<br><font size=2 face="sans-serif">Hi,</font>
<br>
<br><font size=2 face="sans-serif">i am sorry if i miss the point, but
i did not read the older messages.</font>
<br>
<br><font size=2 face="sans-serif">But why not use 802.3ad aka dynamic
link aggregation, we have that successfully configured with HP and Nortel
switches ( also with machines running drbd over those links ).</font>
<br>
<br><font size=2 face="sans-serif">As far as i know cisco switches also
support 802.3ad.</font>
<br>
<br><font size=2 face="sans-serif">Regards</font>
<br><font size=2 face="sans-serif">Soeren</font>
<br>
<br>
<br>
<br>
<br>
<table width=100%>
<tr valign=top>
<td width=40%><font size=1 face="sans-serif"><b>Ralf Gross <Ralf-Lists@ralfgross.de></b>
</font>
<br><font size=1 face="sans-serif">Sent by: drbd-user-bounces@lists.linbit.com</font>
<p><font size=1 face="sans-serif">19.06.2008 17:49</font>
<td width=59%>
<table width=100%>
<tr valign=top>
<td>
<div align=right><font size=1 face="sans-serif">To</font></div>
<td><font size=1 face="sans-serif">drbd-user@lists.linbit.com</font>
<tr valign=top>
<td>
<div align=right><font size=1 face="sans-serif">cc</font></div>
<td>
<tr valign=top>
<td>
<div align=right><font size=1 face="sans-serif">Subject</font></div>
<td><font size=1 face="sans-serif">Re: [DRBD-user] 10Gb ethernet ?</font></table>
<br>
<table>
<tr valign=top>
<td>
<td></table>
<br></table>
<br>
<br>
<br><tt><font size=2>Lars Ellenberg schrieb:<br>
> On Thu, Jun 19, 2008 at 01:03:11PM +0100, Lee Christie wrote:<br>
> > In any event, I'm no expert on channel bonding, but in a 2-server<br>
> > configuration, where the Ips and MAC addresses are fixed at either
end,<br>
> > how can you use all 4 channels ? I was always under the impression
that<br>
> > the bonding used an algorithm based on src/dest IP/Mac to choose
which<br>
> > link to send data down, so in a point to point config it would
always be<br>
> > the same link. <br>
> <br>
> "balance-rr" aka mode 0 for linux bonding schedules packets
round robin<br>
> over the available links.<br>
<br>
balance-rr will not help in a cisco environment, because the switch<br>
will still use the same ports and does not perform rr load balancing.<br>
The only way to get rr working with cisco switches was to use 2<br>
different vlans.<br>
<br>
eth0 <---- vlan x ----> eth0<br>
eth1 <---- vlan y ----> eth1<br>
<br>
At least this is my experience and our CCNP's (or what they are<br>
called) told me the same.<br>
<br>
With the above trick I was able to get ~1.6x GbE throughput with<br>
the netpipe benchmark (after tuning the reorder kernel parameter). I<br>
didn't used the conenction for drbd, I tried to speed up our backup.<br>
But the funny thing was, it slowed down. Even the ftpd or samba was<br>
slower over this link as it was with the xor mode or just one GbE NIC.<br>
<br>
> but still, for a single tcp connection, given some tcp_reorder tuning,<br>
> the strong gain you get from 2x 1GbE (1.6 to 1.8 * that of one channel)<br>
> degrades again to effectively less than one channel if you try to
use 4x.<br>
> <br>
> again, "more" is not always "better".<br>
> for the usage pattern of drbd (single tcp connection with bulk data)
the<br>
> throughput-optimum linux bonding seems to be 2x, with 3x you are back
to<br>
> around the same throughput as 1x, with 4x you are even worse than
1x,<br>
> because packet reordering over bonded GbE and tcp congestion control<br>
> don't work well together for sinlge tcp links.<br>
<br>
Very true.<br>
<br>
Ralf<br>
_______________________________________________<br>
drbd-user mailing list<br>
drbd-user@lists.linbit.com<br>
http://lists.linbit.com/mailman/listinfo/drbd-user<br>
</font></tt>
<br>