On Sat, Aug 24, 2013 at 3:00 PM, Chris Cappuccio wrote:
The errors you got are due to errors reported by the IDE controller
back to the pciide driver. That's a layer (or two) lower than the
filesystem itself. If moving the drive around fixed it, perhaps there
was a poor connection somewhere? (Hell, I just bought an old stereo
receiver and had to take it apart and reseat every board inside to
clean up the sound coming out...)
Since the dc driver doesn't know how to read the MAC addresses from
your system, your best bet to make those work is to use the 'lladdr'
flag so that something other than 0 is used on the wire...
It looks like I have an intermittent connection or a controller that's
going bad. When I try to get networking to work, the system hangs at
starting the network. Some times it completes and I get a login prompt,
other times I don't. At times it gives the same errors as I posted
before. I'll experiment with the other units I have to see if the
problem follows the hard drive.
I noticed both the IP350 and IP380 have a MAC address label on the
system board. I suspect this is the starting address used to assigned
addresses to the fxp and dc interfaces. Where that address is stored is
a good question. Odd that the fxp interfaces have FF and the dc
interfaces have 00.
I was sent a link about an Intel utility called eeupdate.exe that can be
used to write the MAC address to the eeprom. But, it looks like that
utility is an OEM download from Intel. It looks like one can put lladdr
in the hostnam.if file and set the address that way.
The 100/1000 card looks like it has the addresses onboard. I haven't
tried copper SFPs to see if they will work in place of the fiber SFPs.
Would be nice if they do.
Thanks again to everyone for the help with this.
Stan