Re: Bonding support for eth1394?

Re: Bonding support for eth1394?

[Stefan Richter]

(Adding Cc: netdev)

Karl Svec wrote to linux1394-devel:
> I'm trying to create a bonded network interface for an Xdmx setup out  
> of a regular ethernet device (sungem) and a firewire ethernet  
> (eth1394) device. However, ifenslave fails on the firewire device,  
> and gives me the following output:
> 
> Master 'bond0' : Error : SIOCBONDENSLAVE failed : Operation not  
> supported
> 
> I'm assuming that this error message means that bonding support is  
> not implemented in the eth1394 driver. If this is the case, I'd like  
> to request that bonding support be added to the eth1394 driver. Is  
> this a reasonable request (i.e. is this something that was left out  
> of the driver because it is difficult to do, or was it left out  
> because nobody anticipated anyone using this feature)?
> 
> I'm happy to do any testing.
> 
> Thanks for your time,

What criteria has a networking driver to fulfill to be interoperable
with the bonding driver?  From a brief look at the top of the files in
linux/drivers/net/bonding/, it appears that it is meant for Ethernet
drivers.

Unlike eth1394's name suggests, it is not Ethernet over FireWire but
IPv4 over FireWire, in a partial implementation of RFC 2734.
http://www.faqs.org/rfcs/rfc2734.html
What is missing to fully support RFC 2734 is multicast capability.
http://lxr.linux.no/source/drivers/ieee1394/eth1394.c#L25

The bonding sources have a few occurrences of EOPNOTSUPP.  Unless I
missed something, they are all related to setting the hardware address
of the interface.  AFAICS this is impossible with IP over FireWire.  If
it is crucial to bonding to be able to change the slaves' hardware
addresses, then you are out of luck.
-- 
Stefan Richter
-=====-=-=== =-=- -==-=
http://arcgraph.de/sr/

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/

Re: Bonding support for eth1394?

[Roland Dreier]

 > The bonding sources have a few occurrences of EOPNOTSUPP.  Unless I
 > missed something, they are all related to setting the hardware address
 > of the interface.  AFAICS this is impossible with IP over FireWire.  If
 > it is crucial to bonding to be able to change the slaves' hardware
 > addresses, then you are out of luck.

There are a few changes to the bonding driver pending that will add
support for bonding IP-over-InfiniBand interfaces.  IPoIB also cannot
change its HW address, so the patches address that issue.

Once those patches land, bonding eth1394 interfaces may "just work".

 - R.

Re: Bonding support for eth1394?

[Stefan Richter]

Roland Dreier wrote:
> There are a few changes to the bonding driver pending that will add
> support for bonding IP-over-InfiniBand interfaces.  IPoIB also cannot
> change its HW address, so the patches address that issue.
> 
> Once those patches land, bonding eth1394 interfaces may "just work".

Sounds promising.  I will keep an eye on it.
-- 
Stefan Richter
-=====-=-=== =-=- -==-=
http://arcgraph.de/sr/

Re: Bonding support for eth1394?

[Bill Fink]

On Sat, 13 Oct 2007, Stefan Richter wrote:

> Roland Dreier wrote:
> > There are a few changes to the bonding driver pending that will add
> > support for bonding IP-over-InfiniBand interfaces.  IPoIB also cannot
> > change its HW address, so the patches address that issue.
> > 
> > Once those patches land, bonding eth1394 interfaces may "just work".
> 
> Sounds promising.  I will keep an eye on it.

While that might allow multiple eth1394 interfaces to be bonded,
I believe the user wanted to bond an eth1394 interface with a normal
Ethernet interface, and I don't think that will work even with the
IPoIB bonding changes, since bonding of different fundamental types
of network interfaces still won't be supported, and I'm pretty sure
eth1394 is not considered a standard Ethernet interface (different
MAC address format for one thing).

						-Bill

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/

Re: Bonding support for eth1394?

[Stefan Richter]

Bill Fink wrote:
> I believe the user wanted to bond an eth1394 interface with a normal
> Ethernet interface, and I don't think that will work even with the
> IPoIB bonding changes, since bonding of different fundamental types
> of network interfaces still won't be supported, and I'm pretty sure
> eth1394 is not considered a standard Ethernet interface (different
> MAC address format for one thing).

True, the MAC addresses, frame headers (OSI layer 2?) and so on are
different.  Of course the case which the original poster had in mind,
bonding FireWire with Ethernet, would be at least as interesting as
bonding several FireWire interfaces, since most FireWire equipped
machines come with only one FireWire link and one or two Ethernet links.
-- 
Stefan Richter
-=====-=-=== =-=- -===-
http://arcgraph.de/sr/

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/

Re: Bonding support for eth1394?

[Rick Jones]

Failover between disparate link-types sounds like a job for IP and routing.

rick jones

Re: Bonding support for eth1394?

[Stefan Richter]

Roland Dreier wrote on 2007-10-13:
>  > The bonding sources have a few occurrences of EOPNOTSUPP.  Unless I
>  > missed something, they are all related to setting the hardware address
>  > of the interface.  AFAICS this is impossible with IP over FireWire.  If
>  > it is crucial to bonding to be able to change the slaves' hardware
>  > addresses, then you are out of luck.
> 
> There are a few changes to the bonding driver pending that will add
> support for bonding IP-over-InfiniBand interfaces.  IPoIB also cannot
> change its HW address, so the patches address that issue.
> 
> Once those patches land, bonding eth1394 interfaces may "just work".

Bill Fink wrote on 2007-10-14:
> While that might allow multiple eth1394 interfaces to be bonded,
> I believe the user wanted to bond an eth1394 interface with a normal
> Ethernet interface, and I don't think that will work even with the
> IPoIB bonding changes, since bonding of different fundamental types
> of network interfaces still won't be supported, and I'm pretty sure
> eth1394 is not considered a standard Ethernet interface (different
> MAC address format for one thing).


Karl, you could try kernel 2.6.24(.2) which AFAIU features the mentioned 
changes.
-- 
Stefan Richter
-=====-==--- --=- =--==
http://arcgraph.de/sr/

-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/