Open gfodor opened 4 years ago
https://github.com/mozilla/hubs/discussions/2746
Another user confirmed this as an issue for DO deploys!
Also happens when enable/disable cloudflare workers, when there is automatic hubs-cloud updates being pushed (at least it happens around the time when Mozilla posts new hubs-cloud branch updates in #hubs-cloud channel) . I think it also happens on server restarts but havent confirmed it.
I'm using Basic / 2 GB / 1 vCPU (shared), with a managed database as recommended with DO documentation. Sometimes the server goes nuts (probably due to memory overloaded, it's just a cheap test project) and this will also reset(restart?) the server (I get "can't find hub.html" reticulum error for a while then it goes back to the non-custom-client interface (settings from admin panel still intact, just the custom client changes are gone)).
Can also confirm. This also happens if you reboot the server.
User reported their custom client was blown away on DigitalOcean when we did a hubs client push. Confirmed that their runhook still was doing a copy.
┆Issue is synchronized with this Jira Task