LinuxQuestions.org
Review your favorite Linux distribution.
Go Back   LinuxQuestions.org > Forums > Linux Forums > Linux - Newbie
User Name
Password
Linux - Newbie This Linux forum is for members that are new to Linux.
Just starting out and have a question? If it is not in the man pages or the how-to's this is the place!

Notices


Reply
  Search this Thread
Old 02-26-2009, 01:17 PM   #1
deathsfriend99
Member
 
Registered: Nov 2007
Distribution: CentOS 6
Posts: 195

Rep: Reputation: 22
fdisk gives: Unable to read /dev/sdc


I had a drive go bad in RAID5. I failed the drive, removed the drive, and replaced it with a new drive. The drives that are currently in the machine are 1TB seagates. I am replacing it with a 1TB Western Digital.

I attmpted to add the new drive with the same paritioning info as the old drives:
sfdisk -d /dev/sdd | sfdisk /dev/sdc

Then added the device back to RAID5. That went fine, but as the RAID was rebuilding, I started getting errors:

Buffer I/O error on device sdc, logical block 0
Buffer I/O error on device sdc, logical block 1
Buffer I/O error on device sdc, logical block 2
Buffer I/O error on device sdc, logical block 3

RAID failed the drive and stopped the rebuild.

I removed the drive from RAID with mdadm and did an fdisk. I get Unable to read /dev/sdc

I find it hard to believe that a 2nd drive, from a diff manufacturer is bad. Have I done something wrong here or do I just have a bad drive?

/var/log/messages:
Feb 26 12:04:55 vv-atl1 smartd[2896]: Device: /dev/sdc, failed to read SMART Attribute Data
Feb 26 12:34:55 vv-atl1 smartd[2896]: Device: /dev/sdc, not capable of SMART self-check

but smartctl -H /dev/sdc gives health OK
 
Old 02-26-2009, 02:28 PM   #2
thorkelljarl
Senior Member
 
Registered: Jun 2008
Posts: 1,787

Rep: Reputation: 211Reputation: 211Reputation: 211
Have you tried this?

1) Have you tried any of WDs tools to see what they tell you? Is the HDD on this the page?

http://support.wdc.com/product/downl...vel1=6&lang=en

2) If you failed in the attempt to read and transfer the partitioning scheme using the command sfdisk and the command was without effect, would not fdisk give you the result "unable to read.." since no partitions have been created? You might try to create the needed partitions by hand with cfdisk or fdisk and see if the disk will accept partitioning.

I would have written Nr.2 as Nr.1 had I been thinking quicker. Good luck

Last edited by thorkelljarl; 02-26-2009 at 02:33 PM.
 
Old 02-26-2009, 05:26 PM   #3
gergely89
Member
 
Registered: Feb 2009
Posts: 100

Rep: Reputation: 21
The sfdisk man pages tells us that it is very dangerous to repartition a device with sfdisk and also that sfdisk doesn’t understand the GUID Partition Table (GPT) and it is not designed for large partitions. I would consider a 1 TB disk as quite large single partition and I guess that your problem could just as well be caused by all these sfdisk peculiarities.

Maybe format and partition your new disk the traditional way (/bin/fdisk), maybe even add the filesystem yourself - and afterwards try again to fit it into your RAID setup.

linux

Last edited by gergely89; 02-28-2009 at 12:05 AM.
 
Old 02-26-2009, 06:42 PM   #4
deathsfriend99
Member
 
Registered: Nov 2007
Distribution: CentOS 6
Posts: 195

Original Poster
Rep: Reputation: 22
The drive passes WD's diagnostics. I re-partitioned it in XP with NTFS and put it back into the Linux machine. Then fdisk could read it. I removed the NTFS partition and added a linux partition. I added the drive back to the array. I'll cross my fingers. It may be that this thing really wanted to be partitioned by WD's software first. That would be bizarre.
 
Old 02-26-2009, 07:23 PM   #5
thorkelljarl
Senior Member
 
Registered: Jun 2008
Posts: 1,787

Rep: Reputation: 211Reputation: 211Reputation: 211
Nah

I think that it was blank, that is all zeros from WD, and sfdisk didn't do anything with it, leaving it still blank, that is no flag, no label, no partitions, nothing. You came out of it well enough in the end, whatever was the problem. I trust you can recreate your partitions without too much trouble. Then again, doesn't a system administrator always get through in the end?

Last edited by thorkelljarl; 02-27-2009 at 07:55 PM.
 
Old 02-27-2009, 07:08 PM   #6
deathsfriend99
Member
 
Registered: Nov 2007
Distribution: CentOS 6
Posts: 195

Original Poster
Rep: Reputation: 22
Just to update. I still have the same problem. I'm thinking we have a bad SATA port on our controller. We may have to change out the MB on this one. The SATA controller is onboard.
 
Old 02-27-2009, 07:53 PM   #7
thorkelljarl
Senior Member
 
Registered: Jun 2008
Posts: 1,787

Rep: Reputation: 211Reputation: 211Reputation: 211
What about this.

Did you try switching the SATA cable to the drive just in case that is bad? Cables can produce funny faults, but at least they're cheap.

I assume that you have done all the fault finding hardware swaps, but I can't understand how the first problem, formatting a new disk, is related to a faulty controller. There have been other symptoms, I presume. I'm just curious; I might learn something.

Last edited by thorkelljarl; 02-27-2009 at 07:56 PM.
 
Old 02-28-2009, 12:08 PM   #8
deathsfriend99
Member
 
Registered: Nov 2007
Distribution: CentOS 6
Posts: 195

Original Poster
Rep: Reputation: 22
Yes, I replaced the SATA cable with a new one that came with the new HD.

A full explanation may clear up some questions:
I recently (less than 2 weeks ago) built a new Samba/LDAP server and put it in RAID5 with 4 brand new 1TB Seagate HD's. I noticed shortly after that one of the drives in the array had failed. It was in RAID5 so no data was lost, but I wanted to get the array to a healthy state. I didn't have time to return the drive, but I did have a new Western Digital 1TB drive still in its box, so I plugged that in and copied the partition. The array worked fine rebuilding until it got to about 5-7% and then the drive failed. It appeared at first that the partition didn't take. I looked at it and it had changed size. I found it hard to believe that 2 brand new drives of different brands each with a different SATA cable would be bad within days, so I tested each of them with their respective diagnostic tools. Both drives passed. All I can conclude at this point is that I have a bad SATA port on the MB. Hopefully I can return it.

Last edited by deathsfriend99; 02-28-2009 at 12:10 PM. Reason: I can't spell
 
Old 03-01-2009, 07:00 PM   #9
thorkelljarl
Senior Member
 
Registered: Jun 2008
Posts: 1,787

Rep: Reputation: 211Reputation: 211Reputation: 211
Thank you

Thanks for the update. I did learn a little something says a non-administrator.
 
  


Reply


Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is Off
HTML code is Off



Similar Threads
Thread Thread Starter Forum Replies Last Post
Installing RIP LinuX on a USB drive - device name /dev/sdc becomes /dev/sda Mleahy Linux - Software 1 07-30-2008 09:57 PM
Unable to read /dev/sdc madunix Linux - Enterprise 2 11-25-2007 09:06 AM
grub on /dev/sdc and live-usb on /dev/sdc1 fitzov Linux - General 2 04-25-2007 12:11 PM
fdisk unable to read /dev/sda frankie_DJ Slackware 1 09-08-2005 09:44 PM
fdisk gives unable to read /dev/hda... mbr is gone... what can it be? slackie1000 Linux - Hardware 5 01-27-2004 05:29 AM


All times are GMT -5. The time now is 12:12 PM.

Main Menu
Advertisement
My LQ
Write for LQ
LinuxQuestions.org is looking for people interested in writing Editorials, Articles, Reviews, and more. If you'd like to contribute content, let us know.
Main Menu
Syndicate
RSS1  Latest Threads
RSS1  LQ News
Twitter: @linuxquestions
Facebook: linuxquestions Google+: linuxquestions
Open Source Consulting | Domain Registration