If you want to run the SQL Server 2017 image, see the SQL Server 2017 version of this article.

    The previous command pulls the SQL Server 2019 container image based on Ubuntu. To instead use container images based on RedHat, see Run RHEL-based container images. To see all available images, see the mssql-server-linux Docker hub page.

    For the bash commands in this article, sudo is used. On macOS, sudo might not be required. On Linux, if you do not want to use sudo to run Docker, you can configure a docker group and add users to that group. For more information, see Post-installation steps for Linux.

  2. To run the container image with Docker, you can use the following command from a bash shell (Linux/macOS) or elevated PowerShell command prompt.

    Note

    The password should follow the SQL Server default password policy, otherwise the container can not setup SQL server and will stop working. By default, the password must be at least 8 characters long and contain characters from three of the following four sets: Uppercase letters, Lowercase letters, Base 10 digits, and Symbols. You can examine the error log by executing the docker logs command.

    By default, this creates a container with the Developer edition of SQL Server 2019.

    The following table provides a description of the parameters in the previous docker run example:

    ParameterDescription
    -e 'ACCEPT_EULA=Y'Set the ACCEPT_EULA variable to any value to confirm your acceptance of the End-User Licensing Agreement. Required setting for the SQL Server image.
    -e 'SA_PASSWORD=<[email protected]>'Specify your own strong password that is at least 8 characters and meets the SQL Server password requirements. Required setting for the SQL Server image.
    -p 1433:1433Map a TCP port on the host environment (first value) with a TCP port in the container (second value). In this example, SQL Server is listening on TCP 1433 in the container and this is exposed to the port, 1433, on the host.
    --name sql1Specify a custom name for the container rather than a randomly generated one. If you run more than one container, you cannot reuse this same name.
    -h sql1Used to explicitly set the container hostname, if you don't specify it, it defaults to the container ID which is a randomly generated system GUID.
    mcr.microsoft.com/mssql/server:2019-latestThe SQL Server 2019 Ubuntu Linux container image.
  3. To view your Docker containers, use the docker ps command.

    You should see output similar to the following screenshot:

  4. If the STATUS column shows a status of Up, then SQL Server is running in the container and listening on the port specified in the PORTS column. If the STATUS column for your SQL Server container shows Exited, see Troubleshooting SQL Server Docker containers.

The -h (host name) parameter as discussed above, changes the internal name of the container to a custom value. This changes the internal name of the container to a custom value. This is the name you'll see returned in the following Transact-SQL query:

Setting -h and --name to the same value is a good way to easily identify the target container.

  1. As a final step, change your SA password because the SA_PASSWORD is visible in ps -eax output and stored in the environment variable of the same name. See steps below.

Change the SA password

The SA account is a system administrator on the SQL Server instance that gets created during setup. After creating your SQL Server container, the SA_PASSWORD environment variable you specified is discoverable by running echo $SA_PASSWORD in the container. For security purposes, change your SA password.

  1. Choose a strong password to use for the SA user.

  2. Use docker exec to run sqlcmd to change the password using Transact-SQL. In the following example, the old and new passwords are read from user input.

  1. Choose a strong password to use for the SA user.

  2. In the following example, replace the old password, <[email protected]>, and the new password, <[email protected]>, with your own password values.

  1. Choose a strong password to use for the SA user.

  2. In the following example, replace the old password, <[email protected]>, and the new password, <[email protected]>, with your own password values.

Connect to SQL Server

The following steps use the SQL Server command-line tool, sqlcmd, inside the container to connect to SQL Server.

  1. Use the docker exec -it command to start an interactive bash shell inside your running container. In the following example sql1 is name specified by the --name parameter when you created the container.

  2. Once inside the container, connect locally with sqlcmd. Sqlcmd is not in the path by default, so you have to specify the full path.

    Tip

    You can omit the password on the command-line to be prompted to enter it.

  3. If successful, you should get to a sqlcmd command prompt: 1>.

Create and query data

The following sections walk you through using sqlcmd and Transact-SQL to create a new database, add data, and run a query.

Create a new database

The following steps create a new database named TestDB.

  1. From the sqlcmd command prompt, paste the following Transact-SQL command to create a test database:

  2. On the next line, write a query to return the name of all of the databases on your server:

  3. The previous two commands were not executed immediately. Type GO on a new line to execute the previous commands:

Insert data

Next create a new table, Inventory, and insert two new rows.

  1. From the sqlcmd command prompt, switch context to the new TestDB database:

  2. Create new table named Inventory:

  3. Insert data into the new table:

  4. Type GO to execute the previous commands:

Select data

Now, run a query to return data from the Inventory table.

  1. From the sqlcmd command prompt, enter a query that returns rows from the Inventory table where the quantity is greater than 152:

  2. Execute the command:

Exit the sqlcmd command prompt

  1. To end your sqlcmd session, type QUIT:

  2. To exit the interactive command-prompt in your container, type exit. Your container continues to run after you exit the interactive bash shell.

Connect from outside the container

You can also connect to the SQL Server instance on your Docker machine from any external Linux, Windows, or macOS tool that supports SQL connections.

The following steps use sqlcmd outside of your container to connect to SQL Server running in the container. These steps assume that you already have the SQL Server command-line tools installed outside of your container. The same principles apply when using other tools, but the process of connecting is unique to each tool.

  1. Find the IP address for the machine that hosts your container. On Linux, use ifconfig or ip addr. On Windows, use ipconfig.

  2. For this example, install the sqlcmd tool on your client machine. For more information, see Install sqlcmd on Windows or Install sqlcmd on Linux.

  3. Run sqlcmd specifying the IP address and the port mapped to port 1433 in your container. In this example, that is the same port, 1433, on the host machine. If you specified a different mapped port on the host machine, you would use it here. You will also need to open the appropriate inbound port on your firewall to allow the connection.

  4. Run Transact-SQL commands. When finished, type QUIT.

Other common tools to connect to SQL Server include:

Remove your container

If you want to remove the SQL Server container used in this tutorial, run the following commands:

Warning

Stopping and removing a container permanently deletes any SQL Server data in the container. If you need to preserve your data, create and copy a backup file out of the container or use a container data persistence technique.

Docker demo

After you have tried using the SQL Server container image for Docker, you might want to know how Docker is used to improve development and testing. The following video shows how Docker can be used in a continuous integration and deployment scenario.

Next steps

For a tutorial on how to restore database backup files into a container, see Restore a SQL Server database in a Linux Docker container. Explore other scenarios, such as running multiple containers, data persistence, and troubleshooting.

Also, check out the mssql-docker GitHub repository for resources, feedback, and known issues.

Initially, the inception of Docker containerization started out with Linux as its base platform. However, over the years, Docker and Microsoft have continuously grown their partnership, creating a conveniently consistent interface for building, shipping, and running applications without the usual dependence hurdles associated with virtual machines.

Though a huge number of enterprises are already using Docker on Windows platforms, there has been a number of subtle functionality disparities between Windows and Linux containers. However, Windows Server 2019 (1809 build) has successfully addressed most of the inconsistencies between Docker containers in Linux and Windows environments.

Requirements for Installation of Docker on Windows

Docker containers are powered by a Docker engine. Though initially designed for Linux, extensive work has been done to allow Docker containers to run on Windows and macOS environments.

To run Docker containers on a Windows platform, one prerequisite is the installation of a Windows server. You can do this in a physical server machine, on a cloud environment running in Azure, or an on-premise virtual machine.

Install the Hyper-V feature on your Windows server 2019

There are two distinct modes to run Decker containers on Windows platforms: Process isolation and Hyper-V isolation. With the Process isolation mode, the Docker containers share the OS kernel with the host platform, hence they are lightweight and identical to Linux system Docker containers.

On the other hand, the running of Docker containers in the Hyper-V mode is confined to a special nominal virtual machine. This enables improved compatibility and secure kernel-level. To run Docker containers in this mode, you must first enable Hyper-V in the host operating system.

The default operation mode for Docker installation on a Windows server is the operation mode (enabling Hyper-V is optional). However, it’s a prerequisite to enable the Hyper-V isolation mode if you need to run Linux containers on a Windows Server interface.

The OS build is another crucial determinant on the need for Hyper-V mode as Windows containers should be of the same build version as the container host OS’s version. Still, Windows container images with a lower build version than the container host OS can run with Hyper-V isolation.

To install Hyper-V on Windows Server 2019, run the PowerShell as Administrator and run the commands below:

Enable-WindowsOptionalFeature –Online -FeatureName Microsoft-Hyper-V –All -NoRestart

Install-WindowsFeature RSAT-Hyper-V-Tools -IncludeAllSubFeature

Next, restart your Windows Server VM.

Prerequisites for the container host

Install Container

Yum Install Container-selinux

You must enable virtualization in the hosting Windows server platform to utilize Hyper-V isolation in your containers: enable hardware virtualization for a container host running on hardware and nested virtualization in the base interface for a container host running on a cloud space or Hyper-V.

Running Docker Containers on Windows Server 2019

Before running multiple isolated applications using Windows Containers, you need to activate (enable) the containers feature and install Docker on your Windows Server 2019. Here’s the process:

  1. Enable the containers feature in Windows Server 2019.

Run PowerShell as an Administrator and run this command:

Install-Module -Name DockerMsftProvider -Repository PSGallery -Force

This command will install the Docker-Microsoft Package Management Provider from the PowerShell Gallery.

When prompted to install and import NuGet provider, type Y and hit ENTER

  1. Install Docker on your Windows Server 2019

After installing the Containers feature on Windows Server 2019, it’s time to install the latest versions of Docker Engine and Docker Client. Run this command in your PowerShell session:

Install-Package -Name docker -ProviderName DockerMsftProvider

Accept the installation by selecting “Yes”, “Y” or “A” to Agree to all the installation requests.

After the completion of this installation, reboot your computer.

Restart-Computer –Force

You can check your installed Docker version via the PowerShell command:

Get-Package -Name Docker -ProviderName DockerMsftProvider

Install Containers Hyper V


You can also confirm the installed Docker version using the docker –version command:

docker –version

You can opt to upgrade anytime by running the commands below on PowerShell:

Install-Package -Name Docker -ProviderName DockerMsftProvider -Update -Force

Then start the docker service.

Start-Service Docker

  1. Launch (Run) Docker Containers on Windows Server 2019

Run the following commands on PowerShell:

Start-Service Docker

After starting the Docker Engine service, proceed to download the pre-created .NET sample image on the Docker Hub registry:

docker pull microsoft/dotnet-samples:dotnetapp-nanoserver-1809

After the download, you can deploy a simple Docker container that runs the .Net ‘Hello World’ application:

docker run microsoft/dotnet-samples:dotnetapp-nanoserver-1809

After running the command, an ASCII image will be printed to the shell accompanied by the “Hello” message.

Running Linux Containers on your Window Server 2019

By default, Docker on Windows only runs Windows containers. To launch Linux containers on Windows Server, use the Docker Enterprise Edition Preview that comes with a full LinuxKit system to run Docker Linux containers.

  1. First, uninstall the already installed Docker CE.

Uninstall-Package -Name docker -ProviderName DockerMSFTProvider

  1. Enable Nested Virtualization in case you’re running Docker Containers on a Linux Virtual Machine running on Hyper-V.

Get-VM WinContainerHost Set-VMProcessor -ExposeVirtualizationExtensions $true

NOTE:WinContainerHost is the name of your virtual machine

  1. Install the Module Docker Provider

Install-Module DockerProvider

Install-Package Docker -ProviderName DockerProvider -RequiredVersion preview

A restart will be required after this operation

  1. Enable LinuxKit system to run Linux containers

[Environment]::SetEnvironmentVariable(“LCOW_SUPPORTED”, “1”, “Machine”)

  1. Restart the Docker Service after the change above and restart the Service Docker

Restart-Service docker

To switch back to running Windows containers, execute the following command in PowerShell:

Install Container

[Environment]::SetEnvironmentVariable(“LCOW_SUPPORTED”, “$null”, “Machine”)

You have finally installed and configured Docker your Windows Server machine to run both Linux and Windows containers. We hope this guide was insightful.

Most Viewed Posts