SUMMARY - mangled disk error - puzzled

From: Lisa Weihl <lweihl_at_cs.bgsu.edu>
Date: Thu Feb 05 2004 - 13:34:26 EST
I've received enough replies now to realize I've been mistaken.  I could've 
sworn 7 years ago when I went through Sun training that if one partition 
ENDED on a cylinder number the other began with that number.  Meaning they 
didn't really overlap.  0-3716 means that partition stopped at 3716 but 
obviously I'm wrong.  If I'd just looked for one partition table on the web 
I would've seen that.  Sorry guys (and gals).  I will be reformatting and 
restoring from tape.  Thanks for helping me out.

The following people were kind enough to reply and I appreciate their help.

Minoti Koul
Anthony Talltree
pitsanu@pacbell.net
Trevor McLachlan
Ray Brownrigg
Kevin Buterbaugh
Michael Schulte
Hichael Morton
Jay Lessert


*****************************************************************************************
Original message....

I've looked online and through the sunmanagers archives for solutions to my 
problem but have to admit that I'm still puzzled.

I have an Ultra 5 running as a server. It's running Solaris 2.6 and has an 
Ultra Wide SCSI card. On one channel I have a DLT tape drive.  On the other 
channel I have 2 9GB hard drives.  They been up and running without fail 
since 1998.  Recently one of the filesystems began to give me 
trouble.  Lots of error messages like:  /filesystem bad dir ino 2 at offset 
242 : mangled entry.  Running fsck (yes I did unmount it) on it would find 
lots of unknown file type I=# Clear? errors. Originally fsck fixed it but I 
wasn't comfortable that all files were still connected.  Within a few 
minutes another fsck revealed more errors so I decided to newfs and restore 
from tape. The filesystem was residing on one of the disks in a 4GB 
partition with one other 4GB parition and a 500 mb partition.  Neither of 
the other filesystems was complaining.  Still I had an emtpy 4GB parition 
on the second drive so I moved the filesystem over there.  I issued the 
newfs command and restored the data from tape and ran fsck after the 
restore, it was fine.  Now here we are less than 1 week later and I'm 
getting the same mangled entry errors.  Still none of the other filesystems 
are complaining.  Ideas on what could be causing this?  Ideas I saw on the 
web were overlapping partitions or bad SCSI cable.  Here's the partition 
info from the format command, I'm just showing the relevant slices....

0	0-3716
2	0-7498
4	3716-7432

As far as I'm aware those cylinders don't overlap.  I haven't tried chaning 
SCSI cables yet.  I'm thinking that may not be the issue because wouldn't 
the other filesystems be reporting errors also?

I hope I haven't left out any relevant information.  Will summarize.  TIA 
for any help.

Lisa
**********************************************************************************
Lisa Weihl, System Administrator          E-mail:      lweihl@cs.bgsu.edu
Department of Computer Science          Office:        Hayes 225
Bowling Green State University              Phone:     (419) 372-0116
Bowling Green, Ohio 43403-0214           Fax:         (419) 372-8061
_______________________________________________
sunmanagers mailing list
sunmanagers@sunmanagers.org
http://www.sunmanagers.org/mailman/listinfo/sunmanagers
Received on Thu Feb 5 13:34:19 2004

This archive was generated by hypermail 2.1.8 : Thu Mar 03 2016 - 06:43:27 EST