[
https://issues.apache.org/jira/browse/IGNITE-5597?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16081776#comment-16081776
]
Andrei Yakushin commented on IGNITE-5597:
-----------------------------------------
Patch available
https://github.com/apache/ignite/pull/2268
> Wrong javadoc in Affinity and AffinityFunction for REPLICATED cache
> -------------------------------------------------------------------
>
> Key: IGNITE-5597
> URL: https://issues.apache.org/jira/browse/IGNITE-5597
> Project: Ignite
> Issue Type: Bug
> Affects Versions: 1.7
> Reporter: Evgenii Zhuravlev
> Assignee: Andrei Yakushin
> Labels: javadoc
>
> RendezvousAffinityFunction.getPartitions() Javadoc says:
> {code:java}
> * Note that for fully replicated caches this method should always
> * return {@code 1}.
> {code}
> but it's not true, it works the same as PARTITIONED cache.
> Affinity.mapKeyToNode(K key) javadoc says:
> {code:java}
> * <li>
> * For fully replicated caches first node ID returned by {@link
> AffinityFunction}
> * is returned.
> * </li>
> * <li>For partitioned caches, primary node for the given key is
> returned.</li>
> {code}
> it looks confusing, while REPLICATED cache has primary nodes for keys as
> PRATITIONED.
> Also,
> {code:java}
> * Provides affinity information to detect which node is primary and which
> nodes are
> * backups for a partitioned cache.
> {code}
> Affinity matter for REPLICATED cache too.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
