Polygon Heimdall V2 Mainnet Restored After Consensus Bug Disruption

According to Cointelegraph, the Polygon Heimdall V2 mainnet, a crucial component of the Polygon proof-of-stake chain, experienced a temporary outage on Wednesday due to a suspected consensus bug. The issue, which affected the communication between node validators, was resolved within an hour. Despite the disruption, the Bor layer, responsible for block production and transaction processing, remained unaffected. Polygon representatives confirmed that the outage was triggered by an unidentified validator's exit from the network. During the downtime, block production on the Bor mainnet continued uninterrupted, and any discrepancies in network uptime and block explorer data are being addressed. Polygon is actively working with RPC providers to resolve sync inconsistencies that emerged following Heimdall's recovery. One provider has already resumed operations, albeit with a slight delay in synchronization. The incident underscores the importance of consistent network uptime for blockchain networks, which aim to offer a reliable alternative to traditional financial systems. The Heimdall V2 upgrade, launched in early July, aimed to enhance the network's performance by reducing finality times to approximately five seconds and updating the tech stack with CometBFT and Cosmos-SDK v0.50. This upgrade is considered the most technically complex hard fork since Polygon's launch in 2020, according to Polygon co-founder Sandeep Nailwal. While the upgrade has improved block times and network throughput, it has also introduced additional complexity and potential points of failure. Previously, the Heimdall V1 layer was also a source of network downtime. In March 2022, Polygon faced several hours of downtime due to a software bug in the Heimdall layer, which caused validators to operate on different blockchain versions. The ongoing efforts to enhance network performance highlight the challenges blockchain networks face as they strive to maintain uptime amid increasing complexity.