Patches posted to the dev list tend to get lost if none looks at them
straight away. I suggest you create a bugzilla entry and add your
patch to that.
Mark
Rafael H. Schloming wrote:
> Is no interest in merging this patch? I'd like to see it accepted, so if
> there is anything I can do, please le
Is no interest in merging this patch? I'd like to see it accepted, so if
there is anything I can do, please let me know.
On Fri, 2005-12-02 at 11:23 -0500, Rafael H. Schloming wrote:
> Attached is an updated patch for the directory listings cache. I've made
> the following changes:
>
> * the ca
I can confirm that the benchmarks I was doing were without XSLT. I was
using a directory with 2000 empty files on a dual processor 2.8Ghz Xeon
with 1G of RAM. I didn't make any changes to the default config produced
by ant deploy.
On Thu, 2005-12-01 at 19:18 +, Mark Thomas wrote:
> Tim Funk wr
Tim Funk wrote:
It looks like this issue only occurs when the XSLT transformation is
done on the directory listing. Which is not an out of box configuration
so there is no worry for DOS (in case anyone is wondering).
Actually, there is an issue here without the XSLT. This is why
directory lis
Thanks for the feedback. I'll make the changes you suggest as soon as I
have a chance and repost the patch.
On Wed, 2005-11-30 at 19:47 -0500, Yoav Shapira wrote:
> Hi,
> It's an interesting patch and an interesting idea in general. I'm
> impressed with the performance findings, although I haven'
It looks like this issue only occurs when the XSLT transformation is done on
the directory listing. Which is not an out of box configuration so there is
no worry for DOS (in case anyone is wondering).
Otherwise - looks interesting. cacheTTL and cacheMax should be configurable
at servlet init
Hi,
It's an interesting patch and an interesting idea in general. I'm
impressed with the performance findings, although I haven't seen
anyone recently complaining about many concurrent directory listings
requests.
I have a couple of requests for the patch, if you don't mind some
feedback. I'd pr
