Built so a machine can fail and nobody notices
Every receipt is backed on two machines before it reaches you, and the standby takes over in about a second.
For operators and finance teams who need to know what happens to money in flight when hardware does what hardware eventually does.
Every payment system has a bad day in its future. A disk fails, a process hangs, a data center loses power. The question is not whether, but what happens to the payment that was in flight. On Paxeer X the answer is simple. If you hold a receipt, your transaction lands. There is no window where a receipt was issued but the only copy sat on a machine that just died.
If you hold a receipt, your transaction lands. That is the whole rule.
Two machines, one promise
What your systems see
| Situation | What happens to your transaction |
|---|---|
| The primary fails after your receipt | It lands at its promised position in the next block, about a second later than usual. |
| The primary fails before your receipt | You never got a receipt, so nothing is ambiguous. Resend the same transaction and it is accepted exactly once. |
| The standby is unreachable | New sends are refused with an error until it is back. No receipt is issued that cannot be backed. |
| Everything is healthy | You notice none of this, which is the point. |
Why this matters more for software than for people
A person who sees a failed payment screen tries again and moves on. An agent making a thousand payments an hour has no patience and no judgment; it needs a system where every answer is true. A receipt it holds must land. A send that cannot be backed must be refused, loudly, so its retry logic does the right thing. Paxeer X is designed around that software, and people get the benefit for free.
The hand-over between machines is fenced so that two machines can never both believe they are in charge, and a receipt once issued is never re-signed or altered. Operators rotate hardware during business hours without telling anyone, and nobody needs to be told.
Related
See who keeps the lights on
Node roles, what each one does, and how the network stays up when one of them does not.

