Ok it looks like we are back up!
1 Like
Its back? Yes!!!
1 Like
Doesnāt cloudflare have something called Always Online that loads cached webpages when the server is down? I think that might be for paying users only
Yep. I just got an email too.
1 Like
āā¦because the server stopped respondingā. Are we having so much users on the support because of this that we are causing an unintentional ddos attack. Discourse does have to ping the server like every 5 seconds to download new post updates.
1 Like
Oh greatā¦ Had downtime yeserday too
Iām pretty sure support.glitch.com uses discourse hosting. Check the links to images and you will notice they are hosted on aws.discourse-cdn.com
1 Like
Website isāt working
The community site uses Cloudflare (AWS).
The software and the hosting plan provided by discourse, yes. I donāt think they self hostā¦
1 Like
They need to do something similar to how Discord Bots do sharding. So if a project container goes out, not all projects go offline - especially the paying usersā projects.
That would be a good idea, however that would probably make the amazon bill go through the roof. They would need to use at least twice as many VMs.
They could always select like a group of 10 to 50 containers to test deploys on before deploying on all other containers so we can limit problems.
3 Likes
True, but I feel like that would be the only good solution. /shrug
Iām also not really sure why the outages are happeningā¦
It really shouldnāt be an upstream service because amazon has 99.99% uptime. Maybe its a code error internally? Or maybe they are matience.
1 Like
Why does that even happen?
1 Like