
Bitcoin2 min read
BTCLE , UNCX incident
BTCLE Official Security Update | October 14, 2026 Following the recent security incident, we would like to share an important update with our community. 1. UNCX Security Breach UNCX Network h
A global outage that would turn off Bitcoin, Ethereum, and Solana machines for five minutes would stop these networks. It would not erase your cryptos however. In this thought experiment, all

A global outage that would turn off Bitcoin, Ethereum, and Solana machines for five minutes would stop these networks. It would not erase your cryptos however. In this thought experiment, all validation machines shut down, while their disks remain intact. The blocks already recorded therefore remain, and only new block production is interrupted.
Let’s imagine that all equipment of the major blockchains switch off simultaneously. Five minutes later, power returns. The machines restart, but not all at the same pace. This scenario allows separating two often confused ideas: preserving a history and continuing to extend it. It also sheds light on the notion of finality, the moment when a transaction becomes final, which Ethereum researchers want to accelerate.
The hypothesis remains artificial however. A public network outage would not shut down everything, because many data centers have batteries and generators. So we assume a complete shutdown of all machines. Disks, databases, and backups remain intact.
Bitcoin does not produce a block every ten minutes on the dot. Mining remains a lottery. According to the Bitcoin technical documentation, the network adjusts the difficulty every 2,016 blocks. It aims for about two weeks, so ten minutes per block on average according to our calculation.
This pace allows a simple calculation. Suppose blocks arrive randomly, at a constant rate. The probability of seeing none in five minutes then reaches 60.65%. Five minutes without a block does not therefore prove an outage. This is a theoretical model, not a measurement of the real network.
During a total shutdown however, no miner looks for a block anymore. When power returns, miners resume from the last chain they know. If groups restart separately, two branches may coexist for a time. Nodes then follow the chain that required the most work and abandon the other. Bitcoin’s white paper also anticipates that nodes leave and come back freely.
Nodes can leave the network and rejoin at will, accepting the proof-of-work chain as proof of what happened during their absence.
Satoshi Nakamoto, Bitcoin white paper, 2008
Ethereum follows another clock. Each slot lasts 12 seconds, and 32 slots make an epoch of 6.4 minutes. A 300-second outage therefore covers 25 slots. These lost opportunities are not made up: the network simply resumes with the following slots.
Finality poses a more serious challenge. To finalize blocks, validators representing at least two-thirds of the staked ether must agree on checkpoints. During the outage, no votes circulate, and finality freezes. However, finalized blocks remain confirmed. Only a recent transaction, not yet finalized, requires re-verification after the restart.
Ethereum also has an emergency procedure. According to its page on rewards and penalties, it only starts after more than four epochs without finality, i.e., more than 25.6 minutes. Five minutes of outage is therefore insufficient to trigger it. Yet two-thirds of the staked ether must vote again within about twenty minutes afterwards. Moreover, the same page specifies there is “no penalty for not proposing a block”. However, each missed vote incurs a small penalty, equivalent to the reward not earned. An absence does not equate to “slashing,” a sanction reserved for serious faults.
Ethereum has already experienced a real loss of finality. According to the Prysm team’s report, the network finalized nothing for 9 epochs on May 12, 2023, about one hour. Several clients struggled to process some votes, and validator participation dropped to 30.7%. The emergency procedure was then triggered.
Solana changed parameters twice in October 2026. On October 9, the slot duration was reduced to 200 milliseconds, down from 400 in August. This reduction happened in four steps. On October 21, the network activated Alpenglow, a new consensus replacing TowerBFT. According to the Solana Foundation, it targets a finality of about 150 milliseconds.
Alpenglow tolerates up to 20% offline stake, in addition to 20% malicious stake. A global outage would far exceed this threshold. The 150 milliseconds say nothing about the recovery delay after a total shutdown.
Solana already experienced a complete shutdown of a different kind. On December 4, 2020, a validator launched in duplicate broadcast contradictory blocks. It revealed a known flaw in block propagation, and confirmations stopped. According to the Solana report, 393 validators had to coordinate a restart, achieved in less than six hours. About thirty provisionally confirmed blocks were lost. According to the report, they contained no transfers.
After resumption, three cases arise.
Your wallet does not contain your cryptos. It keeps the keys that allow spending them. After an outage, the challenge remains to recover these keys and a coherent chain.
Five minutes without power does not mean five minutes of interruption. A data center, an internet service provider, or a connection node each restart at their own pace. No specification therefore allows calculating an overall recovery delay. Furthermore, exchanges and DeFi applications could remain unavailable even after blocks return.
Rotor, the second phase of Alpenglow, is to replace Solana’s block propagation system. As of November 1, 2026, it had no deployment date yet, according to the Solana Foundation. Our article on Alpenglow activation details this schedule.