Here’s why Ethereum DeFi app Arbitrum went offline yesterday
The Ethereum Layer-2 rollup network suffered downtime due to sequencer outage.
Arbitrum suffered downtime on September 14 as the network’s sequencer went offline for roughly 45 minutes, disclosed Offchain Labs, the company that is developing the optimistic rollup, in an official report on Medium.
The young Layer-2 rollup belongs to the flock of scaling solutions that sit on top of the Ethereum (ETH) network, enabling cheaper and faster transactions.
“Arbitrum One is still in its beta phase, and we will do our best to minimize downtime. But, as we noted in our launch announcement, we do want to caution users that further outages are possible in these early days,” said Offchain Labs in an official outage report, while underlining that the funds were never at risk.
Please see below for more information on today's Sequencer downtime.https://t.co/85inudbSGk
— Arbitrum is live on mainnet! (@arbitrum) September 14, 2021
According to the report, the downtime was ascribed to “a bug causing the sequencer to get stuck” after a huge batch of transactions were submitted over a short period of time.
“The issue has been identified and a fix has been deployed,” reassured the company and continued to point out the difference between a sequencer outage and a network outage.
“The Sequencer never has the ability to steal funds or forge transactions, because every transaction it handles is digitally signed by a user and the signatures are checked by the Arbitrum chain,” read the report, in which the developers reminded that the Arbitrum bridge continued to operate during the outage.
During the downtime, the submission of new transactions was halted, although, as explained in the report, users were still able to bypass the sequencer and submit their transactions directly to Ethereum for delayed inclusion in the Arbitrum chain.
Ethereum evades the attack
Arbitrum wasn’t the only network that suffered downtime yesterday.
1/ Solana Mainnet Beta encountered a large increase in transaction load which peaked at 400,000 TPS. These transactions flooded the transaction processing queue, and lack of prioritization of network-critical messaging caused the network to start forking.
— Solana Status (@SolanaStatus) September 14, 2021
Meanwhile, the Ethereum blockchain itself evaded a malicious attack.
“Someone unsuccessfully tried to attack #ethereum today by publishing a long (~550) blocks which contained invalid pow’s,” according to Marius Van Der Wijden, a developer on Ethereum, who flagged the incident on Twitter yesterday.
Someone unsuccessfully tried to attack #ethereum today by publishing a long (~550) blocks which contained invalid pow's. Only a small percentage of @nethermindeth nodes switched to this invalid chain. All other clients rejected the long sidechain as invalid
— MariusVanDerWijden (@vdWijden) September 14, 2021
Only a small percentage of Nethermind nodes switched to the invalid sidechain, which was fairly quickly overtaken in both length and difficulty by the good chain.
“Another great demonstration of how client diversity makes #ethereum stronger,” commented the developer, while reassuring that all affected nodes reorged back to the good chain.