Are you certain you applied the max.deployments.restart = 25? You can
always lower that number a bit too, for instance 15 might work for you.
That simply controls the number of tests which are run in the server
before starting a new one which will recover all the memory that is lost
from running the tests.
These settings should exist in your local.build.properties file under
the jboss-as directory.
On Tue, 2010-02-23 at 12:59 +0100, Jozef Hartinger wrote:
With Martin's settings I am not able to get rid of
tried even more generous settings (2GB heap and 1GB PermSize) with no luck.
On 02/22/2010 06:10 PM, Pete Muir wrote:
> Yes, the ootb settings are not generous enough, I already adjusted them some, but
perhaps we need to increase them more. Jozef, could you try with Martin's settings? If
they work, I will update the tag, zips and trunk with this.
> On 22 Feb 2010, at 15:44, Martin Gencur wrote:
>> I ran the CDI-TCK test bundle and results are following:
>> Tests run: 835, Failures: 0, Errors: 0, Skipped: 0, Time elapsed:
>> 2,039.701 sec :)
>> Maybe it was because I used these parameters:
>> max.deployments.restart = 25
>> On Mon, 2010-02-22 at 15:55 +0100, Jozef Hartinger wrote:
>>> I created the following issues:
>>> - Interceptors spec audit and coverage report missing in the
>>> distribution https://jira.jboss.org/jira/browse/CDITCK-115
>>> - Typo in dd-override-report.txt
>>> - release-notes.txt empty https://jira.jboss.org/jira/browse/CDITCK-117
>>> - JBoss maven repository declaration is missing
>>> - Documentation issues https://jira.jboss.org/jira/browse/CDITCK-119
>>> Furthermore, I was not able to get a clean TCK run against JBoss AS 6 M2
>>> (with updated Weld) in my environment - I am getting 100+ failures
>>> although nightly build seems OK in hudson.
>> weld-dev mailing list
weld-dev mailing list