Hi,
At the moment I am experimenting with a two node setup. Both systems have
installed QNX Momentics 6.2:
uname -a
QNX node2 6.2.0 2002/05/21-12:17:24edt x86pc x86
patched with patchA.
On both systems QNET is started with
mount -Tio-net -o"crc=0,resolve=ndp" /lib/dll/npm-qnet.so
as the last line in /etc/rc.d/rc.local
When both systems are up and running everything seem normal:
ls /net
node1 node2
The systems can ping each other, telnet works etc…
However, when I do on node2
ls -l /net
the system responds with
ls: Host is down (/net/node1)
total 4
drwxr-xr-x 4 root root 2048 Apr 03 08:53 node2
Other commands trying to do something with /net/node? also fail. At this
point ping etc… still works.
Obviously QNET is not entirely down:
ls -l /net/node2
works.
And when I give node1 a different hostname with hostname node1new
that name appears at node2 in /net.
The strange thing is that it has worked before: I was able to put my login
directory on node2 at /net/node1/fs/hd0-qnx4-2/home/pim and had no problems
until some fatal moment (on node1 make 
).
Is there a way to find out where and why things go wrong now ?
If you need more information I will be glad providing so.
Thanks,
Pim Bollen
What is the network driver? After node2 dropped node1, can “ls /net/node1”
(on node2)
bright it back? Does nicinfo said anything interesting? Is takeing off
“crc=0” helpful?
-xtang
Pim Bollen <pbollenNOSPAM@vimec.nl> wrote in message
news:b6jdth$55s$2@inn.qnx.com…
Hi,
At the moment I am experimenting with a two node setup. Both systems have
installed QNX Momentics 6.2:
uname -a
QNX node2 6.2.0 2002/05/21-12:17:24edt x86pc x86
patched with patchA.
On both systems QNET is started with
mount -Tio-net -o"crc=0,resolve=ndp" /lib/dll/npm-qnet.so
as the last line in /etc/rc.d/rc.local
When both systems are up and running everything seem normal:
ls /net
node1 node2
The systems can ping each other, telnet works etc…
However, when I do on node2
ls -l /net
the system responds with
ls: Host is down (/net/node1)
total 4
drwxr-xr-x 4 root root 2048 Apr 03 08:53 node2
Other commands trying to do something with /net/node? also fail. At this
point ping etc… still works.
Obviously QNET is not entirely down:
ls -l /net/node2
works.
And when I give node1 a different hostname with hostname node1new
that name appears at node2 in /net.
The strange thing is that it has worked before: I was able to put my login
directory on node2 at /net/node1/fs/hd0-qnx4-2/home/pim and had no
problems
until some fatal moment (on node1 make > > ).
Is there a way to find out where and why things go wrong now ?
If you need more information I will be glad providing so.
Thanks,
Pim Bollen
Xiaodan Tang wrote:
What is the network driver? After node2 dropped node1, can “ls /net/node1”
(on node2)
bright it back? Does nicinfo said anything interesting? Is takeing off
“crc=0” helpful?
Removing crc=0 did the trick.
Thank you.
Pim