Does Linux work with an internal 1 TB SATA disk?

dunkel · 09-05-2008, 05:00 PM

Seagate labels the 1 TB disk drive "Linux ready"... My 2.6.18-92.1.10.el5 desktop works fine with two PATA and a small capacity SATA internal HD. My CentOS-5.2-i386-bin-DVD.iso md5sum is correct, "growisofs -dvd-compat -speed=1 -Z /dev/hdc=CentOS-5.2-i386-bin-DVD.iso" writes the DVD+R without errors, and the installation disk check finds no DVD problems. But when only the 1 TB Seagate Barracuda 7200.11 ST31000340AS is connected, the CentOS DVD default installation spends many minutes formatting and then fails with the popup:

"Error
-----------
An error occurred trying to format VolGroup00/LogVol00. This problem
is serious, and the install cannot continue.

Press <Enter> to reboot your system."

After this "reboot", no more hard disk can be found. Did anybody succeed to install an internal 1 TB SATA HD on CentOS 5.2? My ESC P4M900T-M motherboard only supports SATA 150 and the "1.5 Gbits per second jumper" of this Seagate drive appears appropriate. KNOPPIX 5.3 does not succeed to "fsck /dev/sda" or "mkfs.ext3 /dev/sda2". What else could be wrong?

ThanX

drewbug01 · 09-05-2008, 05:05 PM

Perhaps it is not the fault of Linux. You may have a faulty piece of hardware.

Question(s): Are you able to install other distros on it? If you take it to another computer and connect it, does it work? Does Windows install on it?

I would advise trying all of those things and seeing where you get on it.

Also, instead of using LVM, can you just partition it normally?

-- Drew

CRC123 · 09-05-2008, 05:31 PM

I found this article on google.

Specifically, it says this:

Quote:

n order to limit the Linux kernel memory usage, there is a limit of 65,536 physical extents (PE) per logical volume (LV). Hence, the LVM PE size will directly determine the maximum size of a logical volume (LV)! For example, 4MB PE size (the default PE size) will limit single logical volume (LV) to 256GB, 16MB PE size will limit single LV to grow beyond 1TB, etc.

Perhaps your Physical Extent size is not set big enough.

i92guboj · 09-05-2008, 05:45 PM

I don't know anything about LVM, never bothered with it really.

Linux shouldn't have a problem with that. I've had a 1TB Western Digital SATA drive working on this very machine, and it has not a 1TB Seagate disk, neither of them gave a single problem.

I would double check the cables, play around a bit with them, try changing the ordering of the drives (just in case the problem is in one of the ports in your mother board). If nothing works, and it fails while formating, the problem is probably in the drive. Test is in another box.

amani · 09-05-2008, 07:30 PM

In knoppix you must be root to mkfs

It must be having qtparted and parted too.

What are the bios settings for SATA?

enable ahci

For that drive any of the settings will work.

and check the PE

billymayday · 09-05-2008, 07:36 PM

Questions - do you need LVM, or would you be just as happy with a straight ext3 partition?

Have you tried setting up the LVM from a live CD (maybe F9)?

dunkel · 09-06-2008, 01:26 AM

Thank you for the responses! First I tried using the SATA1 (instead of SATA2) port on the motherboard with a new SATA cable. No improvement. Then during the installation, I manually partitioned /dev/sda:

/dev/sda1 /boot ext3 Format 101 MB 1 13 (force primary)
/dev/sda2 swap Format 4000 MB 14 523
/dev/sda3 / ext3 Format 949764MB 524 121601

The installation gave the same formatting popup, now complaining about sda3 instead of VolGroup00/LogVol00.

Then I booted using the CentOS-5.1-i386-LifeCD and started qtparted v 0.4.5-cvs. /dev/sda1 and /dev/sda2 are formatted, /dev/sda3 exists but is unformatted. Formatting it shows in the Progress popup: "Scanning all disk partitions. There was a problem with mkfs.ext3." So as root, I issued from the command line:

# mkfs.ext3 /dev/sda3
mke2fs 1.39 (29-May-2006)
Filesystem label=
OS type: Linux
Block size=4096 (log=2)
Fragment size=4096 (log=2)
121585664 inodes, 243139758 blocks
12156987 blocks (5.00%) reserved for the super user
First data block=0
Maximum filesystem blocks=0
7421 block groups
32768 blocks per group, 32768 fragments per group
16384 inodes per group
Superblock backups stored on blocks: [3 lines of numbers]
Warning: Could not read block 0: Attempt to read block from filesystem resulted in short read
Writing inode tables: done
ext2fs_mkdir: Attempt to read block from filesystem resulted in short read while creating root dir

#fsck /dev/sda
fsck 1.39 (29-May-2006)
e2fsck 1.39 (29-May-2006)
fsck-ext2: Attempt to read block from filesystem resulted in short read while trying to open /dev/sda
Could this be a zero-length partition?

I also tried to install Windows XP Professional on this disk instead. XP reports it would inspect the computer - and within half an hour it does neither report anything nor start the XP installation.
Am I certain these problems originate from the new faulty Seagate drive?

- Reinhard

CRC123 · 09-06-2008, 10:31 AM

If xp and Linux have problems with the disk, you probably have a faulty disk. Where did you buy the drive from? Exchange it or call Seagate support and get them to send a new one.

jay73 · 09-06-2008, 10:35 AM

I have 1TB Samsung spintpoints that work fine with Ubuntu, Fedora and Centos, yet I have read complaints from other spinpoint users that run similar to yours. This would suggest that it is not the drive itself that is to blame. I would guess that you may need to update your BIOS.

amani · 09-06-2008, 11:21 AM

Is fsck /dev/sda supposed to work?

#man fsck says something else

what does sdparm or hdparm say?

You can always check the drive with Seagate's utility too.

i92guboj · 09-07-2008, 12:27 AM

Quote:

Originally Posted by amani

Is fsck /dev/sda supposed to work?

No. Unless you formated the device as a whole, instead of a partition, which is highly odd (though not impossible, I must admit).

Fsck only checks filesystems. You usually format partitions, like sda1, sda2... but not sda (which is the whole device).

Electro · 09-07-2008, 02:01 AM

It seems to me those utilities are too old, so that could be one of your problems. The other problem is you are using a Seagate drive. They are OK if you want more heat in your computer and some problems in the future because it does not like the quality of power from an aging power supply. Seagate drives do have hardware problems with some controllers. Western Digital or Hitachi are a lot better.

Yes, it will help to update the BIOS even though Linux goes around the BIOS to calculate the geometry of all the hard drives.

I suggest use a small hard drive to install CentOS. Then update the disk utilities. Next partition the 1 TB hard drive and copy the files over. If during partitioning and formating operations that the utilities fail on the 1 TB hard drive, the BIOS may have to be updated.

dunkel · 09-08-2008, 01:05 AM

Thank you for these responses! SATA was introduced in 2003. So I should not have been surprised that a 2002 XP Professional installation disk fails to install on a 1 TB Seagate SATA drive. I likely need to install some newer Windowze driver before I could install this XP. I don't know how to do so, but I will install CentOS as host and VMware with Windows clients. This is what I already have - except I want to eliminate several low capacity internal disk drives.

My CentOS 5.x is installed on an internal 80 GB SATA 150 drive. I have used it like that for a year. I am awaiting the response of support@escusa.com what changed since version 1.0 of their P4M900T-M motherboard supplied BIOS. The statement Linux calculates its own disk geometry helps! As my 80 GB SATA 150 disk works, the new 1 TB Seagate ST31000340AS should work as its replacement! I bought this disk from Brain Power, Inc. two weeks ago. I am waiting their response to my email. I only have their address and email, neither phone nor FAX number. If the answer does not arrive tomorrow, I'll get an RMA from Seagate and ship this S/N=9QJ1GASA back for repair under its 5 year warranty. I am confident now, this Seagate disk is to blame for my described problems!

- Reinhard

farslayer · 09-09-2008, 01:36 PM

could just as easily be a controller issue with the specific SATA controller on that motherboard, not properly supporting a drive that size... VIA seems to have all the information on the VT8237A very generalized. I am unable to find specifics on the limitations of the chipset...

Ahh I see why I can't find the info I want on that chipset...

Code:

* VIA Technologies, Inc. VT8235, VT8237R, VT8237A, VT8251
+    Datasheet: available on request and under NDA from VIA
+
 Authors:
 	Ky�sti M�lkki <kmalkki@cc.hut.fi>,
 	Mark D. Studebaker <mdsxyz123@yahoo.com>,
@@ -39,6 +42,8 @@ Your lspci -n listing must show one of t
  device 1106:8235   (VT8231 function 4)
  device 1106:3177   (VT8235)
  device 1106:3227   (VT8237R)
+ device 1106:3337   (VT8237A)
+ device 1106:3287   (VT8251)

You can easily test the drive with Seatools to see if it's defective, as a matter of fact Seagate will probably insist on it before providing a RMA number.

You can download a bootable seatools iso here: http://www.seagate.com/www/en-us/sup...loads/seatools
Select DOS as your system type, agree to the terms, fill in their little survey then choose the bootable .iso download.

This is a rather interesting snippet as well..

Quote:

The Barracuda 7200.11 is a native SATA product and supports speeds of SATA 3.0Gb/s. Unlike some of the first batches of SATA drives, a native drive will bypass the legacy Task File reads and writes, as well as avoiding the limitation of 133 Mbytes/sec for Ultra DMA Mode 6 transfers. While the 7200.11 supports SATA 3Gb/s transfer speeds, it is backwards compatible with SATA 1.5GB/s hosts. SATAII mode is also supported by the drive and it is enabled by default. Provided your motherboard supports it, it would be a good idea to leave it this way, but be sure to flip the jumper position if not. One of our forum readers in the past had an issue with a MSI P4N Diamond (no SATAII support) and a Seagate drive and changing the jumper corrected his problem of the drive not being detected.

I don't think your controller is SATA II so I'd have to ask if you saw a jumper listed in the Hard Drive manual that should be changed to make the Drive compatible with your older controller ? Check Page 22 of the Hard Drive owners manual.

A couple other items to look at anyway..

Electro · 09-09-2008, 04:34 PM

farslayer, there is no such thing as SATA II and this is what sata.org have said too. There are SATA 1.5 Gb or SATA 3 Gb. Though manufactures are using names for SATA to refer to different generations like they did for IDE even though the naming is not industry standard like 1.5 Gb and 3 Gb.