(no subject)

[tingwei liu]

[-- Attachment #1: Type: text/plain, Size: 241 bytes --]

Hi, dears:

      On Suse 11 sp2 3.0.13 kernel. After a few days, Nginx reply
packets with src ip 0.0.0.0 and  dst ip 0.0.0.0。
      Just like https://bbs.archlinux.org/viewtopic.php?id=129304. Is
there a bug?


Thanks very much!

[-- Attachment #2: 2.png --]
[-- Type: image/png, Size: 44507 bytes --]

Re:

[Jiri Slaby]

On 06/20/2013 02:28 PM, tingwei liu wrote:
> Hi, dears:
> 
>       On Suse 11 sp2 3.0.13 kernel. After a few days, Nginx reply
> packets with src ip 0.0.0.0 and  dst ip 0.0.0.0。
>       Just like https://bbs.archlinux.org/viewtopic.php?id=129304. Is
> there a bug?

You should better report this via suse's standard channels (suse
support/bugzilla). We cannot help you otherwise.

thanks,
-- 
js
suse labs

Re:

[tingwei liu]

On Thu, Jun 20, 2013 at 8:51 PM, Jiri Slaby <jslaby@suse.cz> wrote:
> On 06/20/2013 02:28 PM, tingwei liu wrote:
>> Hi, dears:
>>
>>       On Suse 11 sp2 3.0.13 kernel. After a few days, Nginx reply
>> packets with src ip 0.0.0.0 and  dst ip 0.0.0.0。
>>       Just like https://bbs.archlinux.org/viewtopic.php?id=129304. Is
>> there a bug?
>
> You should better report this via suse's standard channels (suse
> support/bugzilla). We cannot help you otherwise.
>
> thanks,
> --
> js
> suse labs

The problem has been fixed. It is a known bug has fixed by Eric Dumazet.

commit dfd25ffffc132c00070eed64200e8950da5d7e9d
Author: Eric Dumazet <eric.dumazet@gmail.com>
Date:   Sat Mar 10 09:20:21 2012 +0000

    tcp: fix syncookie regression

    commit ea4fc0d619 (ipv4: Don't use rt->rt_{src,dst} in ip_queue_xmit())
    added a serious regression on synflood handling.

    Simon Kirby discovered a successful connection was delayed by 20 seconds
    before being responsive.

    In my tests, I discovered that xmit frames were lost, and needed ~4
    retransmits and a socket dst rebuild before being really sent.

    In case of syncookie initiated connection, we use a different path to
    initialize the socket dst, and inet->cork.fl.u.ip4 is left cleared.

    As ip_queue_xmit() now depends on inet flow being setup, fix this by
    copying the temp flowi4 we use in cookie_v4_check().

    Reported-by: Simon Kirby <sim@netnation.com>
    Bisected-by: Simon Kirby <sim@netnation.com>
    Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
    Tested-by: Eric Dumazet <eric.dumazet@gmail.com>
    Signed-off-by: David S. Miller <davem@davemloft.net>

Re:

[Jiri Slaby]

On 06/24/2013 03:43 AM, tingwei liu wrote:
> The problem has been fixed. It is a known bug has fixed by Eric Dumazet.
> 
> commit dfd25ffffc132c00070eed64200e8950da5d7e9d
> Author: Eric Dumazet <eric.dumazet@gmail.com>
> Date:   Sat Mar 10 09:20:21 2012 +0000

Which is a part of 3.0.26, pushed into the SLE kernel in Apr 2012...

-- 
js
suse labs

Re:

[tingwei liu]

Hi， dears

      I have found many RetransSegs of tcp by /proc/net/snmp, but I
can't capture these packet on local system by tcpdump.
      What could be the reason? Droped by qdisc level or device
driver? How to confirm that?

      I also find patch "[PATCH] tcp: reflect SYN queue_mapping into
SYNACK packets" provided by Eirc, how to confirm this?

 Thanks very much!