[ 
https://issues.apache.org/jira/browse/MINIFICPP-308?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

marco polo resolved MINIFICPP-308.
----------------------------------
    Resolution: Fixed

Closed with commit 5bd67470e3922c0b51a1f60998c5b28fcbcdd363 (HEAD -> master, 
upstream/master, origin/master, origin/HEAD)
Author: Andrew I. Christianson <a...@andyic.org>
Date:   Wed Nov 15 19:22:49 2017 -0500

    MINIFICPP-308 Ensure that all core resources are deterministically released 
at the end of onTrigger calls so that script engines using GC do not hold onto 
them and waste resources
    
    This closes #190.
    
    Signed-off-by: Marc Parisi <phroc...@apache.org>

> Script objects should release underlying minifi objects within expected scope
> -----------------------------------------------------------------------------
>
>                 Key: MINIFICPP-308
>                 URL: https://issues.apache.org/jira/browse/MINIFICPP-308
>             Project: NiFi MiNiFi C++
>          Issue Type: Improvement
>            Reporter: Andrew Christianson
>            Assignee: Andrew Christianson
>            Priority: Blocker
>             Fix For: 0.3.0
>
>
> ExecuteScript sessions are holding onto some underlying minifi objects, such 
> as FlowFile instances, longer than we would like. This is due to the GC 
> implementations of the scripting environments, for example:
> https://github.com/ThePhD/sol2/issues/129
> ExecuteScript should therefore offer proxy objects (FlowFile is already 
> proxied by ScriptFlowFile) which release any resource-intensive minifi 
> objects as soon as conditions are met where the script should not longer use 
> them (after onTrigger of context/session, after transfer or remove for flow 
> files).



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)

Reply via email to