[
https://issues.apache.org/jira/browse/IGNITE-12705?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17040980#comment-17040980
]
Maxim Muzafarov commented on IGNITE-12705:
------------------------------------------
[~slava.koptilin]
Of course, we should remove the order gracefully with the following assumption:
1. `ignite-sys-cache` must be rebalanced first (and he is a candidate for
removing to, because of distributed metastore has implemented)
2. SYNC caches must be rebalanced prior to ASYNC
3. We should provide static getCacheRebalanceOrder metrics, so a user can find
out a particular cache order.
Let's continue on dev-list? I think this discussion is also valuable for other
community members.
> Schedule rebalancing of SYNC caches with high priority
> ------------------------------------------------------
>
> Key: IGNITE-12705
> URL: https://issues.apache.org/jira/browse/IGNITE-12705
> Project: Ignite
> Issue Type: Improvement
> Affects Versions: 2.8
> Reporter: Vyacheslav Koptilin
> Assignee: Vyacheslav Koptilin
> Priority: Major
> Fix For: 2.9
>
> Time Spent: 10m
> Remaining Estimate: 0h
>
> In current implementation order in which rebalancing is triggered seems to be
> arbitrary, regardless of the configured rebalanceMode. In this case
> Ignition.start() method is forced to wait not only for rebalancing of SYNC
> caches, but also the ones that are scheduled for rebalancing before it.
> This behavior should be fixed by making SYNC caches rebalance first, before
> going to any other caches.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
