To achieve high availability we operate 4 relays, 1 Block Producer and an additional Stand-by Block Producer to allow a switchover with minimal latency. High available network disks (triple-stored) allow to auto-recover the machines in case of hardware failure.
Only the Relay Node Port is exposed to the public. All other traffic (SSH, Monitoring) is tunneled through VPN. Node/Relay communication through internal network.
2-FA Authentication to Management Console. SSH hardened (Root Login Disabled, Key based authentication with hardware wallet.
Root Disabled, No Pasword Login, Fail2Ban, Auto Security Updates, Service User for Cardano Service, Firewall configuration as described in Network Security
32/24 (Cloud/Private) GB RAM, 8 (Cloud/Private) Cores, High-available Network Disk (each block stored 3x), 160GB HDD, Inplace Upgrade possibility with minimal downtime through Standby approach.
Daily Snapshots and major version snapshots. Recovery time below 10 minutes
Email and SMS Alerts based on TIP check. If the database is not synced for more then 5 minutes.