The Solana seven-hour outage took place between April 30 and May 1. Network operators say the incident was due to massive transactions from non-fungible token (NFT) minting bots.
A record-breaking 4 million transactions, or 100 gigabits of data per second, crowded the network, knocking validators out of consensus and causing Solana to go dark at 8 p.m. UTC on April 30.
Validators could not effectively restore the mainnet until seven hours later, at 3 a.m. UTC on May 1.
Solana seven-hour outage: What exactly happened?
“Solana Mainnet Beta lost consensus after an enormous amount of inbound transactions (4m per second) flooded the network, surpassing 100gbps,” Solana said in a tweet last April 30.
According to the Twitter post, engineers were still analyzing why the network could not recover from the transaction overload that triggered the outage at the time.
Finally, late Saturday night, through the Solana discord, one of the validators supplied a Google doc with instructions that allowed the operators to restart the cluster at slot 131973970.
The bots hoarded Candy Machine, a popular tool used by Solana NFT initiatives to launch collections. Metaplex stated that bot traffic on their app was partially to blame for the network meltdown.
According to Solana, “high compute transactions” reduced network capacity to “several thousand” transactions per second (TPS). However, the company said that figure is well below the projected 50,000 TPS.
Outage crashes the price of SOL
The outage led the price of SOL, the blockchain’s native coin, to plummet nearly 7% to $84, though prices have subsequently recovered to slightly over $89.
Metaplex said it would impose a 0.01 SOL or $0.89 fee on wallets that attempt to complete an invalid transaction. The company described as “typically done by bots that are blindly trying to mint.”
Solana also went offline in September 2021 after being swamped by hundreds of thousands of transactions per second. However, the outage lasted had lasted longer, for approximately 17 hours.