[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20180226134610.GB2063@nanopsycho>
Date: Mon, 26 Feb 2018 14:46:10 +0100
From: Jiri Pirko <jiri@...nulli.us>
To: Andrew Lunn <andrew@...n.ch>
Cc: Ido Schimmel <idosch@...lanox.com>, netdev@...r.kernel.org,
linux-kselftest@...r.kernel.org, davem@...emloft.net,
shuah@...nel.org, jiri@...lanox.com, dsahern@...il.com,
roopa@...ulusnetworks.com, nikolay@...ulusnetworks.com,
f.fainelli@...il.com, vivien.didelot@...oirfairelinux.com,
mlxsw@...lanox.com
Subject: Re: [PATCH net-next 00/14] selftests: forwarding: Add VRF-based tests
Mon, Feb 26, 2018 at 02:35:24PM CET, andrew@...n.ch wrote:
>On Mon, Feb 26, 2018 at 10:59:34AM +0200, Ido Schimmel wrote:
>> One of the nice things about network namespaces is that they allow one
>> to easily create and test complex environments.
>>
>> Unfortunately, these namespaces can not be used with actual switching
>> ASICs, as their ports can not be migrated to other network namespaces
>> (NETIF_F_NETNS_LOCAL) and most of them probably do not support the
>> L1-separation provided by namespaces.
>>
>> However, a similar kind of flexibility can be achieved by using VRFs and
>> by looping the switch ports together. For example:
>>
>> br0
>> +
>> vrf-h1 | vrf-h2
>> + +---+----+ +
>> | | | |
>> 192.0.2.1/24 + + + + 192.0.2.2/24
>> swp1 swp2 swp3 swp4
>> + + + +
>> | | | |
>> +--------+ +--------+
>
>Hi Ido
>
>So you decided to stick to a bridge of just two interfaces. You
>clearly can write tests using this, but it seems very limiting. You
>cannot test bridging vs flooding. You cannot tests packets don't leak
>between parallel bridges. It i hard to test igmp snooping, etc.
>
>It seems to me, using 8 ports, not 4, would give you a much more
>flexible and interesting test setup.
You can add tests that would use how many ports you like. It is up to
the individual test how many ports it needs. Top of this patchset is 4.
>
> Andrew
Powered by blists - more mailing lists