Previous in Forum: error 1304.error writing to file c:\   Next in Forum: C++
Close
Close
Close
4 comments
Rate Comments: Nested
Anonymous Poster

Unix Problem

07/01/2008 12:50 AM

Can any body tell help me, I have Digital make Alpha station 200 4/166 with Digiatl unix4.0 OS. Now I got a error while rebooting. That is "medium error at block:9846, hard error detected- non recoverable medium error then came out to promt disply.so system not booting. This unit is connected to a real time application processor units with a special application software developed by ABB.

Please help me now, How to get Hard disk to Hard disk copy in Unix and how to solve this error problem.

Reply
Interested in this topic? By joining CR4 you can "subscribe" to
this discussion and receive notification when new comments are added.
Associate

Join Date: Jul 2007
Location: BANGALORE, INDIA.
Posts: 45
#1

Re: unix problem

07/01/2008 2:45 AM

Dear guest, whatisyourname, there is no disk to disk copy like the msdos diskcopy. But you can use dd (diskdump) command. You will have to connect a freshly formatted HDD with the exact slices as the original disk. I don't much Digital Unix but old version of unix in general allowed s7 (Slice 7) covering entire disk, but this is not so nowadays. So you create matching slices and using dd command make disk-to-disk copy. I don't about Digital Unix, as already told, but in Solaris we use it like dd if=/dev/dsk/c0t0d0s0 of=/dev/dsk/c0t1d0s0 etc. Try your luck and let me know, meanwhile I get some help from colleague who familiar with Digitial Unix.

__________________
Dream is not what you see in your sleep, but is the one which does not let you sleep - APJ ABDUL KALAM
Reply
Active Contributor

Join Date: Jun 2008
Location: NJ USA
Posts: 13
#2
In reply to #1

Re: unix problem

07/01/2008 10:45 PM

I'm not so sure that dd will work. It seems like there is a hard read error on the boot slice or on a slice that is read during bootup. If you try to dump the slice that has the hard error, you will end up with a corrupted slice that most likely will be unreadable. You may need to go in with a binary or hex editor and try to "fix" the corrupted area (that assumes that you know what is SUPPOSED to be in the corrupted area). Then again, I may be misunderstanding the problem (as has happened before... 8^( ) and I may be thinking the problem is more complicated than it really is. Anyway, Good Luck

__________________
ARMY MARS - THE VOICE OF THE ARMY
Reply
Anonymous Poster
#3

Re: Unix Problem

07/02/2008 1:10 AM

First Of All the failure must be cleared. For solaris the command "format" and then "repair" works in most cases (in some cases you may not recover the data). And then if you can clone the whole HDD that would be nicer, using ufsdump and then ufsrestore (these were the commands for solaris).Step by step after you make the partitions and partition sizes similar to the original one and NEWFS them: 1-boot from CDROM 2-mount /dev/dsk/c0t2d0s0 /mnt (t2 is the controller name for the new HDD) 3-ufsdump 0f - /dev/dsk/c0t3d0s0 | (cd /mnt; ufsrestore xf -) 4-if this is not a booting HDD switch off the machine and plug in the new HDD instead of old one. If it is booting HDD #umount /mnt,#cd /usr/platform/`uname -i`/lib/fs/ufs,#installboot bootblk /dev/rdsk/c0t2d0s0. Good luck

Reply
Anonymous Poster
#4
In reply to #3

Re: Unix Problem

07/02/2008 8:36 AM

Always need a plan B. The next time make a backup!

Reply
Reply to Forum Thread 4 comments
Copy to Clipboard

Users who posted comments:

Anonymous Poster (2); ARMY-MARS-GUY (1); ganapa (1)

Previous in Forum: error 1304.error writing to file c:\   Next in Forum: C++
You might be interested in: Hard Drives, Error Correction Chips, RAID Disk Arrays

Advertisement