how to correct a mistake in establishing new kernel

[szhiyong]

Hi,

I make a mistake in establishing new kernel. I first change the makdsiz as 2GB in the hp/ux 10.20. I forget to select the process for the new kernel. Then I reboot it, I found I can not go into the graphical user interface. I can only work on sh.

Can anyone help me to solve this problem?

Thanks!

Zhiyong

[Sridhar Bhaskarla · Accepted solution]

I am a bit confused with your problem. If you want to revert back to the previous kernel, you can copy /stand/vmunix.prev back as /stand/vmunix and then reboot the system. Or you can change the maxdsiz to the previous value (or copy back the system.prev as system) and then rebuild the kernel.

If you are not getting the CDE back, it may not be due to the maxdsiz or the new kernel. There are a lot other reason why the CDE doesn't start. Please give us more details.

-Sri

[szhiyong]

Hi,Sri

Thanks for your reply.

I am still confused why I can not go into the CDE. Before I reboot, I just go to sam--kernel configuration--change the maxdsiz.After I reboot, I found those error message:
.profile[27]:tset:not found.
.profile[31]:tabs:not found.
infs:Do not know how to handle
driver "netbios"
infs:Do not know how to handle
driver "netbisodg"
infs: Do not know how to handle
driver "lmxloop"

I will try to recovery the original kernel.

Best Regards!

Zhiyong

[szhiyong]

Hi,Sri,

Thanks for your reply and help!

I really appreciated. I try the "bdf /usr", I found the error message "bdf does not found", then I check usr directory, there is nothing there. By the way, my workstation is hp/ux 10.20. This is output of "ls -al /dev/vg00"
total 10
root 1024 June,1999 .
bin 4096 Oct,.. ..
sys 64 Sep,5 0x00000a TMPTMP
-- -- -- 0x000009 USR2
-- -- Nov,16 0x000000 group
-- -- Sep,5 0x000001 lvol1
-- -- -- 0x000002 lovl2
. .. ........................
-- -- Sep,5 0x000008 lovl8
-- -- May.27 0x00000a rTMPTMP
-- -- -- 0x000009 rUSR2
-- -- -- 0x000001 rlvol1
-- -- -- 0x000002 rlvol2
..................................
-- -- -- 0x000008 rlovl8

The output of "ls -al /ect/fstab"
sys 556 Sep.27 /etc/fstab

I am a beginner of this list. I am sorry for post many message.

Best Regards!

Zhiyong

[Sridhar Bhaskarla]

Sziyong,

Sorry for the late reply. Well, bdf not found is due to /usr missing. Please check your other thread. To verify which logical volume was mounted as /usr before, do like this

#mount |grep vg00|awk '{print $3|'

This gives the list of all logical volumes that are mounted now. Now find out the logical volumes that are missing in this output but that are there in /dev/vg00.

Let's say they are lvol8,lvol9 etc., Now to find out which logicial volume that was mounted as /usr, do

#ncheck /dev/vg00/lvol8 |more

(this may take a while)

/dev/vg00/usr:
3 /lost+found/.
4 /local/.
5 /bin/.
7 /sbin/.
50 /lib/.
121 /newconfig/.
142 /share/.
151 /lbin/.

If you get output like this, then it is /usr. Otherwise do it on other unmounted logical volumes. The one which gives the above structure is your /usr and you can mount it as

mount /dev/vg00/lvol8 /usr. Now your commands like bdf, test etc will work.

Now edit your /etc/fstab file and add this logical volume in it. You can use any of the entries that are already there in /etc/fstab.

Then reboot. You should get your CDE back.

-Sri