Topic: Seagate ST220000A Drive installation problem

[John Lewis]

Greetings,

I've just installed an IDE Seagate STT220000A drive, following Darrell's instructions (set it for master, and cdrom for slave)

It did not appear to work, so I then, based on suggestions in these forums did the following:

First, I replaced nst0 with nht0 in /etc/flexbackup.conf

I also ran the following:

/sbin/e-smith/db configuration setprop backup Device /dev/nht0
/sbin/e-smith/signal-event conf-backup

But now get the following when I run flexbackup -newtape:

flexbackup version 0.9.8
/etc/flexbackup.conf syntax OK

|------------------------------------------------
| Trying "mt defblksize" instead of "mt setblk"
Error setting block size
Neither of these commands worked:
  mt -f /dev/nht0 setblk 0
  mt -f /dev/nht0 defblksize 0

It is a brand new tape.  Where/what can I do to confirm the server recognizes it  (it does see it in BIOS), and what else can I do to just get drive to write to the tape...

Thanks.

[John Lewis]

Here's what I see in my logs when I try to force a backup:

Nov  3 23:21:21 gluon e-smith[15899]: Backup aborted: No tape loaded
Nov  3 23:21:21 gluon e-smith[15899]: S10BackupNow=action|Event|BackupNow|Action|S10BackupNow|Start|1004847680 614959|End|1004847681 690230|Elapsed|1.075271

I have inserted a brand new Seagate STTM20 TR-5 cartridge...

[John Lewis]

After a reboot, here is what I'm getting in messages:

Nov  3 23:52:35 gluon kernel: ide-tape: hdc <-> ht0, 1000KBps, 6*54kB buffer, 5400kB pipeline, 130ms tDSC
Nov  3 23:52:35 gluon kernel: ide-tape: ht0: I/O error, pc = 34, key =  4, asc = 80, ascq = 86
Nov  3 23:52:35 gluon kernel: ide-tape: ht0: I/O error, pc =  1, key =  4, asc = 80, ascq = 86
Nov  3 23:52:35 gluon kernel: ide-tape: ht0: I/O error, pc = 34, key =  4, asc = 80, ascq = 86
Nov  3 23:52:35 gluon kernel: ide-tape: ht0: I/O error, pc =  1, key =  4, asc = 80, ascq = 86
Nov  3 23:52:35 gluon kernel: ide-tape: ht0: I/O error, pc = 34, key =  4, asc = 80, ascq = 86
Nov  3 23:52:35 gluon kernel: ide-tape: ht0: I/O error, pc =  1, key =  4, asc = 80, ascq = 86
Nov  3 23:52:35 gluon kernel: ide-tape: ht0: I/O error, pc =  1, key =  4, asc = 80, ascq = 86
Nov  3 23:53:32 gluon kernel: ide-tape: ht0: I/O error, pc = 34, key =  4, asc = 80, ascq = 86
Nov  3 23:53:32 gluon kernel: ide-tape: ht0: I/O error, pc =  1, key =  4, asc = 80, ascq = 86
Nov  3 23:53:32 gluon kernel: ide-tape: ht0: I/O error, pc = 34, key =  4, asc = 80, ascq = 86
Nov  3 23:53:32 gluon kernel: ide-tape: ht0: I/O error, pc =  1, key =  4, asc = 80, ascq = 86
Nov  3 23:53:32 gluon kernel: ide-tape: ht0: I/O error, pc = 34, key =  4, asc = 80, ascq = 86
Nov  3 23:53:32 gluon kernel: ide-tape: ht0: I/O error, pc =  1, key =  4, asc = 80, ascq = 86

[Bruce Friedman]

John,

I assume you searched the archives here and saw my notes on this same drive.  I saw the problems you mentioned.  I have also opened a bug report on this item with Mitel.

I have successfully used taper with nst0.  It does give errors in /var/log/messages, but does not seem to affect the files written to tape.

I have successfully backed up using nts0 with flexbackup - again the errors don't seem to corrupt the data.

I do not recommend the use of nht0 as you listed because that seems worse than using nst0 (I didn't try the backups with it, though - did you?).

[Dan G.]

Did you look at the How-to for Flexbackup on Darrell's site?  ---->  http://myezserver.com/docs/mitel/flexbackup-howto.html

I was using a SCSI version of that same drive, and initially got I/O errors.  The part of that How-to that seemed to fix it was:

     How to set flexbackup to always perform an erase

     /sbin/e-smith/db configuration setprop flexbackup erase_rewind_only false
     /sbin/e-smith/signal-event console-save

My backups run flawlessly now.

Good luck.

[John Lewis]

Thanks, I ran those, and still get the following when I try to backup:

Nov  4 22:36:26 gluon e-smith[21525]: Backup aborted: No tape loaded
Nov  4 22:36:26 gluon e-smith[21525]: S10BackupNow=action|Event|BackupNow|Action|S10BackupNow|Start|1004931384 693580|End|1004931386 43795|Elapsed|1.350215
Nov  4 22:37:46 gluon kernel: ide-tape: ht0: I/O error, pc = 34, key =  4, asc = 80, ascq = 86
Nov  4 22:37:46 gluon kernel: ide-tape: ht0: I/O error, pc =  1, key =  4, asc = 80, ascq = 86
Nov  4 22:37:46 gluon kernel: ide-tape: ht0: I/O error, pc = 34, key =  4, asc = 80, ascq = 86
Nov  4 22:37:46 gluon kernel: ide-tape: ht0: I/O error, pc =  1, key =  4, asc = 80, ascq = 86
Nov  4 22:37:46 gluon kernel: ide-tape: ht0: I/O error, pc = 34, key =  4, asc = 80, ascq = 86
Nov  4 22:37:46 gluon kernel: ide-tape: ht0: I/O error, pc =  1, key =  4, asc = 80, ascq = 86

I've tried two tapes.  When I insert them (forgive me, this is the first time I have worked with Travan's), nothing happens.  When I run flexbackup - newfile

I get:

flexbackup version 0.9.8
/etc/flexbackup.conf syntax OK

|------------------------------------------------
| Rewinding & erasing tape...
/dev/nht0: Input/output error
Error from mt rewind

And still more I/0 errors in the log.

Is it software or hardware?

[Dan G.]

Forgive me if this sounds elementary, but are you getting a power/activity light on the tape drive?  You say "nothing happens" --- if that is the case, this is a sign.  When you initially insert a Travan cartridge, the drive should immediately begin "feeling it up," and will make telltale whirring/seeking sounds, and the groovy light should blinky-blinky.  If this is not happening, you have a problem at the most basic hardware level.  Either the drive is defective, or a cable is disconnected or seated improperly.

[John Lewis]

No, not elementary at all.  When I stick in the tapes, nothing really happens.  No light etc.  It does sound like there is something going on, but its very faint.

However, when I stick in a cleaning cartridge, it immediately recognizes it and begins the cleaning process.  So somthing is working.  I've got the right tapes, so I guess I'll recheck all of the cables etc.  

Could this be a master/slave issue?

Why would I get kernel i/o issues if it was a hardware problem?

Thanks.

[Dan G.]

A properly functioning drive should begin seeking the beginning of the tape when a cartidge is inserted --- way before the kernel attempts to control the device.  It even does this if only the power connector is connected, an it is evident at bootup, before the OS is initialized.  This is all handled in the firmware of the drive, AFAIK.   Are you using the proper cartridge?  I don't know if a "wrong" one would physically fit in the drive or not --- but I only use the 10/20 GB Travan TR5 tapes, and they work perfectly.  Just another thing to check.

Now, I'm no kernel wizard, but I'd venture your I/O errors are exactly what you would expect to see if the kernel looks for the drive, and gets no response from it.  Try unhooking the device totally, and attempting a backup with no drive.  If you get an identical error message, you'll know.

[John Lewis]

A belated followup to this thread.

I sent the drive back to Seagate, and they replaced it.

Everything is working fine now.  When I load tapes, they successfully do their re-tensioning.  Backups perform on schedule, and I have successfully done a restore of several files as a test.

Thanks to everyone for their help and suggestions.

What is the general life of a TR-5 tape?  At $33 each, I'm hoping they will last some time before having to be replaced.

-JL