Adding Docker to an ASP.NET Core Application

I have a series of post about deploying an ASP.NET Core application to different cloud providers which centered around clouds that provided a service that supported ASP.NET Core. The following are all the post in that vein if you are interested.

Google Cloud Platform
Amazon Web Services
Microsoft Azure

I was asked if I was going to do a post on deployment to DigitalOcean. While not in my original plan it was a good idea. The thing is DigitalOcean doesn’t have a service to support ASP.NET Core, but they do support Docker so this post is going to cover adding Docker support to a sample ASP.NET Core application. Look for a future post on how to deploy to DigitalOcean.

Docker

What is Docker? To quote Wikipedia:

Docker is a computer program that performs operating-system-level virtualization also known as containerization. It was first released in 2013 and is developed by Docker, Inc.

Docker is used to run software packages called “containers”. In a typical example use case, one container runs a web server and web application, while a second container runs a database server that is used by the web application. Containers are isolated from each other and use their own set of tools and libraries; they can communicate through well-defined channels. All containers use the same kernel and are therefore more lightweight than virtual machines. Containers are created from “images” which specify their precise contents. Images are often created by combining and modifying standard images downloaded from repositories.

Installing Docker

Head to this link and click the Get Docker link to download the installer. After the install is complete you will have to log out and back in. When I logged back in I got a message about Hyper-V not being enabled.

After logging back in I then got the following message about hardware-assisted virtualization not being enabled.

After tweaking my BIOS settings and logging back in I was greeted by the Docker welcome screen.

Open a command prompt and run the following command.

You should output that starts with the following if your installation is working.

Hello from Docker!
This message shows that your installation appears to be working correctly.

Sample Application

This is the same as in the cloud series, but am including it in case you missed that post. The sample application that we will be deploying is the basic Razor Pages applications created using the .NET CLI. The following commands are what I used to create the application, create a solution, and add the project to the solution. I ran all these commands in a  CloudSample directory.

Add Docker Support

Open the solution created above in Visual Studio and right-click the project and select Add > Docker Support.

Next, you will have the select the OS you want to target.

This will add the following four files to your project.

The process will also change your project to start in Docker when you hit run from in Visual Studio.

Publishing

In order to publish we need an account with a registry that we want to publish to. There are multiple options, but I am going to be using Docker Hub. Go to Docker Hub and sign up for an account. After you get signed in on your Repositories tab click Create Repository.

On the next page enter a name for your repository, select if you want it to be private or public and click Create. I’m using the name TestRepository and making it private.

Now that we have a Docker Hub account hope back to Visual Studio and right-click on your project and select Publish.

On the next on the dialog select the Container Registry option and then Docker Hub. Then click the Publish button.

On the next dialog enter your username and password for Docker Hub and click Save.

After saving Visual Studio will build and publish the application to Docker Hub.

Issues

If you get luck the above will work for you. I hit a couple of issues that made the above fail for some reason. The following is the first error I got.

Error: Service ‘cloudsample’ failed to build: The command ‘/bin/sh -c dotnet restore /CloudSample.csproj’ returned a non-zero code: 1

After some digging, I found this GitHub issue. The gist of the fix is to change the following in the DockerFile.

The above fixed the build when in release mode, but publish still failed on me with the following error when the tools were attempting to tag the image.

System.Exception: Running the docker.exe tag command failed.
at Microsoft.VisualStudio.Web.Azure.Publish.ContainerRegistryProfileVisual.DockerTag(ProcessStartInfo startInfo, String serverUrlString)
at Microsoft.VisualStudio.Web.Azure.Publish.ContainerRegistryProfileVisual.<PostPublishAsync>d__24.MoveNext()
— End of stack trace from previous location where exception was thrown —
at System.Runtime.CompilerServices.TaskAwaiter.ThrowForNonSuccess(Task task)
at System.Runtime.CompilerServices.TaskAwaiter.HandleNonSuccessAndDebuggerNotification(Task task)
at Microsoft.VisualStudio.ApplicationCapabilities.Publish.ViewModel.ProfileSelectorViewModel.<RunPublishTaskAsync>d__116.MoveNext()

I haven’t found a way to get around this issue. I ended up using the Docker CLI to push my image. This takes a few steps. First, use the following command to log in.

Next, use the  docker images command to find the image you are trying to push. In my case, it was the image with ID f83e5adab340. When the ID we can now use the following command to tag the image.

It is important that the image is tagged with the name of the repository it is going to be pushed to. Finally, run the following command to push the image with the specified tag into your repository.

Wrapping Up

The teams at Microsoft have made adding Docker support to an existing project super simple. Docker is something that I have had a surface level awareness of for a while, and I am looking forward to trying an actual deployment to see how that works with DigitalOcean.

Setup PostgreSQL on Windows with Docker

Over the weekend I finally got the chance to start reading A Curious Moon by Rob Conery which is a book on learning PostgreSQL by following the fictional Dee Yan as she is thrown into database administrator role at an aerospace startup.

I have a lot of experience using Microsoft’s SQL Server, but up until now, I haven’t touched PostgreSQL. For personal projects SQL Server’s cost and be prohibitive and the release of Rob’s book added up to a good time to give PostgreSQL a try.

Install Directly or not?

On the download section of the official Postgres site, there is an option to download an installer. This is the route I was going to at first, but in Rob’s book, he suggests using a VM for Postgres installation on Windows. This kicked off a lot of searching on my part and didn’t find a good definitive answer on why that is or isn’t the way to do.

In the end, I decided to try and run the Postgres process using Docker instead installing directly on Windows or dealing with a full VM.

Installing Docker

Head to this link and click the Get Docker link to download the installer. After the install is complete you will have to log out and back in. When I logged back in I got a message about Hyper-V not being enabled.

After logging back in I then got the following message about hardware-assisted virtualization not being enabled.

After tweaking my BIOS settings and logging back in I was greeted by the Docker welcome screen.

Open a command prompt and run the following command.

You should output that starts with the following if your installation is working.

Hello from Docker!
This message shows that your installation appears to be working correctly.

What about Postgres?

Getting up and going with a container running Postgres was pretty simple and could be done with the following command which will create a container and expose the port used by Postgres so it can be accessed from the host.

The problem with this approach is if you ever need to rebuild the container for some reason, like a new version of Postgres is released, your data will be lost. Thankfully I found this blog post which shows how to use a secondary container for the data leaving the Postgres container able to be destroyed and recreated as needed. The following is the command I used to create my data container.

The above creates a container named PostgresData based on the Alpine image. It is important that the  -v parameter matches the path that Postgres expects.

Now that we have a container that will keep our data safe let’s create the actual Postgres container with the following command.

The only difference from the first example run command is the addition of  --volumes-from PostgresData which tells the container to use the PostgresData container.

If you run the  docker ps -a command it will show you all your containers.

As you can see in my example I have two containers only one of which is actually running. Make sure you don’t remove the data container just because it will never show as running.

Connect to Postgres

To verify all was working I downloaded pgAdmin from here. Run the installer and then open the application. Right-click on Server and click Create > Server.

On the Create Server dialog enter a Name for your server and then switch over to the Connection tab.

On the Connection tab for Host use localhost and in the Password field use the password you used for POSTGRES_PASSWORD on the docker run command.

Click Save to close the dialog and connect to the server. The following is an example screenshot of what you will see showing the available databases on the server, which is just the default database in this case.

Wrapping Up

Make sure to check out the official docs here for more information as needed.

Other than the storage portion getting Postgres up and running in Docker was pretty simple. I hope like me this will give you a good jumping off point to learn more about both Docker and Postgres.

If anyone has any alternate ways to deal with persistent storage please leave a comment.