A customer of mine is running version 4.23 on a 12 node network. Recently I
have been having a problem doing an ls -l on one of the nodes from any other
node. The ls command runs very slowly, and then only displays a few
directories. However, if I do a listing from the “bad” node itself,
everything shows up just fine.
I also cannot always copy a file FROM the bad node. If I try to cp a file
from this node to another node, cp reports “no route to host”, even though I
can see all the other nodes and their directories from the bad node. I can
however copy files TO the bad node.
Is this a network issue? Bad network card and/or cable? I’ve never seen
this behavior before - does anyone have any thoughts?
A customer of mine is running version 4.23 on a 12 node network. Recently
I
have been having a problem doing an ls -l on one of the nodes from any
other
node. The ls command runs very slowly, and then only displays a few
directories. However, if I do a listing from the “bad” node itself,
everything shows up just fine.
I’ve had similar problems, although in version 4.25E. In my case the root of
the problem was that I was using a proprietary Net.ether82559 driver given
to me by our PC board supplier. The driver turned out to not be entirely
compatible with the Intel 82559ER chip on their PC boards. It manifested
itself exactly as you describe it, plus the problems of this single node
upset the licensing mechanism for the entire QNX network, you had to wait a
minute to get a license to use Watcom C for example.
I switched to the original QNX Net.ether82557 and the problems went away.
This probably didn’t help you, but yes I’ve experienced the problem…
A customer of mine is running version 4.23 on a 12 node network. Recently
I
have been having a problem doing an ls -l on one of the nodes from any
other
node. The ls command runs very slowly, and then only displays a few
directories. However, if I do a listing from the “bad” node itself,
everything shows up just fine.
I also cannot always copy a file FROM the bad node. If I try to cp a file
from this node to another node, cp reports “no route to host”, even though
I
can see all the other nodes and their directories from the bad node. I
can
however copy files TO the bad node.
Is this a network issue? Bad network card and/or cable? I’ve never seen
this behavior before - does anyone have any thoughts?
Thanks in advance -
Chris Holt
Chris,
my small expierence says that the following steps can be will help
you:
use “netinfo -l” for checking how your NIC works ( bad packet presents
?)
“nameloc”( starting in sysinit. ) must run only on the “server”
( or “main”) node. On another nodes use " nameloc -k …" or " slay -f
nameloc" at the end of sysinit.