[
https://issues.apache.org/jira/browse/IGNITE-16363?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Denis Chudov updated IGNITE-16363:
----------------------------------
Description:
We need to be sure that recovery should perform on the node after it has joined
physical topology and has been validated via node join protocol.
Necessary prerequisites for the recovery start are following:
* the node has retrieved the information and metastorage group from Vault;
* receiving a join response by the node in case of non-standalone mode,
meaning that the node is validated and is allowed to join the cluster. This
step can be mocked for now;
* all components have started and subscribed on configuration updates and
events. After this, notifications should be allowed and the recovery actually
starts.
was:We need to be sure that recovery should perform on the node after it has
joined physical topology and has been validated via node join protocol.
> Provide a guarantee of completeness of pre-recovery actions
> -----------------------------------------------------------
>
> Key: IGNITE-16363
> URL: https://issues.apache.org/jira/browse/IGNITE-16363
> Project: Ignite
> Issue Type: Improvement
> Reporter: Denis Chudov
> Priority: Major
> Labels: ignite-3
>
> We need to be sure that recovery should perform on the node after it has
> joined physical topology and has been validated via node join protocol.
> Necessary prerequisites for the recovery start are following:
> * the node has retrieved the information and metastorage group from Vault;
> * receiving a join response by the node in case of non-standalone mode,
> meaning that the node is validated and is allowed to join the cluster. This
> step can be mocked for now;
> * all components have started and subscribed on configuration updates and
> events. After this, notifications should be allowed and the recovery actually
> starts.
--
This message was sent by Atlassian Jira
(v8.20.1#820001)
