[
https://issues.apache.org/jira/browse/IGNITE-22807?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Roman Puchkovskiy updated IGNITE-22807:
---------------------------------------
Labels: iep-128 ignite-3 (was: ignite-3)
> Repair CMG on node start
> ------------------------
>
> Key: IGNITE-22807
> URL: https://issues.apache.org/jira/browse/IGNITE-22807
> Project: Ignite
> Issue Type: Improvement
> Reporter: Roman Puchkovskiy
> Assignee: Roman Puchkovskiy
> Priority: Major
> Labels: iep-128, ignite-3
> Fix For: 3.0.0-beta2
>
> Time Spent: 4h 10m
> Remaining Estimate: 0h
>
> If, during an Ignite node startup, it finds a ClusterResetMessage stored in
> its Vault, it does the following:
> # Before starting the networking, assign cluster ID from the message instead
> of the cluster ID stored in the local CMG data for the network handshake
> validation
> # Start the network
> # If the cluster ID stored in the local CMG data is different from the one
> from the message, then:
> ## Destroy the local CMG data and CMG Raft storages (meta, snapshots, log)
> ## Initialize a new CMG Raft node using the new CMG nodes (from the message)
> as voting set
> ## Submit a CmgInitCommand (having new CMG nodes, MG nodes from the message,
> new cluster name and ID, new cluster ID history) to the new CMG
> # Remove the message from the Vault
> # Proceed with the normal startup
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
