Hi! not sure if this is the correct forum to reach out to, however, here i am anyways.
I currently run a lemmy server that has been going for a few years now, and i was recently informed by our host that they were shutting down. I was wondering if it would be possible to migrate the instance to your services, as well as getting an estimated cost based on our current server configuration vs your services; I know you have listed out lemmy plans, but those don’t seem to go in to too much detail as to what is included.
So, it’s a bit tricky to tell you exactly “what is included” because I don’t have isolated VMs for each deployment. They run in a cluster and then each deployment gets just set up with some “soft-limit” for CPU/RAM/S3 storage. The only place where it’s isolated per deployment is the database volume.
But to give you a rough idea, these are the current limits for Lemmy:
Small Plan: Database with 512MB RAM and 30GB disk, 1.5GB RAM for the application, 1vCPU, 100GB storage bucket
Medium Plan: Database with 1GB RAM, 50GB disk, 3GB for application, 2vCPU, 250GB storage bucket
To be completely honest though, all Lemmy instances that I had running right now are “small” and not from heavy users. I am listing the medium plan for “around 100 users” but depending on the usage patterns your instance might need more resources than that.
We could definitely work out something for migrating your instance. If you commit to an 1-year plan I can waive any setup fee. You’d need to send me a copy of your Lemmy database, the pict-rs database. You should be able to use tools like rclone to copy the data from your current server to the new one.
Hi, thanks for the quick reply; As of right now, our instance currently uses 800gb of disk space. I think this is everything from lemmy install, DB, backups, and images, etc.
I’m going to try to get an exact breakdown of all pieces of that pie.
The instance currently in question is ttrpg.network. We have about ~150ish users per month on average of around 8000 registered, which i think puts us on the larger side of lemmy instances
Would you mind telling me what was the current plan you had on k&t? I can offer more space for relatively cheap, the limiting factor is the amount of RAM and vCPUs.
Hi, so i got an answer. We apparently have 3 vCPUs on our plan. I was also wrong in that the postgres database had a different quota than the initial 800GB disk space; It looks like our postgres DB is currently sitting around 100 GB
Hi, Yeah, i think that sounds like it should work. I think we might still need a bit more storage space for what we have currently (in addition to the storage is probably likely to grow over time), but i think the listed specs should be fine otherwise; I’m not 100% sure how much each portion of our install currently takes up, since from my perspective it’s all lumped in together, But i’d suspect the images are the bulk of it.
Should i just directly purchase the default medium plan after all the updates and everything?
Either way, i think that should generally be fine, but i assume we’ll need to work out the logistics of actually migrating over, right? I havent started downloading any of our data/backups yet, since im waiting until i have more concrete plans right now.
Hi, are the plans updated? I think it is because it looks like the price changed. Should i sign up using that page? And about migration, what would you like the next steps to be?
Hey! Sorry, I think I think I missed the notification from this last message.
I did update the prices, so yes, please do sign up to it and we can start working on the migration. I do think we need to come up with a comfortable time window, because this is not an automated process, so I guess it will take me about 4-6 hours to ensure that everything is in order:
You’d need to:
take your server offline,
get a backup of your database (both the postgresql dump and pict-rs)
send me a copy of all your media storage folder (or make it available at a place where I can make a copy)
update your DNS server to point to gateway.communick.host
It would be better/easier if we can schedule this for mid-December. Would that work for you?
