Re: [PATCH net-next] bonding: 3ad: bonding of links with different data rate

[Steven Hsieh]

Thanks Jay for reviewing patch request.
We decided not to move forward.
Instead of adding different speed links to one aggregator,
we can assign them to multiple aggregators and use ad_select=bandwidth
to pick highest bandwidth aggregator.

Thanks
Steven H.

> From: Jay Vosburgh <jay.vosburgh@xxxxxxxxxxxxx>
> To: Steven Hsieh <steven.hsieh@xxxxxxxxxxxx>
> Cc: Andy Gospodarek <andy@xxxxxxxxxxxxx>,
> 	"David S. Miller" <davem@xxxxxxxxxxxxx>,
> 	Eric Dumazet <edumazet@xxxxxxxxxx>,
> 	Jakub Kicinski <kuba@xxxxxxxxxx>, Paolo Abeni <pabeni@xxxxxxxxxx>,
> 	Veaceslav Falico <vfalico@xxxxxxxxx>,
> 	linux-kernel@xxxxxxxxxxxxxxx, netdev@xxxxxxxxxxxxxxx
> Subject: Re: [PATCH net-next] bonding: 3ad: bonding of links with different data rate
> Date: Mon, 24 Oct 2022 09:25:09 -0700	[thread overview]
> Message-ID: <15633.1666628709@famine> (raw)
> In-Reply-To: <20221022220158.74933-1-steven.hsieh@xxxxxxxxxxxx>
> 
> Steven Hsieh <steven.hsieh@xxxxxxxxxxxx> wrote:
> 
> >Current Linux Bonding driver supports IEEE802.3ad-2000.
> >Operation across multiple data rates—
> >All links in a Link Aggregation Group operate at the same data rate.
> >
> >In IEEE802.1AX-2014
> >Aggregation of links of different data rates is not prohibited
> >nor required by this standard.
> 
> 	The -2014 and -2020 versions change a lot of things at once; I'm
> not sure we can just cherry pick out one thing (or maybe we can, I'm
> reading through the changes).  Notably, the -2020 version states, in
> reference to changes added at -2014,
> 
> "[...] it explicitly allowed the aggregation of point-to-point links of
> any speed using any physical media or logical connection capable of
> supporting the Internal Sublayer Service specified in IEEE Std
> 802.1AC."
> 
> 	whereas the -2008 standard specifies "CSMA/CD MACs" instead of
> the ISS from 802.1AC.  I'm not yet sure if this makes any relevant
> difference.

> >This patch provides configuration option to allow aggregation of links
> >with different speed.
> 
> 	Have you tested all of the edge cases?  E.g., what is the
> behavior with and without the option enabled when an interface in an
> aggregator changes its speed?
> 
> 	If you have tests, consider including test scripts in
> tools/testing/selftests/drivers/net/bonding/
> 

In current code, when 2nd port is linked up with different speed,
it will not be part of active aggregator.

> >Enhancement is disabled by default and can be enabled thru
> > echo 1 > /sys/class/net/bond*/bonding/async_linkspeed
> 
> 	New option settings like this require (a) support in iproute2
> (to set/get the option like any other bonding option), and (b) updates
> to the documentation (Documentation/networking/bonding.rst).
> 
> 	I'm not completely sold on the name, either, "async" doesn't
> really describe "differing data rates" in my mind.  Perhaps an option
> named "ad_link_speed" with allowed values of "same" or "any"?



-- 
This electronic communication and the information and any files transmitted 
with it, or attached to it, are confidential and are intended solely for 
the use of the individual or entity to whom it is addressed and may contain 
information that is confidential, legally privileged, protected by privacy 
laws, or otherwise restricted from disclosure to anyone else. If you are 
not the intended recipient or the person responsible for delivering the 
e-mail to the intended recipient, you are hereby notified that any use, 
copying, distributing, dissemination, forwarding, printing, or copying of 
this e-mail is strictly prohibited. If you received this e-mail in error, 
please return the e-mail to the sender, delete it from your computer, and 
destroy any printed copy of it.

Attachment: smime.p7s
Description: S/MIME Cryptographic Signature