Hi,
I have this problem where Lets Encrypt will not autoupdate properly.
I think the problem might be that I am deploying multiple projects on the same droplet. Every project has its own trellis-instance, site- and trellis-folder are in the same github repo. I think I remember @ben mentioning in an other threat, that this is not the way trellis was meant to be used?
From looking at the cronjob files, I had the idea that maybe every time I provision from a new trellis instance/project, this project will overwrite the cronjob-files and only the cronjobs of the last provisioned project will work, the others will not be triggered anymore.
Am I right with this and what would be a good way to fix this? Maybe have only one trellis-folder in an own github-repo and multiple site-folders in seperate github-repos?
As Iâm hosting a lot of very small sites it would be unnecessarily expensive to deploy each on an own droplet, so I would like to avoid that if possible.
My fix at the moment is uptime-monitoring and manually reprovisioning - which obviously is a sh**y solution.
I believe that might be correct. The Lets Encrypt cron job simply runs the generated ./renew-certs.py script. But that script file is âglobalâ since itâs only generated once per project to the same file path meaning it should overwrite any existing ones.
Is this the only problem youâve run into with this multi-trellis setup on a single server?
And yes Iâm running this setup since over a year and its working great except this one aspect - canât say for sure tho there arent any other problems I havenât noticed yet.
Any ideas how I could change adapt the lets encrypt regeneration process in order to make it work?
My best guess for a suitable identifier would be using the wordpress site name. Iâm guessing you only have 1 per project? So you could use {{ wordpress_sites.keys() | first }}: