[
https://issues.apache.org/jira/browse/IGNITE-14893?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17371950#comment-17371950
]
Vyacheslav Koptilin commented on IGNITE-14893:
----------------------------------------------
Hello [~Korol],
I left a few comments at your PR. Please take a look.
> Bug in GridCacheWriteBehindStore Flusher thread lookup
> ------------------------------------------------------
>
> Key: IGNITE-14893
> URL: https://issues.apache.org/jira/browse/IGNITE-14893
> Project: Ignite
> Issue Type: Bug
> Components: cache
> Affects Versions: 2.8, 2.7.5, 2.7.6, 2.9, 2.8.1, 2.10, 2.9.1
> Reporter: Mike W
> Assignee: Ilya Korol
> Priority: Major
> Time Spent: 1h
> Remaining Estimate: 0h
>
> There's a bug in GridCacheWriteBehindStore in the flusher method.
> [https://github.com/apache/ignite/blob/master/modules/core/src/main/java/org/apache/ignite/internal/processors/cache/store/GridCacheWriteBehindStore.java#L674]
> The logic states there that if flush thread count is not a power of 2, then
> perform this math, which is not guaranteed to return a positive number.
> {code:java}
> idx = ((h = key.hashCode()) ^ (h >>> 16)) % flushThreadCnt
> {code}
> For example, if you pass this string as a key, it returns a negative number:
> *accb2e8ea33e4a89b4189463cacc3c4e*
> and then throws an array out of bounds exception when looking up the thread.
> http://apache-ignite-developers.2346864.n4.nabble.com/IndexOutOfBoundsException-in-GridCacheWriteBehindStore-Flusher-thread-lookup-IGNITE-14893-tp52827.html
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
