On Thu, 2013-04-18 at 15:09 -0400, David Miller wrote:
> Applied and queued up for -stable, but:
>
> > + sizeof(_l4), &_l4);
> >
>
>
From: Eric Dumazet
Date: Mon, 15 Apr 2013 20:03:24 -0700
> [PATCH] bonding: fix l23 and l34 load balancing in forwarding path
>
> Since commit 6b923cb7188d46 (bonding: support for IPv6 transmit hashing)
> bonding doesn't properly hash traffic in forwarding setups.
>
> Vitaly V. Bursov diagnosed
From: Eric Dumazet
Date: Tue, 16 Apr 2013 07:25:17 -0700
> On Tue, 2013-04-16 at 06:51 -0700, Eric Dumazet wrote:
>
>> Perfect, thanks a lot for all this !
>>
>> Tested-by: Vitaly V. Bursov
>>
>>
>
> By the way, we probably should use skb_flow_dissect() to get proper
> hashing for tunnels u
On Tue, 2013-04-16 at 06:51 -0700, Eric Dumazet wrote:
> Perfect, thanks a lot for all this !
>
> Tested-by: Vitaly V. Bursov
>
>
By the way, we probably should use skb_flow_dissect() to get proper
hashing for tunnels users.
--
To unsubscribe from this list: send the line "unsubscribe linu
On Tue, 2013-04-16 at 12:01 +0300, Vitaly V. Bursov wrote:
> Testing under real load for almost 2 hours now, works as expected.
> xmit_hash_policy=layer3+4
>
> I made a few simple test with layer2+3 policy too, looks OK.
>
> Thanks!
Perfect, thanks a lot for all this !
Tested-by: Vitaly V. Bur
16.04.2013 06:03, Eric Dumazet пишет:
From: Eric Dumazet
On Mon, 2013-04-15 at 17:37 -0700, Eric Dumazet wrote:
On Mon, 2013-04-15 at 16:57 +0300, Vitaly V. Bursov wrote:
Hello,
I have a bonding device (mode=802.3ad xmit_hash_policy=layer2+3 miimon=300) and
for kernels <3.7 forwarded IPv4 tr
From: Eric Dumazet
On Mon, 2013-04-15 at 17:37 -0700, Eric Dumazet wrote:
> On Mon, 2013-04-15 at 16:57 +0300, Vitaly V. Bursov wrote:
> > Hello,
> >
> > I have a bonding device (mode=802.3ad xmit_hash_policy=layer2+3 miimon=300)
> > and
> > for kernels <3.7 forwarded IPv4 traffic distributed f
On Mon, 2013-04-15 at 16:57 +0300, Vitaly V. Bursov wrote:
> Hello,
>
> I have a bonding device (mode=802.3ad xmit_hash_policy=layer2+3 miimon=300)
> and
> for kernels <3.7 forwarded IPv4 traffic distributed fine across multiple
> physical
> links. Ethernet cards are Intel 82576 with igb driver
Hello,
I have a bonding device (mode=802.3ad xmit_hash_policy=layer2+3 miimon=300) and
for kernels <3.7 forwarded IPv4 traffic distributed fine across multiple
physical
links. Ethernet cards are Intel 82576 with igb driver (various versions).
3.7 and 3.8 kernels tend to fully utilize only one l
9 matches
Mail list logo
