quality@glassfish.java.net

Re: v3 not deploying huge EAR anymore

From: Judy Tang <Judy.J.Tang_at_Sun.COM>
Date: Thu, 29 Oct 2009 08:44:58 -0700

Hi Matthias,

Very happy to get your help. This is just the test case we need. We
will work
together on this !

Thanks,
Judy

matthias.fraass_at_tricoder.net wrote:
> \o/
>
> Once again, I'm baffled on the response time of the GF team!
>
> A quick note, so you can put up a test case:
> - the number of files in that jar is about 47.000
> - there are directories with more than 5.000 files.
> - the v2 algorithm seemed to be OK with this :).
>
> -Matthias
>
>
> 2009/10/29 Tim Quinn <Timothy.Quinn_at_sun.com
> <mailto:Timothy.Quinn_at_sun.com>>
>
> Hi, again, Matthias.
>
> The stack trace might turn out to have been very useful.
>
> The GenericAnnotationDetector's scanArchive method invokes the
> InputJarArchive's entries method which returns an Enumeration of
> the entries in the archive - the JAR in this case. The current
> implementation of InputJarArchive.entries is inefficient and could
> account for the problem you are seeing.
>
> Hong and I are talking to see about how that method might be improved.
>
> Stay tuned!
>