You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@felix.apache.org by "Jean-Baptiste Onofré (Jira)" <ji...@apache.org> on 2021/04/15 06:19:00 UTC
[jira] [Commented] (FELIX-6393) FileInstall | In main loop, we have
serious trouble | java.lang.NullPointerException: null
[ https://issues.apache.org/jira/browse/FELIX-6393?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17321930#comment-17321930 ]
Jean-Baptiste Onofré commented on FELIX-6393:
---------------------------------------------
Just got the same report on Karaf:
{code:java}
On Karaf 4.3.0 I see the following error logged every few milliseconds:
2021-04-14T08:09:13,749 | ERROR | fileinstall-/[KARAF_HOME]/deploy | fileinstall | 15 - org.apache.felix.fileinstall - 3.6.8 | In main loop, we have serious troublejava.lang.NullPointerException: null at org.apache.felix.fileinstall.internal.Scanner.processFiles(Scanner.java:158) ~[!/:3.6.8] at org.apache.felix.fileinstall.internal.Scanner.scan(Scanner.java:131) ~[!/:3.6.8] at org.apache.felix.fileinstall.internal.DirectoryWatcher.run(DirectoryWatcher.java:311) [!/:3.6.8]
It looks like this can happen on a system that has too few file handles configured. The Felix DirectoryWatcher then use a Scanner instead of a WatcherScanner implementation.Scanner doesn’t have a null check for the file filter object, and the deploy fileinstall configuration (org.apache.felix.fileinstall-deploy.cfg) does not define a filter, giving the NPEs.It looks like it should also be possible to reproduce this error by configuring the watcher property felix.fileinstall.disableNio2. I will look into addressing this by increasing the account’s available file handles, but I was wondering if there should be additional protection in Karaf (by setting a filter) or Felix (guarding against the NPE)? {code}
I will improve this.
> FileInstall | In main loop, we have serious trouble | java.lang.NullPointerException: null
> ------------------------------------------------------------------------------------------
>
> Key: FELIX-6393
> URL: https://issues.apache.org/jira/browse/FELIX-6393
> Project: Felix
> Issue Type: Bug
> Components: File Install
> Affects Versions: fileinstall-3.6.8
> Reporter: Rakesh Kumar
> Assignee: Jean-Baptiste Onofré
> Priority: Critical
>
> I was trying to configure the Felix file install and after the configuration in Felix WebConsole is done I placed a bundle in configured load directory and it went into an endless exceptions.
> I had to kill the server to get over it.
> Here is the exception trace.
> {code:java}
> 2021-03-07 06:36:09.942 [FelixLogListener] ERROR com.adeptj.modules.commons.logging.osgi.Logger - In main loop, we have serious trouble
> java.lang.NullPointerException: null
> at org.apache.felix.fileinstall.internal.Scanner.processFiles(Scanner.java:158)
> at org.apache.felix.fileinstall.internal.Scanner.scan(Scanner.java:131)
> at org.apache.felix.fileinstall.internal.DirectoryWatcher.run(DirectoryWatcher.java:311)
> {code}
--
This message was sent by Atlassian Jira
(v8.3.4#803005)