Skip to content

Latest commit

 

History

History
52 lines (39 loc) · 2.88 KB

README.md

File metadata and controls

52 lines (39 loc) · 2.88 KB
Raw

Github2Synology

A simple Ash (practically "BusyBox bash") script designed to run on the Synology DS range of file storage servers to backup all repositories (and wikis) for a user from Github.

Running

  • Ensure you have git installed on the Synology - this can be download from the SynoCommunity. The script also needs cUrl and jq but these seem standard on Synologys.

  • Now login to Github and go to https://github.com/settings/tokens and create a personal access token with the following scopes:

    • repo (repo itself including all subs) - Full control of private repositories
    • admin:org read:org - Read org and team membership

  • Add this token as the OAUTH_TOKEN in line 7 on the github2synology.sh script. (OAUTH_TOKEN="[PUT YOUR TOKEN HERE BETWEEN THE QUOTES]")

  • Ensure the backup path is correct on line 9. (BACKUP_PATH="/volume1/serverBackups/github/backup")

  • Copy the script over to your Synology and run it (all via SSH)

Problems?

Getting "Access forbidden/Repository not found" issues?

This is because the Synology doesn't have access to your Github repositories. The "best (most secure)" way to resolve this is just to enable "SSH Key forwarding" from your Mac/PC to the Synology so it uses your SSH keys for authentication.

On Mac and Linux command line, you should be able to just create/edit ~/.ssh/config and add:

Host [synology]
   ForwardAgent yes

(replacing synology with the IP/name of your Synology)

In PuTTy on Windows, this is under Connection->SSH->Auth->Authentication parameters : Allow Agent Forwarding (ensure Pagaent is running with a Github recognised key).

To test if this is setup correctly, try running from the Synology: ssh -T [email protected] you should get back: Hi xxxxx! You've successfully authenticated, but GitHub does not provide shell access.

"It's only backing up 100 repositories, I've access to more"

Due to the script's simplicity, it does NOT currently read the Github provided Link: Http headers which give the next page details.

To work around this, run the command curl -I "https://api.github.com/user/repos?type=all&page=1&per_page=100" -H "Authorization: token [OAUTHTOKEN]" (replacing [OAUTHTOKEN] with your token). You'll then see a line such as: Link: <https://api.github.com/user/repos?type=all&per_page=100&page=2>; rel="next", <https://api.github.com/user/repos?type=all&per_page=100&page=3>; rel="last"

Then just add the Link:...rel="next" entry to the bottom of the script such as:

API_URL="https://api.github.com/user/repos?type=all&per_page=100&page=2"

fetch_fromUrl

and the last and next do not match, just repeat these steps changing the &page=1 increment in the curl command.