Re: bonding.c and proc_fs

2000-10-14 Thread Jorge Nerin

Willy TARREAU wrote:
> 
> Hello !
> 
> I think that you should wait a bit before writing a config in /proc for the
> bonding driver.
> I have rewritten quite a part of it to support link detection and make it a bit
> fail safe.
> Moreover, I had to rewrite partly ifenslave.c (which is included in the same
> patch). Everything
> *seems* to work fine (even in SMP now) and I've submitted this to Thomas Davis
> after
> Constantine Gavrilov has done some tests. If you're interested in, I can send
> you the currently
> working patch (with longer doc). It may change the way you planned your
> configuration in /proc.
> 
> Regards,
> 
> Willy

Sorry for the AOLism, but... Me too, I also want to give a try to the
patch.

I also want to try bonding between two machines, one 486 and one dual
200mmx, and until now without luck, I only managed to lockup the dual
machine by doing ifconfig up & down and insmod & rmmod, mostly because
the lack of documentation.

Thanks.

-- 
Jorge Nerin
<[EMAIL PROTECTED]>
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [EMAIL PROTECTED]
Please read the FAQ at http://www.tux.org/lkml/



Re: bonding.c and proc_fs

2000-10-11 Thread Willy TARREAU

Hello !

I think that you should wait a bit before writing a config in /proc for the
bonding driver.
I have rewritten quite a part of it to support link detection and make it a bit
fail safe.
Moreover, I had to rewrite partly ifenslave.c (which is included in the same
patch). Everything
*seems* to work fine (even in SMP now) and I've submitted this to Thomas Davis
after
Constantine Gavrilov has done some tests. If you're interested in, I can send
you the currently
working patch (with longer doc). It may change the way you planned your
configuration in /proc.

Regards,

Willy
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [EMAIL PROTECTED]
Please read the FAQ at http://www.tux.org/lkml/



bonding.c and proc_fs

2000-10-11 Thread Sudhindra Herle

Hello,

I am writing a /proc interface to the ethernet bonding driver and had the
following questions:

1) Why does a slave interface need an IP address? Isn't it sufficient if
slave->flags is set to IFF_UP|IFF_RUNNING ?

2) What happens if the MTU of differnet slave devices are different? How
does one choose the MTU of the master? Minimum of the MTU of all slaves?

3) Where in /proc would one put the bonding driver? /proc/net/bondNNN? 

TIA,
Sudhi Herle
Staff Engineer
Samsung Telecom America
[EMAIL PROTECTED]


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [EMAIL PROTECTED]
Please read the FAQ at http://www.tux.org/lkml/



bonding.c and proc_fs

2000-10-11 Thread Sudhindra Herle

Hello,

I am writing a /proc interface to the ethernet bonding driver and had the
following questions:

1) Why does a slave interface need an IP address? Isn't it sufficient if
slave-flags is set to IFF_UP|IFF_RUNNING ?

2) What happens if the MTU of differnet slave devices are different? How
does one choose the MTU of the master? Minimum of the MTU of all slaves?

3) Where in /proc would one put the bonding driver? /proc/net/bondNNN? 

TIA,
Sudhi Herle
Staff Engineer
Samsung Telecom America
[EMAIL PROTECTED]


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [EMAIL PROTECTED]
Please read the FAQ at http://www.tux.org/lkml/



Re: bonding.c and proc_fs

2000-10-11 Thread Willy TARREAU

Hello !

I think that you should wait a bit before writing a config in /proc for the
bonding driver.
I have rewritten quite a part of it to support link detection and make it a bit
fail safe.
Moreover, I had to rewrite partly ifenslave.c (which is included in the same
patch). Everything
*seems* to work fine (even in SMP now) and I've submitted this to Thomas Davis
after
Constantine Gavrilov has done some tests. If you're interested in, I can send
you the currently
working patch (with longer doc). It may change the way you planned your
configuration in /proc.

Regards,

Willy
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [EMAIL PROTECTED]
Please read the FAQ at http://www.tux.org/lkml/