I have been observing a Discourse forum with a huge amount of traffic, which is causing slowness. There is a tradition, or so they say, of resetting the forum and starting over. The reasoning is that as the forum grows in data and traffic, it requires more resources like RAM, CPU, and disk, significantly increasing costs.
Is there any basis for this? can anyone with high-traffic experience explain whether high traffic really requires progressively more powerful servers, even with caching and other optimizations?
Context about their server:
Because the current problem isn’t infrastructure (CPU/RAM/storage), but revenue.
The last upgrade was expensive and we can’t afford it.
The reset is necessary to get a cheaper “droplet” (with less storage, since all the content will be deleted).
Throwing away your community’s history doesn’t really make sense unless unless whatever people said last year doesn’t matter at all. Slack deletes stuff that’s some months old by default, I think.
How big is your database? Is your cost issue about Disk space? There are cheaper ways to get more of it (e.g., pushing images to S3). If the database is very big, then indeed, you can need more RAM to support it.
Are you having performance issues? Is the forum slow?
That’s my point this instance not use proxy or caching maybe are using an droplet /cloud priced in 9€ daily so I wonder myself if the fact them are using cloud and not a VPS KVM or Bare Metal is the reason to many outages these days
No! I’m watching outages of a instance what I am a old member and yes the content it’s important but I don’t get how a instance that not allowed upload is suffer many outages for a huge data.
If you don’t have any data about the server ram, cpu, and size of the database it’s silly to speculate what the problem could be. It could be a configuration issue.
If you mean they are using 2gb server for 9 euro a month with 2gb of ram and the forum has been around a long time then it’s not hard to imagine that they need to double the server size.