Polygon, operating on a proof-of-stake (PoS) system, experienced a
brief service disruption on July 30, lasting approximately one hour and affecting various decentralized applications and user accessibility.
According to statements from Polygon’s CEO and founder,
Sandeep Nailwal, the network incident
did not completely halt the generation of new blocks.
However, QuickNode Status
indicated
a temporary cessation in block production, precisely for one hour at block number 74,592,238, echoing
reports from users
shared on
X
(formerly Twitter).
Consensus Challenge
Nailwal mentioned in a post on X, that several RPC (Remote Procedure Call) providers, “and thus, connected apps and users,” encountered difficulties during a two-to-three hour window. While this occurred, the Polygon chain “remained active, continually creating blocks and handling user transactions” through functioning RPC services.
Data extracted from Polygonscan, a block explorer, reveals that the network produced a fresh block just
two seconds
following block number 74,592,238, as of the current reporting time. This is despite the block explorer displaying a period of inactivity in block creation around that particular block height.
Nailwal attributed the incident to a rapid fix implemented and a temporary suspension affecting the consensus layer, specifically the Heimdall component, which was related to a recent and intricate system update. While the execution layer (Bor) continued to function, some RPC nodes became out of synchronization after the hotfix was applied. This resulted in application-level errors that were perceived by end-users as a full network shutdown.
Nailwal offered an apology for the impact on end-users, emphasizing that Polygon is actively partnering with service providers “to ensure everyone is updated,” anticipating no subsequent problems.
Block Production Interrupted
Infrastructure operators reported similar observations during this timeframe. QuickNode indicated that the main network stalled from their perspective at block height 74,592,238 and cautioned that the latest
Heimdall v0.2.16 upgrade
was creating challenges with execution clients, including Bor and Erigon.
QuickNode opted to suspend the upgrade process “until more information is available,” advised operators against continuing, and initiated a resynchronization of full nodes, while simultaneously resetting Erigon instances with the aim of restoring service functionality.
Polygon’s official status page
pinpointed the issue
to the Heimdall-V2 network. The team acknowledged that the primary Heimdall service had ceased responding, impacting validator and checkpoint visibility through Heimdall APIs. However, they emphasized that the Bor layer remained unaffected and continued to operate normally.
Engineers successfully located the root cause of the disruption and implemented a corrective fix, before subsequently marking the incident as resolved.
The event timeline reveals the sequence of the disruption and its impact across various teams. Polygon acknowledged the event at 09:52 UTC on July 30, identified the cause by 09:57, and announced its resolution at 11:01 UTC.
Following this, QuickNode reported the stall at 11:28 UTC, paused the Heimdall v0.2.16 rollout at 11:51 UTC pending guidance from the Polygon Foundation, and subsequently reported at 15:39 UTC that it was actively resynchronizing and resetting nodes to restore service availability.
Nailwal described the event as a misalignment between consensus mechanisms and infrastructure elements rather than a fault within the protocol itself.