Bots again managed to disrupt Solana’s operations this year and brought all operations to a halt. The rogue bot traffic managed to overcome network safeguards and caused an outage that lasted 7 hours.
Stakeholders of Solana rushed to correct the network on Saturday immediately after an insane volume of data flooding the PoS chain. This knocked validators outside consensus block production and came to a grinding halt.
Bots swarmed Candy Machine, the busy minting tool of NFT on Saturday, causing an unprecedented flood in the inbound traffic; around 4M requests for transactions and 100 GB of data per second.
This was the highest faced by the Solana network, one source at Solana Foundation revealed. This unprecedented swarm of bot activity pushed the validators out of the consensus. It was impossible to continue any block production and finally, the network went dark after 4:32 pm.
By 11:00 pm EST, validators used the Discord channels of Solana and Google Doc generated by a validator to restart the cluster.
Solana Founder Anatoly Yakovenko Answered Accusation Of MIA Against Him
Anatoly Yakovenko was on the move during the outage and credited the community of validators for leading the recovery of the mainnet. He was attacked on social media for being ‘missing in action during the crisis that plagued the network.
There had been a 17-hour outage last September but this time the restart didn’t try to resolve it with improved or fresh code populating. Validators simply picked up from where the network crashed 7 hours earlier.
While preparing for a restart, the validators pondered over whether to bring in code to momentarily block transaction transactions over Candy Machine. Some were against it and said that any sort of move would be considered. This hypothetical situation would be possible only if the validators chose to
Other areas of the Solana network stepped in to strengthen the defense. The Solana outage led to the brief but bloody drawdown in the SOL markets. The native token of Solano crashed to $83.13, a 24-hour low.
