I didn't wanted to send a patch for that. The code is poorly documented nor
tested under many circumtances, old license, etc. etc. It was just a
workaround for a problem we had.

But we use it in production environment with about 350.000 pages a day. So,
for my usecase it's stable. And if you suggest to send it as a patch, I do.
Perheaps somebody finishes my work and we'll have it in the next version of
cocoon.

In addition to my ParallelContentAggregator class I have a version called
TonyContentAggregator. I can't remember if Tony did this, it was inspired by
Tony or whatelse. I also attached thisone as patch.

I have posted it as an enhancement to the bugzilla.
So, see attachements to the bugreport
http://nagoya.apache.org/bugzilla/show_bug.cgi?id=31646
for the javacode.

Maybe somebody put this in the scratchpad.

regards
Christoph


----- Original Message ----- 
From: "Ralph Goers" <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Monday, October 11, 2004 4:37 PM
Subject: Re: aggregate takes too much time


> Why not submit this as a patch?  Then we can all see it.
>
> Christoph Gaffga wrote:
>
> >we also had a problem with slow aggregates. We solved it by changing the
> >Aggregator so it works in a multithraded mode.
> >
> >I don't know if it increases the perfomcance in your case because you are
> >only aggregating simple xml files from the filesystem. But in our case
(the
> >aggragated parts are not static) it increases perfomance dramaticly.
> >
> >If you interested in it, I can send you our javacode for the
> >ParallelContentAggregator.
> >
> >regards
> >Christoph
> >
> >
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [EMAIL PROTECTED]
> For additional commands, e-mail: [EMAIL PROTECTED]


---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to