Re: Load balancing across multiple network paths

On Thu, 2004-12-09 at 16:43, Peter T. Breuer wrote:

> I suppose the MTU is 1500.

Yes.

> > which surprisingly shows a 2:1 ratio! /proc/nbdinfo says 
> 
> yes, this conforms with what is visible below too.

I now think that the ratio is just random, but I still cannot explain
why the imbalance.

> > Indeed the daemons cannot be alternating... Also the B/s max seems
> > spurious as these are 100Mbit/s NICs.
> 
> The problem is that that measurement is a bit too "spontaneous". The
> time interval over which it is measured can be short.  Many such
> intervals are averaged (and smoothed) to give the average measurement.
> 
> Actually, I see that the measurements are updated at least once
> every second, in theory, and that the raw data is folded into an
> existing average to give a smoothed rolling average speed. As far as I
> can see, the new data gets weighted according to the amount of time
> that has passed since the last measurement, so old data should have a
> half-life of one second. The "maximum" is merely the greatest such
> smoothed average that has been recorded. I suppose when the device
> started up.
> 
> Feel free to improve the calculations (in enbd_speed).
> 
> If you look while the read is happening, over a long run, the "average" 
> speed should be close to the current speed. 

I may have a look at enbd_speed if I can steal the time.

> > Yes, though the real reason I want dual pathing is redundancy (in cards,
> > switches, cables etc) so ethernet binding is not really an option. Oh, I
> > should have said that this will be part of a networked raid5, so I
> > cannot use crossovers, hence the switches.
> 
> Have you tried FR5 (fr5.sf.net)? I am hoping for testers.

No, but it's on my to-do list. As soon as I can iron out issues like the
above I will give it a go.

> You are not writing, so I don't think I could even suppose there is any
> serialisation through preservation of write order!

Indeed.

> And I think that if you do ethernet binding one of the NICs can fail
> without harming the joint device. But I have not studied the issue.

Ethernet bonding does have failover modes. However as I said above, I
really want two subnets on different switches for the networked raid5,
so bonding is out of the question.

> I'm sorry, I don't see any clear evidence here one way or the other.
> How about simply _timing_ a read with one NIC only, and then with two
> NICs operating.

I did as you say and it is weird because when I run enbd-client on only 
one IP (and with only one channel!) it maxes out the corresponding NIC
(8.5 MB/s). And that happens with both NICs. Tricky... 

Nikos




This email has been independently scanned for viruses and any virus detected has been removed using McAfee anti-virus software