Well, it depends how many users will be using it, and things like that. I have a testing instance over at https://metaman.space which i use as my primary client, it has 2vcpu/2gb of ram/60gb of disk and that is more than enough. Our bigger instances are 2 or 4 vcpu with 4 or 8gb of ram, gleasonator.dev is 4 vcpu 8gb of ram, and this has maaaany many users.

metaman.space runs its own Postgres, the rest all share a single postgres with 320gb of disk, 8 cores, and 16gb of RAM.

I wouldn't worry too much about CPU and RAM honestly, ditto itself is pretty lightweight. Try starting with a single vCPU and 1gb of ram + 1.5gb of swap and seeing how that goes. You can always trivially scale up your VPS if it feels slow.

In my opinion, the bigger scalability problem is actually disk space -- the gleasonator database is like 30gb at the moment. metaman, a much smaller instance, is using about 10gb. So I'd keep an eye on your usage.

Feel free to @ me if you have any more questions!

Reply to this note

Please Login to reply.

Discussion

Thank you so much.

I’m troubleshooting a 502 error.

So far, this is what I’ve confirmed is working:

- nginx is running

- the config file syntax is ok

- a test of the config file was successful

- I don’t think I have conflicts on either port (80,443)

- the DNS is pointing to the right IP

- I’ve double checked the spelling of the domain that replaced the example.com

Any other suggestions?

Is there something in the ditto setup that is most likely to trip a newbie?

I haven’t discovered the issue yet, but I’ll walk through my screenshots to see if I spot something.

Just confirming, but is Ditto itself running? That's what a 502 Bad Gateway typically indicates, that the service being proxied to isn't active, and your triage list didn't actually mention that. Please don't mistake me, just eliminating the obvious.

Next up, if you have some sort of firewall set up it may be blocking nginx's ability to communicate with Ditto.

Also, if you've configured ditto to listen on a different port other than 4036, please check that your nginx config is updated to reflect that.

There have been some issues with nginx configs in the past but as of latest ditto, nothing. Could you also confirm if you're on latest `main`?

Thank you for your reply. I’ve been digging deeper into this rabbit hole. So much to learn…

I’ll try to run it again today. I think it’s an issue with my config.

If I have the VPS not to use root, for improved security, instead using a different user, I must update it with the other user, right?

Firewall’s not blocking the port.

I’m following the docs and it pulls from main.