FC4 possible kernel 2.6.15 issue

Skunk Worx skunkworx at verizon.net
Thu Mar 16 15:42:02 UTC 2006


Dave Jones wrote:
> On Wed, Mar 15, 2006 at 07:32:32PM -0800, Skunk Worx wrote:
>  > I've also verified the problem using the stock kernel.org releases on my 
>  > FC5T3 box.
>  > 
>  > We might be getting closer to the issue. Right now we're trying to 
>  > decide whether or not to write a simple client/server pair of apps that 
>  > demonstrate the problem.
> 
> If you can construct a test case, the upstream developers at
> netdev at vger.kernel.org will be a better target than fedora-test-list.
> 
> 		Dave
> 

Unfortunately, it's not a bug, it's a feature :

"This came up with java debugging already. The problem is when the 
sender writes a message in separate write() system calls, each one 
becomes a separate packet. In 2.6.15 we do a new thing called 
Appropriate Byte Count and that penalizes stupid applications, but 
provides better fairness over the internet by accounting for packets 
better."

I've sent this comment over to the "stupid application" developers at 
GSOAP. I guess they (and I) unreasonably expect lo to be at least as 
fast, or faster, than ether...

---
John




More information about the fedora-test-list mailing list