What halted Starknet this week?

TLDR

Starknet (STRK) halted this week due to a technical mismatch between block execution and the proving system that stalled block production for hours, before a fix restored the network. See the breakdown in a news report.

1. Duration was about four hours, followed by a rollback and restart at roughly 14:02 UTC per a network update.
2. A full postmortem is pending; engineers acknowledged the issue and promised a detailed timeline and root cause in a follow?up note.
3. This is the second major disruption since the Grinta upgrade, which previously required chain reorganizations per a coverage summary.

Deep Dive

1. Duration And Fix

The outage lasted roughly four hours before Starknet applied a software fix and rolled back to stabilize operations. Block production resumed and transactions were processed again around 14:02 UTC, with a note that some transactions between 09:24 and 09:42 UTC may not have been processed correctly, per a network update.

2. Root Cause Under Review

Developers cited inconsistencies between transaction execution and the proving system, with some specific transactions triggering the fault. A full root cause analysis and prevention plan are to be published, per a news report.

1. The team acknowledged downtime and active investigation on X, with technical details to follow, per a coverage summary.
2. Prior outages in 2025 involved sequencer and RPC provider issues, indicating multi?component complexity in the stack, noted in the coverage summary above.

3. Reliability Context

It is the second notable incident since Grinta, an upgrade that introduced a decentralized sequencer and a new fee market, later linked to chain reorganizations that reverted activity and required resubmissions per a coverage summary.

1. The latest halt was resolved and the network restored to normal operations per a network update.
2. Similar L2s have had downtime episodes, but repeated halts raise reliability questions that teams typically address with architectural changes and monitoring, summarized in the coverage summary above.

Conclusion

Starknets halt was caused by execution versus proving inconsistencies that stalled block production, followed by a rollback and fix that restored service. The episode adds to recent reliability concerns since Grinta, making the forthcoming postmortem and mitigation plan important signals for developers and users.