Hi -

I've read the other posts about this issue, but I wanted to know if
anyone had my particular issue before and had a solution.

We have an isolated network with two subnet addresses on the same
physical line. This configuration was first put into play when we were
using vxWorks 5.1.1 some time ago. We have upgraded to vxWorks 5.4,
and it is on this system that we get this arptnew failed on c009e60b

We are running vxWorksd 5.4. on MVME2700 boards. Each board has its
own ethernet connection and an IP address of On the same
physical network we have PCs running Solaris that we use as displays
that have IP addresses of 192.9.230.x. The main bus controller MVME2700
board has an IP address of In the startup script for all
the MVME2700 boards we add a route to the subnet via the
following command.

routeAdd "", ""

so the routing table of the MVME2700 board with an IP address of looks like this

destination gateway flags Refcnt Use
---------------------------------------------------------------------------- 3 0 0 dc0 101 0 0 dc0 3 2 10885 dc0

destination gateway flags Refcnt Use
---------------------------------------------------------------------------- 5 1 0 lo0

This board nfsMounts to and writes to a file on the
Solaris PC. Periodically during the writing we ge the error
0x178cfb8 (tNetTask): arptnew failed on c009e60b.

The following is the result of the mRouteShow.

Destination Mask TOS Gateway Flags RefCnt Use
Interface Proto 0 0 3 0 0
dc0 0 0 0 5 1 0
lo0 0 ffffff00 0 101 0 0
dc0 0 ffffff00 0 3 2 10473
dc0 0
value = 0 = 0x0

Now I assume that I am getting the arptnew fail error because of using
a different subnet 192.9.230.x that is on the same physical network.
How can I get around this, I need to use the 192.9.230.x subnet because
changing it at this point is not an option.

Any help would be greatly appreciated....

Lorrie McDougald