Resilience

Incidents and fallback performance

Demonstrates how shared-entropy selection preserves liveness when proposers fail, DKG slips, or batch publication degrades.

Incident Feed

Selected proposer failed health checks; first fallback produced within the same slot budget.
high
proposer_failure • slot 50-62026-04-04T09:14:00Z
Fallback proposer activated after missed attest window.
medium
fallback_activation • slot 50-42026-04-04T08:40:00Z
Verifier state for epoch 49 arrived late due to partial committee outage.
medium
dkg_delay • slot 49-142026-04-04T06:20:00Z
Missed slot before fallback threshold tuning was applied.
high
missed_slot • slot 49-122026-04-04T05:55:00Z
DA publication lag exceeded SLO for one DeFiRollup batch.
low
batch_publication_delay • slot 50-12026-04-04T04:15:00Z

Why It Matters

Centralized sequencer failure mode

Single operator outage stalls ordering until manual failover or external coordination.

Entropy-based fallback behavior

If the selected proposer fails, fallback proposer takes over without coordination.

Incidents Over Time

Downtime Avoided via Fallback