Matt Welsh wrote:
> Exactly the topic of my research :-) You should check out Jaguar,
> http://www.cs.berkeley.edu/~mdw/proj/jaguar
OK, I now link to it from http://www.kegel.com/c10k.html#java
> We are also working on building large-scale Internet services in Java.
> Many of the problems you talk about on your 'c10k' web page (which is
> very good, by the way) are things we are addressing.
Thanks. The table of contents I just added ought to make it more readable.
> You also might want to read a paper we wrote recently about engineering
> systems for high throughput. What we have in mind here is Java,
> but it applies to any language and O/S:
> http://www.cs.berkeley.edu/~mdw/papers/events.pdf
Looks like a good read. I now link to it.
(Wish my copy of Ghostscript didn't choke on half the pages!)
> I'm also on the Expert Group for JSR 51 (the new I/O APIs for the Java
> platform), so hopefully good things will happen there!
Yes, it should be interesting!
> My personal feeling is that there is a lot that will need to happen at
> both the Java and the O/S level to get great I/O performance. I am not
> sure I agree with many of the discussions in the linux-kernel list
> that the right way to get high I/O bandwidth is just to use some
> bastardization of signals; I think that the folks at Rice are a lot
> closer to the mark with their novel event-delivery mechanism (by
> this I mean http://www.cs.rice.edu/~druschel/usenix99event.ps.gz).
The folks at the Linux Scalability Project seem to agree that the
realtime signal approach isn't the top contender at the moment.
The /dev/poll they implemented for Linux beats the realtime signal
approach in their tests so far.
See http://www.citi.umich.edu/techreports/reports/citi-tr-00-4.pdf
> So ... it's a very interesting space to work in right now. Demanding
> I/O applications place a lot of new demands on Java (and operating systems
> as well). I would be interested in having more discussion with people on
> this list about their experiences!
Ditto. And if anyone interested in this happens to be in LA,
drop in on the lajug javaklatch at Equator in old Pasadena at 1pm this
Saturday and let's talk Java over java :-)
- Dan
--
Entia non sunt multiplicanda praeter necessitatem.
----------------------------------------------------------------------
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
