bad block

[Nostradamus]

Anyone know why I get this during bootup...

NOTICE: Sdsk: Bad block size writing SCSI disk 1 dev 1/64 (ha=0 bus=1 id=0 lun=0) block=0
Block size (0) must be between 512 and 1024.

the block parameter differs from time to time.
/Sören

 

[patw]

Bring the system to single user mode, then run fsck -ofull on your root filesystem.

See if it finds any problems.

IF you see the message "REBOOT NO SYNC" at the end IMMEDIATELY turn off the power switch.

Wait 5 minutes then turn it on and re-run fsck -ofull.

If you wait too long (about 45 seconds or so) the superblock will be updated with a bad image.

Also check /usr/adm/messages for any other SCSI disk complaints.

Finally, it could be bad SCSI termination or a bad SCSI cable - has anyone been inside the box recently to add memory or disks??

 

[Nostradamus]

I can´t bring the system down now. I´ll try fsck the next time I reboot.

Why wait as long as five minutes after power-off?

What will happen if the superblock is overwritten? Will all filesystem-info be erased?

This error has been present since day one. The server was bought in oct. 2000.

Will SCSI-equipment work at all if not properly terminated? I don´t think that´s the problem, but I´ll look at it. Perhaps try some other cable as well. /Sören

 

[dyna176]

-maybe the term 5mins meant just wait til the hdrvs completely quit spinning. i would think that is adequate.
-yes if the superblk is corrupted, the info that stores divvy info will be hosed, therefore fsys is not really erased, but it will be very close to that practically.
-yes, i've seen scsi devs work with wrong termination, and give weird intermittant problems, this is common on alot of tape drv problems i've seen. hope this helps. you might look up sco's technical articles (while still available free), and run a 'cocktail' on this system...

 

[MadhuMs]

try badtrk

madhu_ekm@rediffmail.com

 

[arizonasco]

Don't do anything unless you have a known good full system backup first!

Thanks;
Mike