SUMMARY: newfs problem

From: Daniel Hurtubise (daniel@CANR.Hydro.Qc.CA)
Date: Tue Mar 16 1993 - 20:54:42 CST


My original posting:

platform: sun4/390
          sunos 4.1.2
disk: Seagate Elite-2 ST24200N

I got some "bad blocks" on my disk, so I did a destructive surface
analysis on it to reassign the bad blocks and update the defect list.
I then formatted the disk and relabled it.

I have five other simular disks that gave no problems
and was able to do the newfs successfully.

I get the following error when trying to do a newfs.
What does it mean?

machine# newfs -i 8192 /dev/rvc0c
/dev/rvc0c: 4155984 sectors in 2604 cylinders of 19 tracks, 84 sectors
        2127.9MB in 163 cyl groups (16 c/g, 13.07MB/g, 1536 i/g)
super-block backups (for fsck -b #) at:
 32, 25664, 51296, 76928, 102560, 128192, 153824, 179456, 205088,
 230720, 256352, 281984, 307616, 333248, 358880, 384512, 410144, 435776,
 461408, 487040, 512672, 538304, 563936, 589568, 615200, 640832, 666464,
 etc, etc, 3448064, 3473696, 3499328, 3524960, 3550592, 3576224, 3601856,
 3627488, 3653120,write error: 3678768
 wtfs: I/O error

The error doesn't always happen on the same super-block backup. Actually
after 4 or 5 tries the newfs worked, but I am curious in knowing more
about the error message and if this means that the filesystem will be
less stable.


Well, while waiting for all of the responses to come in, and by the fact
that I managed to successfully newfs the c partition, I proceeded to copy
data to the disk. The result was another "full system" hang. That is, the
server and its console would freeze up tight! Consequently, I stopped using
the disk.

I called the vendor of my third party SCSI-2 controller who explained to
me that the problem seemed to be with the SCSI interface in the disk since
the controller was not able to get a response from the disk after sending
it an interrupt which in turn caused a queue flush and the a stale mate
between the the disk, controller, and vme bus.

Your responses indicated that an i/o error from wtfs can indicate a problem
from one or more of the following:

1) The mkfs utility is having a hard time locating bad blocks
2) The disk interface is malfonctioning
3) A bad cable, bad terminator, bad controller and/or bad disk
4) The SunOS4.1.X filesystem cannot be larger than 2048MB
5) The utility is trying to access more disk than there actually is

All of these points, checked out ok with my configuration. I removed
the suspected culprit from the daisy chain of 3 disks and proceeded
to test the other to reveal no problems.

I'm having the disk replaced!

Thanks for your help:

deb@beaux.ATWC.Teradyne.COM (MOTHER DAEMON)
David Fetrow <> (Michael Klein)
ups!kalli!kevin@fourx.Aus.Sun.COM (Kevin Sheehan {Consulting Poster Child})
Robert Haddick <>

And to the others to whom I responded directly or that
I accidently deleted from my mail list...


Daniel Hurtubise
Hydro Quebec

This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:07:36 CDT