Henrique Haas schrieb:
My network is now working. Was a simple error of
configure the others ortmanagers in the network which
must share information of the process. But let's see
if I can help you!
My machines is a AMD Athlon 2000+ . I'm using a
2.6.8.1 kernel version patched with a rtai-3.1
(hal7-2.6.8.1.patch). A simple documentation is in
http://www.captain.at/programming/rtai/ . After I
installed a RTAI, the tests were very important. Could
appear errors as the /dev/rtf not present, but Google
known. After the RTAI is fine, I compile the rtnet
reportedly a rtnet.org site. In the rtnet.conf I
replace only the name of my NIC driver. I run the
rtping and so on.
In the ORTE compilation I run a:
./configure --with-rtai=/usr/realtime/
--with-rtnet=/usr/local/rtnet/
In the ortemanager initialization I run:
./ortemanager --peer <list-of-IPs-ofothers-machines>
-e &
I hope help you!
We followed your advises.
Unfortunatly it looks like what we got is a user-space compilation of
ORTE. In our environment it is necessary to specify the
"--with-linux=PATH_TO_ADEOS_PATCHED_KERNEL" flag, in order to compile a
realtime version of ORTE. As far as I know, in the realtime-version
liborte and ortemanager both are kernel-modules called orte_rt.ko and
ortemanager_rt.ko. In our installation the compilation of ORTE with a
2.6.x Linux aborts.
We are using a 2.4.31 Linux with adeos-patch, rtai 3.1, RTnet 0.8.3 and
ORTE 0.3.0. With that versions we are able, to compile ORTE and insert
the orte_rt.ko and ortemanager_rt.ko modules. But a rtai-trap is
triggerd, when trying to insert the h_publisher_rt.ko module from the
ORTE hello-world example...
Regards kai
-------------------------------------------------------
This SF.Net email is sponsored by:
Power Architecture Resource Center: Free content, downloads, discussions,
and more. http://solutions.newsforge.com/ibmarch.tmpl
_______________________________________________
RTnet-users mailing list
RTnet-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/rtnet-users
