You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@karaf.apache.org by "Jean-Baptiste Onofré (Jira)" <ji...@apache.org> on 2020/01/09 16:29:00 UTC

[jira] [Resolved] (KARAF-5628) Corrupt gc.log due to unseparated VM settings

     [ https://issues.apache.org/jira/browse/KARAF-5628?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Jean-Baptiste Onofré resolved KARAF-5628.
-----------------------------------------
    Resolution: Fixed

> Corrupt gc.log due to unseparated VM settings
> ---------------------------------------------
>
>                 Key: KARAF-5628
>                 URL: https://issues.apache.org/jira/browse/KARAF-5628
>             Project: Karaf
>          Issue Type: Bug
>          Components: karaf
>            Reporter: Rico Neubauer
>            Assignee: Jean-Baptiste Onofré
>            Priority: Major
>             Fix For: 4.3.0, 4.2.8
>
>
> The VM properties used for the start, stop, status and client scripts all use one common configuration of the VM properties, like Xmx and GC settings.
> This has one particular issue when activating the gc.log (-Xloggc:gc.log with Oracle-VM): Invoking e.g. status while Karaf is running, uses the same settings, thus points to the same file and corrupts it with NUL values (see here for details: [Stackoverflow|[https://stackoverflow.com/questions/8353401/garbage-collector-log-loggc-file-rotation-with-logrotate-does-not-work-properl]] ).
>  
> I would like to request the possibility to configure the VM properties for starting the instance independently from the maintenance scripts - due to the issue and also since e.g. running status with lots of heap (if instance is configured to it) is unneeded.
>  
> Even making it broader, I think it would also make sense to be able to configure the settings completely, since currently some things are hard-coded in the scripts, like the _-Xdum_p and _-Xlp_ settings  for AIX, where I personally would like to configure it to _-Xdump:heap:events=gpf,opts=PHD_, which is currently not possible using the configuration only. Also especially for AIX, there is _LDR_CNTRL=MAXDATA=0xB0000000@DSA_, which might not be desired when running with heaps larger than 3GB.
>  
> If you agree on a separation, I would make a proposal as PR. Please let me know.



--
This message was sent by Atlassian Jira
(v8.3.4#803005)