This is specifically a problem for us when trying to clone our decaturmakers.org repo. It happens because the vagrant provisioning process runs as root and doesn't have access to a forwarded SSH agent (During vagrant ssh, SSH agent forwarding does work, but that doesn't help us here). I have two solutions to the problem and thought this would be a good place to discuss, debate and decide on which solution we'd like to use.
- Part of the initial setup would require you to create a Personal access token on github.com. We'd then have the
vvv-init.sh script clone decaturmakers.org via HTTPS (using the personal access token) for authentication.
| Pros |
Cons |
Everything remains encapsulated in this single vvv-init.sh script |
There is an extra initial step of having to generate a personal access token |
- We use a
Makefile to clone the decaturmakers.org repo on the host machine and then run the initial vagrant up. In this scenario the vvv-init.sh script would basically just do the automated WP installs using the wp-cli.
| Pros |
Cons |
Ideally, the initialization of the dev enviroment is a single command: make |
Can't really thing of any cons |
This is specifically a problem for us when trying to clone our decaturmakers.org repo. It happens because the vagrant provisioning process runs as root and doesn't have access to a forwarded SSH agent (During
vagrant ssh, SSH agent forwarding does work, but that doesn't help us here). I have two solutions to the problem and thought this would be a good place to discuss, debate and decide on which solution we'd like to use.vvv-init.shscript clonedecaturmakers.orgvia HTTPS (using the personal access token) for authentication.vvv-init.shscriptMakefileto clone thedecaturmakers.orgrepo on the host machine and then run the initialvagrant up. In this scenario thevvv-init.shscript would basically just do the automated WP installs using thewp-cli.make