On Wednesday, August 12, the Solana network came within a step of a complete halt, approaching a crash by just 14%, after 28.83% of the $SOL staked in the network fell out of consensus.
An overdue payment on a Teraswitch server, reported by the Marinade Finance staking platform, could have ended Solana's 30-month uninterrupted uptime streak. However, the network reached only 86% of that threshold.
If the volume of unstaked $SOL in the network had reached the 33.34% mark, Solana would have crashed, and the network would have gone offline.
Why Does the Solana Network Crash?
Solana stops finalizing blocks when more than one-third (33.34%) of all $SOL tokens staked in the network become invalid. Invalidity occurs when a network validator suddenly drops out of the consensus participant list.
The network approached the 33.34% trip point, reaching 28.83%, after which the situation began to stabilize.
According to Marinade, the overdue payment affected 90 validators, who lost 333 $SOL in rewards while being offline. Ironically, with $SOL priced at $76.9 at the time of this Cryptopolitan report, the entire Solana network could have been temporarily halted due to a $25,600 anomaly.
Since the crisis was averted, Solana's official status page still shows 100% cluster uptime over the past 90 days. The last full network halt occurred on February 6, 2024, and lasted approximately five hours.
How Was the Fix Applied?
According to Teraswitch's network status page, "customers on the LON1, AMS1, AMS2, AMS3, DUB1, DUB2, FRA2, SGP1, SGP2, TYO1, TYO2, and TYO3 stations experienced loss of access to internet resources, as well as to Teraswitch's internal backbone networks between the affected stations. Other stations in North America were not affected".
The statement also says that as a response measure, the company intentionally excluded MIA1 (Miami, Florida) from the backbone network.
The incident, which Marinade said was "barely recorded anywhere," was resolved after Teraswitch engineers "identified a corrupted route within 10 minutes of its start and removed MIA1 from the backbone network to stop further propagation".
Teraswitch clarified that "the affected sites switched to their local default routes, and service was restored at 04:16:15 UTC".
Which Solana Validators Went Offline?
Out of 74 validators tested by Marinade, only three recovered without issues: laine by Solana Strategies and Cogent Crypto, as well as Lion3d.
The rest of the validator pool temporarily failed for the duration of the episode. Solana's second-largest validator, Helius, was unavailable throughout the entire 33-minute outage.
Regarding the lost 333 $SOL in rewards, Marinade stated that operators would absorb the losses, and staking participants would not bear any losses.
Solana's Uptime Streak Almost Came to an End
Just two days before this near-crash incident, Solana had been running without interruption for 30 consecutive months. Back then, the root cause was a bug in the LoadedPrograms JIT cache, which caused validators to repeatedly recompile data until consensus stalled on a single block, as reported by Solana Compass. Anza patched the bug, and the network resumed operation after five hours.
The rationale for reliability is based on three changes implemented in 2023 and 2024: the QUIC transport layer with stake-weighted quality of service to limit spam, a priority fee market which now accounts for roughly 88% of daily fee revenue, and Firedancer, an independent validator client that Jump Crypto launched on the mainnet in late 2025.
The launch of a second client means a bug in one piece of code can no longer halt the entire network.
end-content
