[
https://issues.apache.org/jira/browse/IGNITE-23079?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Roman Puchkovskiy updated IGNITE-23079:
---------------------------------------
Description:
When destroying a Raft group storages (log, snapshots, Raft meta), we should do
it durably. That is:
# Register an intent to destroy a particular group storages
# On group recovery (on node start, maybe on partition start) perform
non-finished destructions
was:
Currently, we just write a message to log if Raft storages could not be
destroyed for some Raft group, and that's all. This is not too visible for the
user. We could improve on this. A few ideas follow:
# Fail the node with the Failure Handler
# Retry deletion attempts automatically
# Add some metric
> Make Raft storages destruction durable
> --------------------------------------
>
> Key: IGNITE-23079
> URL: https://issues.apache.org/jira/browse/IGNITE-23079
> Project: Ignite
> Issue Type: Improvement
> Reporter: Roman Puchkovskiy
> Priority: Major
> Labels: ignite-3
>
> When destroying a Raft group storages (log, snapshots, Raft meta), we should
> do it durably. That is:
> # Register an intent to destroy a particular group storages
> # On group recovery (on node start, maybe on partition start) perform
> non-finished destructions
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
