SSH Error: data could not be sent to remote host "35.186.171.128". Make sure
this host can be reached over ssh
fatal: [kinsta_staging]: UNREACHABLE! => {"changed": false, "unreachable": true}
to retry, use: --limit @/Users/afsp/OneDrive - American Foundation of Suicide Prevention/Tutorials/trellis/colorfulpuddle.com/trellis/deploy.retry
PLAY RECAP ************************************************************************
kinsta_staging : ok=1 changed=0 unreachable=1 failed=0
This is definitely a duplicate of Kinsta deploy - data could not be sent but since that thread is rather old, I thought it might be okay to open a new one. Also, I’m having no issue ssh-ing into my account, unlike the OP of that thread, so I’m not sure what additional steps I can take to fix it.
One more question: is this something that the folks at Kinsta could help with? Since they’re sponsors of Roots, I imagine they’re fairly familiar with the setup, but I don’t know how much their support people will know about Trellis specifically. Thanks!
Are you deploying from Windows? If so, your deploy happens from inside the Trellis VM, so you probably need to set up key forwarding to forward your local key to the VM to use for deployment.
Yeah. In order for Trellis deploys to work with Kinsta you’ll need to make sure that your personal SSH key has been added to your profile in the My Kinsta dashboard as well.
You can confirm whether or not your SSH key was added by trying to connect without a password:
$ ssh username@KINSTA_IP -p PORT
And if that does work but Trellis deploys don’t, it’s likely an SSH forwarding issue.
Kinsta support wouldn’t cover a custom Trellis/Bedrock setup unfortunately, but hopefully we can help you out over here
I’ve done a few paid migrations onto Kinsta lately which also includes my sync script for moving the database and uploads between all possible environments (development, staging, and production - with horizontal syncing supported), and a CircleCI config for deploying the site whenever you push to master or staging) Let me know if you wanna chat more about that.
Thanks Ben. It actually wound up being an oversight on my part in setting up main.yml. I’d forgotten to change the web_user to reflect the user in Kinsta. Once I made that fix, deploy went well. Thank you and @MWDelaney for your help.
Let me know if you wanna chat more about that.