If you’re already full… make a new node on the largest single HDD you can afford. There are definately SNOs that have-filled/will-fill 20TB nodes. If you are running multiple nodes… but they don’t have their own /24 IP… then you won’t be filling them any faster (as all nodes behind the same IP share ingress)
But I’m not aware of any hard limits. Like if the largest generally-available consumer HDD is about 22TB: that will work fine.
I think we reach 18/20tb equilibrium in the last months… today everything changed. We lost 30/50% data on big old nodes. We are still alive … little bit depressed but alive:)
Yeah… I think SNOs had a general idea of reasonable growth rates…
…and then the network convulsed: with forever-free/suspended data getting trashed out the back while capacity-reservation/performance-testing data got rammed in the front.
So all bets are off.
(…but I still bet that SNOs can make good money if they stick around!)
This is probably the worst moment to ask this question as suddenly a lot is in motion and very unpredictable. A lot of test data and free account data is being removed and at the same time rapid test rotation has started. If you have the drives, just add them when your space is filling up. Which may happen fairly fast during testing. But due to the TTL used with current tests, that growth may also suddenly stop or reverse. It’s a little up in the air right now.