[ovs-dev] [PATCH/RFC] ofproto-dpif: Make bonding balance test more robust

[Andy Zhou]

The initial bond distribution indeed improved significantly with
e58f91a1.  Thanks for solving one of the mysteries.

I have already pushed Simon's patch assuming test failure was still
observed with this patch. Using 256 is not a bad idea at any rate --
there is no particular good reason to pick 91 to begin with.  We
should however raise the minimally expected distribution to catch a
bad random selection, such as the one that got fix.  I will send out a
patch soon once I figure out what will be a good value.



On Thu, May 1, 2014 at 9:11 PM, Simon Horman <horms at verge.net.au> wrote:
> On Thu, May 01, 2014 at 06:46:08PM -0700, Andy Zhou wrote:
>> On Thu, May 1, 2014 at 6:23 PM, YAMAMOTO Takashi <yamamoto at valinux.co.jp> wrote:
>> >> It is my observation that when sending 91 packets to a mode=balance
>> >> bond interface with three underlying ports in some cases not all ports
>> >> receive at least 7 packets. This causes the test to fail.
>> >
>> > was it with or without commit e58f91a1?
>>
>> Good question, I will look into it and report back.
>
> Thanks.
> I have not been able to reproduce the problem using e58f91a1.
> I think we can drop my patch.
>
>> >
>> > YAMAMOTO Takashi
>> >
>> >>
>> >> It may be that failure indicates that the bonding code
>> >> needs improvement. But it seems to me that its more likely to
>> >> be a valid behaviour.
>> >>
>> >> I have found that by sending 256 packets instead of 91 the test passes
>> >> more reliably: I am yet to see it fail.
>> >>
>> >> Signed-off-by: Simon Horman <horms at verge.net.au>
>> >> ---
>> >>  tests/ofproto-dpif.at | 2 +-
>> >>  1 file changed, 1 insertion(+), 1 deletion(-)
>> >>
>> >> diff --git a/tests/ofproto-dpif.at b/tests/ofproto-dpif.at
>> >> index a32bc41..c46e997 100644
>> >> --- a/tests/ofproto-dpif.at
>> >> +++ b/tests/ofproto-dpif.at
>> >> @@ -178,7 +178,7 @@ ovs-appctl time/warp 100
>> >>  ovs-appctl lacp/show > lacp.txt
>> >>  ovs-appctl bond/show > bond.txt
>> >>  (
>> >> -for i in `seq 10 100` ;
>> >> +for i in `seq 0 255` ;
>> >>      do
>> >>      pkt="in_port(7),eth(src=50:54:00:00:00:05,dst=50:54:00:00:01:00),eth_type(0x0800),ipv4(src=10.0.0.2,dst=10.0.0.1,proto=6,tos=0,ttl=64,frag=no),tcp(src=8,dst=$i),tcp_flags(0x010)"
>> >>      AT_CHECK([ovs-appctl netdev-dummy/receive p7 $pkt])
>> >> --
>> >> 1.8.5.2
>> >>
>> >> _______________________________________________
>> >> dev mailing list
>> >> dev at openvswitch.org
>> >> http://openvswitch.org/mailman/listinfo/dev
>>