[
https://issues.apache.org/jira/browse/TINKERPOP3-847?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14934245#comment-14934245
]
Marko A. Rodriguez commented on TINKERPOP3-847:
-----------------------------------------------
I do this in {{SparkGraphComputer}}.
{code}
try (final JavaSparkContext sparkContext = new
JavaSparkContext(sparkConfiguration)) {
....
}
{code}
{{Closeable}} and {{try{}} work like that. This is the only thing I see that
would hold resources like this. Does the stack trace tell you where in the code
you were when {{Too many open files.}} ?
> SparkGraphComputer resource leak
> --------------------------------
>
> Key: TINKERPOP3-847
> URL: https://issues.apache.org/jira/browse/TINKERPOP3-847
> Project: TinkerPop 3
> Issue Type: Bug
> Components: console
> Affects Versions: 3.0.1-incubating
> Reporter: Daniel Kuppitz
> Assignee: stephen mallette
> Fix For: 3.1.0-incubating, 3.0.2-incubating
>
>
> Yesterday, after working quite some time in the console, doing a lot of
> {{SparkGraphComputer}} stuff, I suddenly got an error: "Too many open files".
> Further attempts to start more OLAP jobs failed immediately. Then, a few more
> seconds later, the console was flooded with error messages coming from Spark
> (although I didn't do anything at this time).
> Restarting the console helped, everything was immediately back to normal.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
