[
https://issues.apache.org/jira/browse/NIFI-4858?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16357670#comment-16357670
]
ASF subversion and git services commented on NIFI-4858:
-------------------------------------------------------
Commit 4428fe28bfac53ea859741cec2cb04212c5a6034 in nifi's branch
refs/heads/master from [~markap14]
[ https://git-wip-us.apache.org/repos/asf?p=nifi.git;h=4428fe2 ]
NIFI-4858: Expose Request Timeout from HTTP Context Map and use that to set as
the timeout of the 'AsyncContext' in HandleHttpRequest. Otherwise, the request
will never timeout (which is OK because the HttpContextMap will handle this).
However, Jetty behind the scenes is adding a task to Scheduled Executor for
each request with a delay of whatever the timeout is set to. Since it's
currently set to Long.MAX_VALUE, that task will never be run and as a result
the ExecutorService's queue will grow indefinitely, eventually exhausting the
JVM Heap
This closes #2460.
Signed-off-by: Bryan Bende <[email protected]>
> HandleHttpRequest has a memory leak
> -----------------------------------
>
> Key: NIFI-4858
> URL: https://issues.apache.org/jira/browse/NIFI-4858
> Project: Apache NiFi
> Issue Type: Bug
> Components: Extensions
> Reporter: Mark Payne
> Assignee: Mark Payne
> Priority: Major
> Fix For: 1.6.0
>
>
> When using HandleHttpRequest, if I issue millions of GET requests things seem
> to work okay. However, if I try to issue millions of POST requests, the JVM
> eventually runs out of heap.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
