Topic: Constant Tape Access - Problem

[rick1908]

Software: SME 5.6 all updates, flexbackup, software RAID1

Hardware:
hda: Maxtor 40 gig HD
hdb: NEC CD ROM Drive
hdc: Maxtor 40 gig HD
hdd: Seagate STT20000A Tape Drive

Problem is that whenever I put a tape into the drive the tape is accessed by the tape drive every few (1-2, varries) minutes. What I mean by accessed is that the tape starts and resets itself like when you insert a tape for the first time or when the computer starts up. It is almost like I am pulling out and inserting the tape every so often(that's what it sounds like anyhow). Backups also fail with the following log:

|------------------------------------------------
| Trying "mt defblksize" instead of "mt setblk"
| Doing level 0 backup of all using dump
| Retensioning tape...
| Rewinding & erasing tape...
| Creating index key 200312220203.11
| Tape #0
| Filesystems = / /boot
|------------------------------------------------
At block 21.
|------------------------------------------------
| File number 1, index key 200312220203.11
| Backup of: /
| Date of this level 0 backup: Mon Dec 22 02:03:48 2003
| Date of last level 0 backup: the epoch
|------------------------------------------------
| (dump -0 -b 10 -a -f - / | gzip -4) | buffer -m 3m -s 10k -u 100 -t -p 75 -B \
|  -o /dev/nst0
|------------------------------------------------
  DUMP: Date of this level 0 dump: Mon Dec 22 02:03:48 2003
  DUMP: Dumping /dev/md1 (/) to standard output
  DUMP: Added inode 8 to exclude list (journal inode)
  DUMP: Added inode 7 to exclude list (resize inode)
  DUMP: Label: /
  DUMP: mapping (Pass I) [regular files]
  DUMP: mapping (Pass II) [directories]
  DUMP: estimated 10511532 tape blocks.
  DUMP: Volume 1 started with block 1 at: Mon Dec 22 02:04:20 2003
  DUMP: dumping (Pass III) [directories]
  DUMP: dumping (Pass IV) [regular files]
  DUMP: 1.75% done at 614 kB/s, finished in 4:40
  DUMP: 3.45% done at 604 kB/s, finished in 4:39
  DUMP: 4.75% done at 554 kB/s, finished in 5:00
  DUMP: 6.23% done at 545 kB/s, finished in 5:01
  DUMP: 7.54% done at 528 kB/s, finished in 5:06
  DUMP: 8.80% done at 513 kB/s, finished in 5:10
  DUMP: 10.14% done at 507 kB/s, finished in 5:10
buffer (writer): write of data failed: Input/output error
bytes to write=10240, bytes written=-1, total written     967170K
  DUMP: Broken pipe
  DUMP: The ENTIRE dump is aborted.
|------------------------------------------------
| Backup start: Mon Dec 22 02:03:48 2003
| Backup end:   Mon Dec 22 02:40:06 2003
|------------------------------------------------
/dev/nst0: Input/output error
mt: The device is offline (not powered on, no tape ?).
|------------------------------------------------
| File number 2, index key 200312220203.11
| Backup of: /boot
| Date of this level 0 backup: Mon Dec 22 02:40:17 2003
| Date of last level 0 backup: the epoch
|------------------------------------------------
| (dump -0 -b 10 -a -f - /boot | gzip -4) | buffer -m 3m -s 10k -u 100 -t -p \
|  75 -B -o /dev/nst0
|------------------------------------------------
  DUMP: Date of this level 0 dump: Mon Dec 22 02:40:17 2003
  DUMP: Dumping /dev/md0 (/boot) to standard output
  DUMP: Added inode 8 to exclude list (journal inode)
  DUMP: Added inode 7 to exclude list (resize inode)
  DUMP: Label: /boot
  DUMP: mapping (Pass I) [regular files]
  DUMP: mapping (Pass II) [directories]
  DUMP: estimated 4693 tape blocks.
  DUMP: Volume 1 started with block 1 at: Mon Dec 22 02:40:18 2003
  DUMP: dumping (Pass III) [directories]
  DUMP: dumping (Pass IV) [regular files]
buffer (writer): write of data failed: Input/output error
bytes to write=10240, bytes written=-1, total written          0K
  DUMP: Broken pipe
  DUMP: The ENTIRE dump is aborted.
|------------------------------------------------
| Backup start: Mon Dec 22 02:40:17 2003
| Backup end:   Mon Dec 22 02:40:21 2003
|------------------------------------------------
/dev/nst0: Input/output error
mt: The device is offline (not powered on, no tape ?).
|------------------------------------------------
| Rewinding...
/dev/nst0: Input/output error
mt: The device is offline (not powered on, no tape ?).
|------------------------------------------------


Sorry to post the whole log but I wasn't sure what is important.

BTW, the two Maxtor drives are on seperate IDE channels because I read somewhere (here?) that this would speed up data transfers. Is this correct? If not and this is determined to be the problem, how do I change it back (software wise: RAID, tape, CD-Rom, etc.)?

Anyway, thanks for your help.

Sincerely,
Rick Evans

[ldkeen]

Have you tried using a brand new tape? You could also try running a cleaning tape through the drive. What is the output of #mt -f /dev/nst0 status ?

[rick1908]

Output for "mt -f /dev/nst0 status" is:

SCSI 2 tape drive:
File number=-1, block number=-1, partition=0.
Tape block size 0 bytes. Density code 0x0 (default).
Soft error count since last status=0
General status bits on (50000):
 DR_OPEN IM_REP_EN

I tried using several different tapes but not a completly new one.

Thanks for your quick reply.

Sincerely,
Rick

[ldkeen]

It definitely appears to be a problem with the tape. The DR_OPEN from status means either there is no tape in the drive or the tape is faulty.

[rick1908]

Sorry there was no tape in the drive when I executed the command (Sorry about that). Here are the results with a tape in the drive:

SCSI 2 tape drive:
File number=0, block number=0, partition=0.
Tape block size 512 bytes. Density code 0x47 (TR-5).
Soft error count since last status=0
General status bits on (41010000):
 BOT ONLINE IM_REP_EN

I also tried with a brand new tape (opened the package   ) and I get the tape tape access results described in my original post.

Please...please...please tell me you have some experience with this kind of error.

Thanks again,
Rick

[bushinc]

> the tape is accessed by the tape drive every few
> minutes.

This doesn't seem right.  I must admit that I've only used SCSI drives, but I would guess that something on the IDE bus is conflicting with the tape drive.  Probably while you are backing up, the same conflict occurs and causes the tape drive to malfunction.
Any way you can test without anything else on the same bus... You mentioned moving the two drives to the same bus, I haven't tried this, so I don't know what steps are required.  Be sure to backup first  

> At block 21.
This seems strange.  The backups on my SME5.6/6.0 servers start at Block 2.

>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
  DUMP: 10.14% done at 507 kB/s, finished in 5:10
buffer (writer): write of data failed: Input/output error
bytes to write=10240, bytes written=-1, total written     967170K
  DUMP: Broken pipe
  DUMP: The ENTIRE dump is aborted.
<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<

I receive this error on my SCSI tape drives if:
1. I have a bad tape in the drive
2. I have dirty tape heads
3. The SCSI bus (my config) is in an unknown state and requires me to reboot the server.  This has happened on occasion  with older version of SME.
4. The tape drive is broken and needs to be replaced

If you get the problem solved, you may want to try increasing the block size of your backups.  I am currently used 32k on my SME5.6 servers, 64k on my SME6 servers.  I also increased the buffer, though this takes up memory.  I have dramatically decreased the amount of time it takes to backup with these changes. YMMV
The templates that can be modified are:
/etc/e-smith/templates/etc/flexbackup.conf/10Blocksize
/etc/e-smith/templates/etc/flexbackup.conf/10Buffering

Sorry I don't have an absolute solution, but I hope this helps.

Steve

[rick1908]

Thanks for the information, at least I have a starting point. If/When I find a solution, I will post the results for use bu others. Thanks for the help.

Sincerely,
Rick

[ldkeen]

Rick,
Just curious, have the backups ever worked in the past or is this a new setup.

[rick1908]

Yes the setup did work before. I used to use the server at my office, we got new equipment, I took the old server.

Anyhow, the only change to the server from it's previous state was I replaced a drive that was getting raid errors, and I upgraded from 5.5 to 5.6. I followed D. Mays How-To when recovering and re-establishing the mirror. I would have included that in my original post but I just did not suspect that it would have an affect on the tape backup.

Does this change things?

Sincerely,
Rick