We are trying to rebuild kernels with a new nodeid however, when we
reboot that node, we have about a 50% failure rate.
Our kernel has been verified to be correct so we are under the
impression that it is a result of the .boot file being copied over an
“active” kernel.
Any help would be appreciated.
–
Dr. Jörg Kampmann - IBK-Consult for Real-Time and Embedded Systems
D-31228 Peine - Tel.:+49-177-276-3140 - Fax: +49-5171-13385 http://www.ibk-consult.de
===== QNX is the better Choice for Real-Time: http://www.qnx.com ====
We are trying to rebuild kernels with a new nodeid however, when we
reboot that node, we have about a 50% failure rate.
Our kernel has been verified to be correct so we are under the
impression that it is a result of the .boot file being copied over an
“active” kernel.
Any help would be appreciated.
–
Dr. Jörg Kampmann - IBK-Consult for Real-Time and Embedded Systems
D-31228 Peine - Tel.:+49-177-276-3140 - Fax: +49-5171-13385 http://www.ibk-consult.de
===== QNX is the better Choice for Real-Time: > http://www.qnx.com > ====
We have tried two different methods to create the kernel.
Found the difference between two kernel’s and changed the nodeid within
the kernel itself.
Not recommended.
used the buildqnx command.
This is the usual method – modify the -L option to the Proc32 command
line, build a new image, cp old image to /.altboot, cp new image to /.boot
then reboot.
Both work at the same success rate. With identical kernels (cksum said
so), the success rate is still 50%.
When they fail, what sort of failure message do you get? Or, how do
they fail?
what kind of kernel failure do you have ? can you post the kernel build file
? you boot from hdd or from another locations for example from network ?
which version/patch of qnx ?
Can you confirm that the new boot is is actually within the first 1024
cylinders (i.e. 0-1023).
Make sure that your new image is not too big. Usually the limit is
512KB, it could be slightly more or less for reasons that I’ve never fully
understood.