[lug] forcing certain services to use eth1 instead of eth0

[John Hernandez]

I see.  I wonder if something else in the IO subsystem was overburdened,
rather than the NIC itself?  1 x Gig-E is a LOT of bandwidth.  Have you
tried tracking (or graphing) the interface utilization over time, such
as with MRTG or similar?

Michael Belanger wrote:
> On Wednesday 04 January 2006 15:18, John Hernandez wrote:
> 
>>It might also be helpful to re-evaluate WHY you want to do this.  Did
>>you diagnose a specific performance bottleneck?  Is it because you are
>>periodically saturating a 100Mbps interface?  Are you running full
>>duplex on a switched network?  Maybe Gig-E is the answer?  Or maybe
>>802.3ad link aggregation?  Or a faster server?
> 
> 
> Well.. It is a dual-gigabit machine that we were only using a single interface 
> on.  There have been instances where other services have disrupted email -- 
> specifically nfs.  

-- 

 |  John Hernandez - NOAA Boulder NOC - 303-497-6392
 |  Mailstop R/OM62. 325 Broadway, Boulder, CO 80305
 |  PGP Public Key ID: 586A7E23