You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@flink.apache.org by "Fabian Hueske (JIRA)" <ji...@apache.org> on 2016/09/29 15:03:20 UTC
[jira] [Updated] (FLINK-4636) AbstractCEPPatternOperator fails to
restore state
[ https://issues.apache.org/jira/browse/FLINK-4636?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Fabian Hueske updated FLINK-4636:
---------------------------------
Assignee: Jagadish Bihani
> AbstractCEPPatternOperator fails to restore state
> -------------------------------------------------
>
> Key: FLINK-4636
> URL: https://issues.apache.org/jira/browse/FLINK-4636
> Project: Flink
> Issue Type: Bug
> Components: CEP
> Affects Versions: 1.2.0, 1.1.2
> Reporter: Fabian Hueske
> Assignee: Jagadish Bihani
> Fix For: 1.2.0, 1.1.3
>
>
> The {{restoreState()}} of the {{AbstractCEPPatternOperator}} restores the a Java {{PriorityQueue}}. For that it first reads the number of elements to insert and then creates a {{PriorityQueue}} object. However, Java's {{PriorityQueue}} cannot be instantiated with an initial capacity of {{0}}, which is not checked.
> In case of an empty queue, the {{PriorityQueue}} should be instantiated with an initial size of {{1}}.
> See http://apache-flink-user-mailing-list-archive.2336050.n4.nabble.com/Problem-with-CEPPatternOperator-when-taskmanager-is-killed-tp9024.html
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)