/dev/nst0 trouble

Issues related to hardware problems
sub31
Posts: 35
Joined: 2007/09/26 19:05:20
Location: Tyumen, Russia

/dev/nst0 trouble

Postby sub31 » 2008/12/24 17:13:52

CentOS 5.2
I've some tape drives to save backups. It's realy long and not intresting process. After upgrading to CentOS 5.2 I've some troubles with drive LTO-2. This drive cannot longer write more then 100Gb to tapes.
I remove that drive and next drive LTO-2 cannot longer write more then 100Gb. I install first drive back and I still cannot write more then 100Gb to tapes in any LTO-2 drive installed as st0.
I don't understand what's wrong with [n]st0? Why [n]st[1-9] is better?

kentyler
Posts: 161
Joined: 2007/06/20 13:31:05
Location: Northeast Ohio
Contact:

/dev/nst0 trouble

Postby kentyler » 2008/12/26 21:02:57

Did you configure the tape device or format the tape? This may help.

http://linux.die.net/man/8/stinit

Please provide the output of:
mt -f /dev/st0 status
and
cat /proc/scsi/scsi

sub31
Posts: 35
Joined: 2007/09/26 19:05:20
Location: Tyumen, Russia

Re: /dev/nst0 trouble

Postby sub31 » 2009/01/09 20:31:02

# mt -f /dev/nst0 status
SCSI 2 tape drive:
File number=0, block number=0, partition=0.
Tape block size 0 bytes. Density code 0x42 (LTO-2).
Soft error count since last status=0
General status bits on (81010000):
EOF ONLINE IM_REP_EN

# mt -f /dev/nst1 status

SCSI 2 tape drive:
File number=1, block number=0, partition=0.
Tape block size 0 bytes. Density code 0x42 (LTO-2).
Soft error count since last status=0
General status bits on (81010000):
EOF ONLINE IM_REP_EN


# cat /proc/scsi/scsi
Attached devices:
Host: scsi0 Channel: 00 Id: 00 Lun: 00
Vendor: AMCC Model: 9500S-8 DISK Rev: 2.08
Type: Direct-Access ANSI SCSI revision: 03
Host: scsi1 Channel: 00 Id: 00 Lun: 00
Vendor: AMCC Model: 9500S-8 DISK Rev: 2.08
Type: Direct-Access ANSI SCSI revision: 03
Host: scsi2 Channel: 00 Id: 10 Lun: 00
Vendor: BROWNIE Model: 1600U3P Rev: 0001
Type: Direct-Access ANSI SCSI revision: 03
Host: scsi2 Channel: 00 Id: 10 Lun: 01
Vendor: BROWNIE Model: 1600U3P Rev: 0001
Type: Direct-Access ANSI SCSI revision: 03
Host: scsi4 Channel: 00 Id: 02 Lun: 00
Vendor: CERTANCE Model: ULTRIUM 2 Rev: 1914
Type: Sequential-Access ANSI SCSI revision: 03
Host: scsi5 Channel: 00 Id: 03 Lun: 00
Vendor: HP Model: Ultrium 2-SCSI Rev: F48D
Type: Sequential-Access ANSI SCSI revision: 03
Host: scsi6 Channel: 00 Id: 00 Lun: 00
Vendor: DotHill Model: R/Evo 2330-2R Rev: J200
Type: Enclosure ANSI SCSI revision: 04
Host: scsi7 Channel: 00 Id: 00 Lun: 00
Vendor: DotHill Model: R/Evo 2330-2R Rev: J200
Type: Enclosure ANSI SCSI revision: 04
Host: scsi6 Channel: 00 Id: 00 Lun: 01
Vendor: DotHill Model: R/Evo 2330-2R Rev: J200
Type: Direct-Access ANSI SCSI revision: 04
Host: scsi7 Channel: 00 Id: 00 Lun: 01
Vendor: DotHill Model: R/Evo 2330-2R Rev: J200
Type: Direct-Access ANSI SCSI revision: 04


Thanks for stinit man page. I don't use it before. I use simple 'mt -f /dev/nst0 defblksize 0' command and look like before your post. Later I learn this specific for CentOS or linux mechanism. But this is not change situation. Before making archive I use mt -f /dev/nst0 rewind command. (Just to remove another questions)
I don't format tapes before use. I can change drives Centrance and HP by busses, but it not provide solution. /dev/nst0 is not write more then 100Gb, and /dev/nst1 is write about 185Gb without troubles for the same tape. I'm using dump program, and tar.
Just for a fix this situation I plan use Exabyte drive as /dev/nst0 and exams a work of /dev/nst2 later. But this computer in productivity and access for changes are limited strongly.

I don't remember when this situation happend. Before CentOS 5 on this computer work RHEL 3 and all drives work well. After install CentOS 5.0 I think drives work slow but use all capacity. After upgrade to CentOS 5.1 we don't use slow drive and cannot describe correct. I'm not sure when this trouble happend. I think this trouble in kernel or near. Now at this computer work CentOS 5.2.

sub31
Posts: 35
Joined: 2007/09/26 19:05:20
Location: Tyumen, Russia

Re: /dev/nst0 trouble

Postby sub31 » 2009/01/10 06:53:47

In previous message I provide wrong tape status becouse this tapes were in use and I try to aproximate recorded in log status.
Now is correct status getted from drives directly.

# mt -f /dev/nst0 status

SCSI 2 tape drive:
File number=0, block number=0, partition=0.
Tape block size 0 bytes. Density code 0x42 (LTO-2).
Soft error count since last status=0
General status bits on (41010000):
BOT ONLINE IM_REP_EN

# mt -f /dev/nst1 status

SCSI 2 tape drive:
File number=0, block number=0, partition=0.
Tape block size 0 bytes. Density code 0x42 (LTO-2).
Soft error count since last status=0
General status bits on (41010000):
BOT ONLINE IM_REP_EN

sub31
Posts: 35
Joined: 2007/09/26 19:05:20
Location: Tyumen, Russia

Re: /dev/nst0 trouble

Postby sub31 » 2009/01/31 12:09:23

After some experiments I've one LTO-2 that working good as /dev/nst1.
I'm install DAT80 drive as /dev/nst0 and remove broken LTO-2 device.
This broken device from Quantum CERTANCE. Write around 100GB very long time and return write error. The same tape in HP 460 writes good much more data and much faster.