[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <4bccd9bf-4858-4c9f-aac1-d25a8083ac09@oracle.com>
Date: Wed, 27 Jun 2018 13:45:13 -0700
From: Santosh Shilimkar <santosh.shilimkar@...cle.com>
To: Ka-Cheong Poon <ka-cheong.poon@...cle.com>,
Sowmini Varadhan <sowmini.varadhan@...cle.com>
Cc: netdev@...r.kernel.org, davem@...emloft.net,
rds-devel@....oracle.com
Subject: Re: [PATCH net-next 2/3] rds: Enable RDS IPv6 support
On 6/27/2018 3:07 AM, Ka-Cheong Poon wrote:
> On 06/26/2018 09:08 PM, Sowmini Varadhan wrote:
>> On (06/26/18 21:02), Ka-Cheong Poon wrote:
[...]
>
>
> I don't expect RDS apps will want to use link local address
> in the first place. In fact, most normal network apps don't.
>
This is not true.
>
>> You're not doing this for IPv4 and RDS today (you dont have to do this
>> for UDP, afaik)
>
>
> Do you know of any IPv4 RDS app which uses IPv4 link local
> address? In fact, IPv4 link local address is explicitly
> disallowed for active active bonding.
>
Yes. Cluster-ware HAIP makes use of link local addresses. That
check was mainly because of RDMA CM issues but that only means
active-active isn't used. The bonding works just fine and if
needed cluster-ware can also use TCP transport.
Lets not add this new behavior for link local and its
actually not relevant to really v6 addressing support.
Regards,
Santosh
Powered by blists - more mailing lists