eg:
* one process, one listen thread, multiple dispatch threads?
* n processes, one listen FD per process, all listening on the same IP:port?
* one process, each thread listening on the same IP:port?
* something else?
Thanks,
-adrian
___
freebsd-net@
Sweet, have any code you can share that can elicit this?
I'm writing a network / disk application layer right now so i can
model/test this stuff out. I'd love to see some pointers/example code.
Thanks!
-adrian
___
freebsd-net@freebsd.org mailing list
On Tue, Jul 23, 2013 at 8:44 AM, Zaphod Beeblebrox wrote:
> What to do when you don't trust the interface? VMWare is obviously
> emulating the hardware and their interpretation of what the hardware "is"
> is possibly different from ours.
>
> If I boot single-user and tcpdump the interface, I see
yep, FreeBSD 9.1-RELEASE-p3
On Mon Jul 22 21:05:55 2013, Adrian Chadd wrote:
On 22 July 2013 14:26, trafdev wrote:
Actually overhead is almost zero, the real problem is in non-equivalent load
distribution between processes.
As https://lwn.net/Articles/542629/ mentions -
"At Google, they have s
What to do when you don't trust the interface? VMWare is obviously
emulating the hardware and their interpretation of what the hardware "is"
is possibly different from ours.
If I boot single-user and tcpdump the interface, I see two transmitted
solicitations. The kernel claims to have sent one.
On 22 July 2013 14:26, trafdev wrote:
> Actually overhead is almost zero, the real problem is in non-equivalent load
> distribution between processes.
> As https://lwn.net/Articles/542629/ mentions -
> "At Google, they have seen a factor-of-three difference between the thread
> accepting the most
Hi,
This patch adds and hw.ixgbe tree and adds rxd/txd. These are already
tunables but it wasn't easy to see if these were being set.
I'd like to commit this soon.
Thanks!
-adrian
adrian@freebsd-10-hack2:~/work/freebsd/head/src/sys/dev/ixgbe % svn diff .
Index: ixgbe.c
===
Actually overhead is almost zero, the real problem is in non-equivalent
load distribution between processes.
As https://lwn.net/Articles/542629/ mentions -
"At Google, they have seen a factor-of-three difference between the
thread accepting the most connections and the thread accepting the
fewe
trafdev wrote this message on Mon, Jul 15, 2013 at 13:04 -0700:
> Yep I think it's wasting of resources, poll manager should somehow be
> configured to update only one process/thread.
> Anyone know how to do that?
This isn't currently possible w/o a shared kqueue, since the event is
level trigger
The following reply was made to PR kern/180430; it has been noted by GNATS.
From: John Baldwin
To: Meny Yossefi
Cc: "bug-follo...@freebsd.org"
Subject: Re: kern/180430: [ofed] [patch] Bad UDP checksum calc for fragmented
packets
Date: Mon, 22 Jul 2013 11:40:08 -0400
On Monday, July 22, 2013
The following reply was made to PR kern/180430; it has been noted by GNATS.
From: Meny Yossefi
To: John Baldwin , "bug-follo...@freebsd.org"
Cc:
Subject: RE: kern/180430: [ofed] [patch] Bad UDP checksum calc for
fragmented packets
Date: Mon, 22 Jul 2013 14:11:51 +
--_000_F2E47A3
The following reply was made to PR kern/179429; it has been noted by GNATS.
From: Tiago Daniel Jacobs
To: bug-follo...@freebsd.org, o...@tdj.cc
Cc:
Subject: Re: kern/179429: [tap] STP enabled tap bridge
Date: Mon, 22 Jul 2013 13:07:56 -0300
It's a bug because STP don't work as expected when u
Note: to view an individual PR, use:
http://www.freebsd.org/cgi/query-pr.cgi?pr=(number).
The following is a listing of current problems submitted by FreeBSD users.
These represent problem reports covering all versions including
experimental development code and obsolete releases.
S Tracker
13 matches
Mail list logo
