[
https://issues.apache.org/jira/browse/IGNITE-12705?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17040961#comment-17040961
]
Vyacheslav Koptilin commented on IGNITE-12705:
----------------------------------------------
Hello [~mmuzaf],
> Can we deprecate and completely remove the relanaceOrder property in this PR?
> I think it will simplify everything.
> See:
> http://apache-ignite-developers.2346864.n4.nabble.com/DISCUSSION-Deprecation-of-obsolete-rebalancing-functionality-td45824.html
Yes, I have seen this discussion and it looks like there is a consensus that
{{rebalanceDelay}} and {{rebalanceOrder}} should be deprecated.
Please take a look at this message:
http://apache-ignite-developers.2346864.n4.nabble.com/DISCUSSION-Deprecation-of-obsolete-rebalancing-functionality-tp45824p45964.html
> Can we deprecate and completely remove the relanaceOrder property in this PR?
> I think it will simplify everything.
I don't think that removing existing APIs in minor releases is a good idea even
though there is an agreement of the community that this API is deprecated.
IMHO we can remove {{rebalanceDelay}}, {{rebalanceOrder}} and
{{CacheRebalanceMode#NONE}} in major release which will be 3.0, and it is
already covered by IGNITE-12662 and IGNITE-11417
I can mark these properties with {{@Deprecated}} in this PR. However, I think
it would be nice to do this under a separate ticket. What do you think?
> 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)
