Skip to content

Devnet rolling update nodes halt issue #2302

@anunay-xin

Description

@anunay-xin

https://xinfin-team.atlassian.net/wiki/spaces/PT/pages/48857090/2026-04-13+devnet+rolling+update+halt

A rolling image update was run across 95 masternodes (31–125) while the network was live. This killed nodes mid-sync. When they restarted, their state databases were incomplete — headers existed up to block 4,244,966 but executed state only reached ~4,244,756. The sync engine entered a silent retry loop that never made progress. Only 30 nodes (1–30, never restarted) had complete state and could vote. 30 votes < 73 required → chain frozen.

Metadata

Metadata

Assignees

Labels

bugSomething isn't working

Type

No type

Projects

No projects

Relationships

None yet

Development

No branches or pull requests

Issue actions