Install sensenet on local Docker

Github logo Edit on GitHub

This article describes how to install sensenet in a docker environment on your local machine. It is a Powershell script that will install sensenet from Docker images. It will create a Docker network and start the necessary containers.

Prerequisites

Before you begin, please take a look at the prerequisites.

The scripts

You'll find the necessary scripts in the deployment folder on GitHub. You can download the contents of the folder and run the script from your local machine. You do not need the whole sensenet repository to run the script.

Docker

The installer uses the docker cli and works with docker containers. You have to have docker installed on your machine.

By default the installer uses the publicly available sensenet Docker images.

Powershell

As this installer is built of Powershell scripts you have to have Powershell installed on your machine.

The installer is built of multiple scripts. By default Windows will ask confirmation before running every script, so it is recommended to set the ExecutionPolicy for the current process:

Set-ExecutionPolicy -Scope Process -ExecutionPolicy Bypass

This change on process scope will only affect the current powershell session. To learn more see the documentation on execution policy.

dotnet cli OR a valid certificate

Sensenet services will use the snapp.pfx certificate file from under ./temp/certificates folder. If it is not present the installer will create a developer certificate in this folder and it uses dotnet cli for this. On linux the dotnet --trust switch won't work. so you have to either create or trust the created certificate manually.

git cli (in case of source install)

By default the installer uses the publicly available sensenet Docker images. In case you want to create the Docker images from source code, you will have to have the git cli installed. The CreateImages switch will download the necessary git repositories in order to create Docker images from sensenet service solutions on-the-fly and it uses the git cli for this.

Simple install

The script is preconfigured so you don't have to know every switch if you just want to try it out. With the preconfigured settings the installer will create an IdentityServer, an MS SQL Server and a sensenet application container. Neither will have mounted volumes so the necessary files will be either copied into the containers or created inside them.

To install sensenet, execute the following script:

.\install-sensenet.ps1

After the script completes, your repository will be accessible on the URL displayed by the script. In case of the default installation this is https://localhost:51016 but it may be different in case of in-memory or search service scenarios.

You can log in by visiting https://admin.sensenet.com and providing the repository url. The default credentials are:

username: admin
password: admin

Using the -OpenInBrowser switch will automatically open the admin ui in your default browser.

Parameters

You can customize the installation with the following parameters:

InMemPlatform

By default the installer will create a sensenet repository that uses an MS SQL database, but there is an in memory sensenet repository that does not need a physical database. This version will start an identity server and a sensenet application container only.

.\install-sensenet.ps1 -InMemPlatform

CreateImages

The -CreateImages switch will download the source code of the necessary repositories from GitHub and create the appropriate Docker images for sensenet services. For example if you executed the installer with the -SearchService switch, it will create the appropriate sensenet application Docker image and the search service image as well, and so on. These temporary folders will be created under the ./temp folder and the images will be created on the host Docker registry.

.\install-sensenet.ps1 -CreateImages

SearchService

Sample sensenet set for NLB environments. Indexing is not handled by the sensenet application but by a separate search service. It will start the following containers:

  • IdentityServer
  • MS SQL Server
  • Search service
  • sensenet application
  • RabbitMQ service for messaging

HostDb with DataSource, SqlUser and SqlPsw

You can try this sensenet demo with your local database on the host machine. The -HostDb switch will initiate this mode. However you will need a SQL user on your MS SQL Server at least with database creation permission. The script will automatically use the host name as datasource. If your MS SQL Server name is different from your host name, you will have to set it with the -DataSource parameter. You can also set the -SqlUser and -SqlPsw parameters when running the installer otherwise it will ask for these at run time. For example:

.\install-sensenet.ps1 -DataSource MyMachineHostName\Sql2019 -HostDb -SqlUser testuserfordockerdemo -SqlPsw Ultr4Secur3P4ssw0rd

UseVolume and VolumeBasePath

This is a bit advanced switch as Docker for Windows with linux containers may have a problem with the default setup. By default containers will be created without volume bind, so it should work on every system. But it is recommended to use volumes with an actual sensenet project in containers. This switch shows how to bind those volumes for sensenet services. However with default settings it may only work on a linux host. If you use Docker for Windows you may have to set the folder path to bind and this path should be on linux. It is because Windows and linux file systems handle file locks differently and the Lucene engine from the linux containers will not be able to lock the necessary files on a Windows file system.

While on linux it is enough to call 

.\install-sensenet.ps1 -UseVolume

on Windows it should be something like this:

.\install-sensenet.ps1 -UseVolume -VolumeBasePath /var/lib/docker/volumes

The example above will bind the WSL paths with the containers.

OpenInBrowser

With this switch the installer will open the admin ui when the repository is created.

Uninstall

This switch is responsible for cleanup after an installation. You should use the same switches as with the install to remove the same resources.

DryRun

This switch can be used to see what processes would be executed but without actually running them.

Verbose

The output is reduced by default to decrease the amount of install information. With the -Verbose switch all the additional technical information will be shown. For example the actual Docker command is shown that can be useful if you need to customize the installation.