Solana CEO hoses down claims network outages caused by on-chain voting

Anatoly Yakovenko, the founder and CEO of Solana Labs, has denied claims that Solana’s network outages were being caused by a high volume of validator messages and its on-chain voting system clogging its consensus layer.

While the Solana Foundation confirmed in a Feb. 27 post that the “root cause” of the recent 20-hour network outage is still not clear, the CEO responded to speculation that Solana’s decision to include on-chain votes as transactions is a “massive design flaw” that has led to its many outages.

The controversial thread claiming that that the high volume of validator messages and on-chain votes were clogging the network was posted by Twitter user DBCryptoX on Feb. 27, days after Solana’s 20-hour network outage.

1/Yesterday #Solana had another 20 hour outage
Just one of about a dozen times the ⛓️ has gone down. But why?
All part of a massive design flaw that I will try to break down in this
So let’s get into it… pic.twitter.com/KmeUPnnlZJ
— DBCryptoX ⚡️ (@DBCrypt0) February 26, 2023

However, in a response tweeted 20 minutes later, Yankovenko called the theory as coming from “pure ignorance.”

In short, he explained that the votes — which are part of a “single giant quorum” — contribute to provide an “exceptional level of security and high throughput and low fees” simultaneously.

Why are votes transactions? Every thread that I’ve seen that talks about this comes form pure ignorance.
Classic BFT consensus requires quadratic messaging overhead.
The more nodes you have in the same quorum, the part of the network that agrees on the state, the more messages… https://t.co/8lOhICb8mn
— toly (@aeyakovenko) February 27, 2023

However, Yakovenko didn’t exactly refute DBCryptoX’s claim that 90-95% of transactions on Solana comprise these validator messages and on-chain votes, which DBCryptoX said has helped “bog down the system.”

*DBCryptoX claims that validator messages and on-chain votes are clogging up the Solana network. Source: Twitter.*

DBCryptoX also claimed that the network outages were last 20 hours because it takes considerable time for validators to meet and reach a consensus (and thus a solution) using off-chain means, such as a messaging system like Discord.

Commentators on DBCryptoX’s initial post also appear to have disagreed with the theory.

Software engineer Alex Kroeger of Solana-powered Wallet Phantom said that there is likely no singular cause of the network outages and that validators of proof-of-stake systems need a lot of network communication to achieve validation.

While the network officially restarted on late on Feb. 25, it appears as though members of the cryptocurrency community are getting tired of the frequent network outages on Solana.

Cointelegraph reached out to Solana Labs for comment but didn’t receive a response by the time of publication.