Part 3: Creating Backups and Sharing DataLad Datasets

Creating a Backup

Command	Description
`git init --bare ~/mydir`	Create a `--bare` repository called `mydir` in the home directory (on Linux/macOS)
`git init --bare %USERPROFILE%\mydir`	Create a `--bare` repository called `mydir` in the home directory (on Windows / CMD)
`git init --bare "$env:USERPROFILE\mydir"`	Create a `--bare` repository called `mydir` in the home directory (on Windows / PowerShell)
`datalad siblings`	List all siblings of the current dataset
`datalad sibings add --name new --url ~/mydir`	Add the repository at `~/mydir` as a new sibling with the name `new`
`datalad push --to new`	Push the dataset content to the sibling named `new`

Exercise 1 List all siblings of the current dataset.

Solution

datalad siblings

Exercise 2 Initialize a --bare git repository at a path outside of this dataset.

Solution

On Linux/macOS

git init --bare ~/penguins_backup

On Windows

git init --bare %USERPROFILE%\penguins_backup

Exercise 3 add a new sibling to the dataset using the path to the newly created git repository as the --url. Then, list all siblings to confirm it was added.

Solution

On Linux/macOS

datalad siblings add --name backup --url ~/penguins_backup
datalad siblings

On Windows

datalad siblings add --name backup --url %USERPROFILE%\penguins_backup
datalad siblings

Exercise 4 Push the dataset to the new sibling twice.

Solution

We need to push tiwce because the first push initializes the repository’s annex ID and the second (and each subsequent) push actually tranfer the annexed files.

datalad push --to backup
datalad push --to backup

Exercise 5 Move to a directory outside of this dataset and clone the new sibling dataset.

Solution

On Linux/macOS

cd ..
datalad clone ~/penguins_backup

On Windows

datalad clone %USERPROFILE%\penguins_backup

BONUS: Sharing your Dataset online

Command	Description
`ssh-keygen`	Generate a public and private authentication key pair
`datalad siblings`	List all siblings of the current dataset
`datalad sibings add --name gin --url git@gin.g-node.org:/user/repo.git`	Add the gin repository at `/https://gin.g-node.org/user/repo` as a new sibling with the name `gin`
`datalad push --to gin`	Push the dataset content to the sibling named `gin`

Exercise 6 Use ssh-keygen to generate a public and private key pair (you don’t have to use a passphrase). Note the location where the public key is stored, e.g. .ssh/id_ed25519.pub. Open the .pub file and copy the whole content — it should look something like this: ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIBOYcoRKZZLWA4FWECpW2K/fTOvuRYXBnBA6gcea2bFq <user>@<computer>

Solution

ssh-keygen

Exercise 7 Login in to your GIN account, go to your user settings and add the copied ssh key. Now datalad should be able to connect to your GIN account!

Exercise 8 Create a new repository on GIN, make sure to NOT initialize it with a README.

Exercise 9 add a new sibling to the dataset using the --url of the newly created gin repository and confirm the connection. Then, list all siblings to confirm it was added.

Solution

For the repository in the image above, the command would look like this:

datalad siblings add --name gin --url git@gin.g-node.org:/adswa/DataLad-101.git

Exercise 10 Push the dataset to the new GIN sibling. Then, open the repository in your browser to confirm the content was pushed.

Solution

datalad push --to gin

Exercise 11 Move to a directory outside of this dataset and clone the new GIN sibling.

Solution

For the repository in the image above, the command would look like this:

cd ..
datalad clone datalad clone https://gin.g-node.org/adswa/DataLad-101

Creating a Backup

BONUS: Sharing your Dataset online

Further reading