[
https://issues.apache.org/jira/browse/IGNITE-8738?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16526548#comment-16526548
]
Alexey Goncharuk commented on IGNITE-8738:
------------------------------------------
[~ezagumennov], I think it would be useful if the output will be consistent on
all nodes. Currently, if nodes leave grid in parallel, node A may print
coordinator change C1->C2->C3 for topology events E2, E3, while some other node
may print coordinator.
Please add a method {{oldestServerNode}} to disco cache which will be similar
to {{oldestAliveServerNode}}, but it should not filter nodes by alive flag and
use it in the print output.
> Improve coordinator change information
> --------------------------------------
>
> Key: IGNITE-8738
> URL: https://issues.apache.org/jira/browse/IGNITE-8738
> Project: Ignite
> Issue Type: Improvement
> Reporter: Alexey Goncharuk
> Assignee: Evgenii Zagumennov
> Priority: Major
>
> When topology changes and coordinator is also changed, we need to print out
> this alongside with topology information.
> An example of such message:
> {{Coordinator changed [prev=node.tostring(), cur=node.tostr()]}}
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
