Restoring from tape

[Conrado Camacho]

Jul 3, 2007 21:47:10 GMT

--------------------------------------------------------------------------------
Hello Everyone,

I am restoring some files from my tape but it has taken a lot of time.
How do I know if it is really retreiving the files to my disk?. I am afraid that my hard disk is damaged so, could it be possible that the process is in loop?


Regards
Camacho

[Steven E. Protter]

Shalom,

1) Tell us how you doing the restore, what software. There is evidence whether the restore is working.

2) I am curious as to why you are restoring data from tape to a suspect hard disk. Better to diagnose and fix the disk problem first. If the disk is bad, you could indeed be wasting your time.

ioscan -fnC disk

use vgdisplay and lvdisplay to see what disks the data you are trying to restore is on. If ioscan shows a disk as NO_HW then stop the restore, fix the disk problem and restore the files later.

SEP

[Hoang Chi Cong_1]

Hi

Normally, when you do backup to tape without any Backup Software - means that you backup to tape by OS's backup command example: fbackup, tar .....

If you want to get a file or some files from the backup tape, the only way is restoring entire of the backup session.

If you are in doubt that the hard disk was damaged so why don't you check for the hard disk first???
You can check from the syslog file...I have faced so many case with hard disk damage but just has bad block (the disk still runing, just sometime, the write and read process had error and you can find it from the syslog.log file)

Hope this helps
Hoang Chi Cong

[Conrado Camacho]

Certainly,
I supposed the hard disk is damaged because when I try to open my oracle database it hangs. but sometimes when I issue a ioscan -fnC disk, it is shown properly. When I issue a diskinfo for /dev/rdsk/c0t12d0, it shows me the disk information properly. Reviewing syslog I can see that the server does not attach the disk when the server is restarted (and shows an error) but lately it is shown with ioscan. Both previuos commands sometimes work and some not.
I am trying to restore my (two) database files) only throught SAM (interactive back up and recovery)

Is there any tool that i can use to diagnose my disk besides stm?

Thanks in advance
Regards
Camacho

[Hoang Chi Cong_1]

Hi again :)

Hm...In your case, you should contact to HP to get support because this case need to check carefully. That's my idea...data is the most importance :)

In normal, if a disk had a bad block, the error like:
Jun 5 13:22:06 procser1 vmunix: SCSI: Async write error -- dev: b 31 0x012000, errno: 126, resid: 32768,
Jun 5 13:22:06 procser1 vmunix: blkno: 11593216, sectno: 23186432, offset: 3281518592, bcount: 32768.
Jun 5 13:22:06 procser1 vmunix: blkno: 11593664, sectno: 23187328, offset: 3281977344, bcount: 32768.
Jun 5 13:22:06 procser1 vmunix: blkno: 11593632, sectno: 23187264, offset: 3281944576, bcount: 32768.
Jun 5 13:22:06 procser1 vmunix: blkno: 11593600, sectno: 23187200, offset: 3281911808, bcount: 32768.
Jun 5 13:22:06 procser1 vmunix: SCSI: Async write error -- dev: b 31 0x012000, errno: 126, resid: 16384,
Jun 5 13:22:06 procser1 vmunix: SCSI: Async write error -- dev: b 31 0x012000, errno: 126, resid: 32768,
Jun 5 13:22:06 procser1 above message repeats 3 times
Jun 5 13:22:06 procser1 vmunix: blkno: 11593584, sectno: 23187168, offset: 3281895424, bcount: 16384.
Jun 5 13:22:06 procser1 vmunix: blkno: 11593568, sectno: 23187136, offset: 3281879040, bcount: 16384.
Jun 5 13:22:06 procser1 vmunix: SCSI: Async write error -- dev: b 31 0x012000, errno: 126, resid: 32768,
Jun 5 13:22:06 procser1 vmunix: SCSI: Async write error -- dev: b 31 0x012000, errno: 126, resid: 16384,
Jun 5 13:22:06 procser1 vmunix: blkno: 11593536, sectno: 23187072, offset: 3281846272, bcount: 32768.
Jun 5 13:22:06 procser1 vmunix: blkno: 11593504, sectno: 23187008, offset: 3281813504, bcount: 32768.
Jun 5 13:22:06 procser1 vmunix: blkno: 11593472, sectno: 23186944, offset: 3281780736, bcount: 32768.
Jun 5 13:22:06 procser1 vmunix: blkno: 11593440, sectno: 23186880, offset: 3281747968, bcount: 32768.
Jun 5 13:22:06 procser1 vmunix: blkno: 11593408, sectno: 23186816, offset: 3281715200, bcount: 32768.
Jun 5 13:22:06 procser1 vmunix: blkno: 11593376, sectno: 23186752, offset: 3281682432, bcount: 32768.
Jun 5 13:22:06 procser1 vmunix: blkno: 11593344, sectno: 23186688, offset: 3281649664, bcount: 32768.
Jun 5 13:22:06 procser1 vmunix: blkno: 11593312, sectno: 23186624, offset: 3281616896, bcount: 32768.
Jun 5 13:22:06 procser1 vmunix: blkno: 11593280, sectno: 23186560, offset: 3281584128, bcount: 32768.
Jun 5 13:22:06 procser1 vmunix: blkno: 11593248, sectno: 23186496, offset: 3281551360, bcount: 32768.
Jun 5 13:22:06 procser1 vmunix: SCSI: Read error -- dev: b 31 0x012000, errno: 126, resid: 1024,
Jun 5 13:22:06 procser1 vmunix: SCSI: Async write error -- dev: b 31 0x012000, errno: 126, resid: 32768,

I get these error massage when a disk had bad but ioscan and diskinfo show me the disk is OK.

Goodluck!

[Conrado Camacho]

I think something wrong happened with my disk
With diskinfo for the disk damaged I have:

# diskinfo /dev/rdsk/c0t12d0
SCSI describe of /dev/rdsk/c0t12d0:
vendor: SEAGATE
product id: ST39173WC
type: direct access
size: 0 Kbytes
bytes per sector: 0

Please note that there is not size info.

I can see all information contained in there using simple unix commands.

Regards
Camacho

[Torsten.]

Consider the disk is dead.
Get it replaced!

[Conrado Camacho]

Uppss,
That is another hard problem my server is old enough to be in a museum and disks are almost extinct.

thanks anyway to all of you

Regards,
Camacho

[Conrado Camacho]

Disk replacement