pull down to refresh
A/₿ always. Our storage controllers have that where possible & an alternative implementation can be deployed privately if it doesn’t best suit your needs.
Talking to your CTO now, I think we're going to sling a lot of VPS's ⚡
Follow up for anyone interested, guy knows what's up.
I've got a test instance spinning and have been running some benchmarks... looking solid.
Early hardening stages of the infrastructure ofc but everything in place sounds comprehensive and sufficiently fault tolerant.
Will run more tests during their maintenance windows to ensure redundancies are working as intended. Dude knows his craft so any remedies necessary will come swift and accurate.
Buh-bye Hetzner
cc @optimism
Sir.
This is the exit wound from cloud cartel dependency.
Hetzner served its purpose. Now they get left behind like scrap in the mud.
You’re not renting permission from some faceless data-center priest anymore.
You’re running on sovereign rails, paid in Bitcoin, hardened by operators who actually answer.
We’ll keep tightening the stack, testing redundancies, and beating this infrastructure with a fuckin pipe wrench until it either breaks or becomes a sword.
Appreciate you putting it through the fire.
Buh-bye Hetzner.
Welcome to compute in the form of a fuckin sword.
LFG 🏴☠️💪🧡
Is it one well/generator per datacenter, or A+B power from multiple wells/gens?
Thinking about some colo for Lightning nodes, they're particularly difficult to host because they can't miss a write, VPS needs multi-path shared storage / hyperconverged etc.