[
Date Prev][
Date Next][
Thread Prev][
Thread Next][
Date Index][
Thread Index]
[
List Home]
|
Re: [jetty-dev] NIO inefficiency?
|
I'm reading the source of "DeflaterOutputStream" and I'm not entirely
convinced you're right about. this. A little deeper of the dump would
easily prove that I am wrong ;)
Kristian
2013/8/9 Viktor Szathmary <phraktle@xxxxxxxxx>:
> Hi,
>
> On Friday, August 9, 2013 at 12:51 PM, Kristian Rosenvold wrote:
>
> Your profiler dump gives a hint:
>
> 2.4% - org.eclipse.jetty.server.HttpChannel.write
> 1.5% - org.eclipse.jetty.server.HttpOutput.flush
> 1.3% - java.util.zip.DeflaterOutputStream.flush
> ...
>
> If I'm not entirely mistaken, the DeflaterOutputStream does
> on-the-fly compression of data, in which case you might be looking at
> the overhead from the "Deflate" compression algorithm ? If you look
> further down the graph I wouldn't be surprised if that's seeping
> through on several of the methods ?
>
> The overhead of compression is clearly on a different path (which is taking
> ~15%), so that's not it.
>
> (Btw, I had a prior contribution for speeding up gzip compression by reusing
> Deflaters: https://bugs.eclipse.org/bugs/show_bug.cgi?id=402885).
>
> Regards,
> Viktor
>
> _______________________________________________
> jetty-dev mailing list
> jetty-dev@xxxxxxxxxxx
> https://dev.eclipse.org/mailman/listinfo/jetty-dev
>
