Edge backup with Seagate Travan tape.
Ulven
there is a tape inside it..
Yes, i am new to unix :-)
Anyone have any tip for me?
q General Resource Information
qqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqx
x Resource Type Tape Drive
x Resource Name [tape0 ] Change as appropriate
x Description [Seagate STT20000N 7.30 ]
x Changer Assoc [Standalone Device]
x Interface [SCSI ]
x Control Node [/dev/xStp0 ]
x q Tape Drive Information
qqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqx
x Data Node [/dev/rStp0 ] [A] TapeAlert(tm)
Support
x No Rewind Node [/dev/nrStp0 ]
x Tape Block Size [-1 ] [C] Partition
x Locate Threshhold [-1 ] [ Manual Check ]
x
x
xq Default Backup Properties
qqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqx
x Volume Size (K) [0 ] [S] Compression
x Edge Block Size [64 ] [Y] Double Buffering
Ulven
# @(#) tape.dfl.src 25.1 94/07/25
#
# Copyright (C) 1988-1994 The Santa Cruz Operation, Inc.
# All Rights Reserved.
# The information in this file is provided for the exclusive use of
# the licensees of The Santa Cruz Operation, Inc. Such users have the
# right to use, modify, and incorporate this code into other products
# for purposes authorized by the license agreement provided they
include
# this notice and the associated copyright notice with any such
product.
# The information in this file is provided "AS IS" without warranty.
#
# default/tape - tape device: tape(C)
#
# for cartridge tapes, use /dev/xct0
# for floppy tapes, use /dev/xctmini
#
device = /dev/nrStp0
curdriver=Stp0
Is this correct?
/FReddy Ulvseth
"Ulven" <fre...@REMOVESPAMulven.com> wrote in message
news:og2C9.13459$TK2.1...@juliett.dax.net...
Ulven
/dev/xct0 linked to /dev/xStp0
/dev/rct0 linked to /dev/rStp0 and /dev/rmt/0b
/dev/nrct0 linked to /dev/nrStp0 and /dev/rmt/0bn
"Ulven" <fre...@REMOVESPAMulven.com> wrote in message
news:On2C9.13460$TK2.1...@juliett.dax.net...
Bill Vermillion
Ulven <fre...@ulven.com> wrote:
>Can anyone see whats wrong here, why cant the program find any medium when
>there is a tape inside it..
>Yes, i am new to unix :-)
>Anyone have any tip for me?
Start at the lowest level to see if the drive is working before
trying the application.
Take some file in the system and try this
dd if=<inputfilename> of=<nameoftapedevice>
eg if=/etc/termcap of=/dev/rStp0
If that fails with an error message then you have the tape drive
configiured improperly or it is broken.
If it works then change to some tmp directory so you don't clobber
anything and do this.
dd if=<nameoftapedevice> of=temcapcopy
Then if that works do this
diff /etc/termcap termcapcopy.
If you get no errors the tape drive is working. Then you can
sort out why your Edge configuration is not working.
I have a gut feeling you'll get errors on the first dd.
Bill
--
Bill Vermillion - bv @ wjv . com
Ulven
Here's the resault:
# dd if=/var/log/backup.txt of=/dev/rStp0
1+0 records in
1+0 records out
#_
Then:
# dd if=/dev/rStp0 of=/var/log/temp
# pwd
/var/log
# ls
backup.txt
# _
# diff /etc/termcap termcapcopy
diff: cannot stat termcapcopy: No such file or directory (error 2)
/Ulven
http://freddy.ulven.com
"Bill Vermillion" <b...@wjv.comREMOVE> wrote in message
news:H5sE5...@wjv.com...
Bill Vermillion
Ulven <fre...@REMOVESPAMulven.com> wrote:
>Hi Bill.
>Here's the resault:
># dd if=/var/log/backup.txt of=/dev/rStp0
>1+0 records in
>1+0 records out
>#_
Who knows how bitg backup.txt is. I just suggested /etc/termcap
as it is decent size.
The concept is to use dd to put a file to tape. 'if' is the name of
the input file and in this case 'of' is the name of the tape.
The we reverse this and use 'if' for the tape devcie and 'of' for
the name of the file we will create as data is taken from the tape.
Then we compare, using diff, to see if the contents of the file
we put on tape match what we just took off the tape.
>Then:
># dd if=/dev/rStp0 of=/var/log/temp
># pwd
>/var/log
># ls
>backup.txt
># _
># diff /etc/termcap termcapcopy
>diff: cannot stat termcapcopy: No such file or directory (error 2)
You have to diff the files you put to tape and take from tape.
I said eg dd if=/etc/termcap eg meaning 'for example. It's a
standard reference but perhaps I should have said e.g. - but
everyone seem to skip that now.
Ulven
Here's the correct version:
# dd if=/etc/termcap of=/dev/rStp0
288+1 records in
288+1 records out
# dd if=/dev/rStp+ of=/termcapcopy
dd: cannot open /dev/rStp+: No such file or directory (error 2)
# dd if=/dev/rStp0 of=termcapcopy
289+0 records in
289+0 records out
# diff /etc/termcap termcapcopy
Warning: missing newline at end of file termcapcopy
3477a3478
something wrong here... does this tells you anything?
/Freddy
"Bill Vermillion" <b...@wjv.comREMOVE> wrote in message
news:H5t2B...@wjv.com...
Ulven
to: DAT
Before i used generic driver.
Here's the resault:
# dd if=/etc/termcap of=/dev/rStp0
288+1 records in
288+1 records out
# dd if=/dev/rStp0 of=termcapcopy
289+0 records in
289+0 records out
# diff /etc/termcap termcapcopy
Warning: missing newline at end of file termcapcopy
3477a3478
>
/Freddy
"Bill Vermillion" <b...@wjv.comREMOVE> wrote in message
news:H5t2B...@wjv.com...
Tony Lawrence
> ahh...
>
> Here's the correct version:
>
> # dd if=/etc/termcap of=/dev/rStp0
> 288+1 records in
> 288+1 records out
> # dd if=/dev/rStp+ of=/termcapcopy
> dd: cannot open /dev/rStp+: No such file or directory (error 2)
> # dd if=/dev/rStp0 of=termcapcopy
> 289+0 records in
> 289+0 records out
> # diff /etc/termcap termcapcopy
> Warning: missing newline at end of file termcapcopy
> 3477a3478
>
> something wrong here... does this tells you anything?
>
It tells me you don't understand dd :-)
If you want to test something like that, do it like this:
cp /etc/termcap /tmp/this
tar cvf /dev/rStp0 /tmp/this
rm /tmp/this
tar xvf /dev/rStp0
diff /tmp/this /etc/termcap
The problem with using dd for this is the differing blocks- notice you
get 288 full blocks plus a PARTIAL block going out to the tape, but the
tape can only give you full blocks back. - therefore there is null data
in the read, and that naturally won't pass diff.
--
Please note new phone number: (781) 784-7547
Tony Lawrence
Unix/Linux Support Tips, How-To's, Tests and more: http://aplawrence.com
Free Unix/Linux Consultants list: http://aplawrence.com/consultants.html
Bill Vermillion
Ulven <notfredb...@ulven.com> wrote:
>ahh...
>Here's the correct version:
># dd if=/etc/termcap of=/dev/rStp0
>288+1 records in
>288+1 records out
># dd if=/dev/rStp+ of=/termcapcopy
>dd: cannot open /dev/rStp+: No such file or directory (error 2)
># dd if=/dev/rStp0 of=termcapcopy
>289+0 records in
>289+0 records out
># diff /etc/termcap termcapcopy
>Warning: missing newline at end of file termcapcopy
>3477a3478
>something wrong here... does this tells you anything?
Yes - it says your tape drive is working.
The tape filled the last block on the dd out and that's why you
got 289+0 on reading it in. And it says there was a missing new
line at the end of termccapcopy - which is normal as the tape would
write full blocks.
We now know that your tape drive is working. Step 1 is done.
Your drive is being accessed properly by the system. Now you just
have to determine where it is failing after that.
Use the steps that Tom put up to step through each process to
determine where it falls down.
So you don't have too fool with mkdev tape or anything like that.
Bill
Bill Vermillion
>> # dd if=/etc/termcap of=/dev/rStp0
>> 288+1 records in
>> 288+1 records out
>> # dd if=/dev/rStp+ of=/termcapcopy
>> dd: cannot open /dev/rStp+: No such file or directory (error 2)
>> # dd if=/dev/rStp0 of=termcapcopy
>> 289+0 records in
>> 289+0 records out
>> # diff /etc/termcap termcapcopy
>> Warning: missing newline at end of file termcapcopy
>> 3477a3478
>> something wrong here... does this tells you anything?
>It tells me you don't understand dd :-)
He did exactly as I asked. I just wanted to see if the device was
working at the very lowest level. This it the first time we've
made it this far.
So don't be hard on him for doing what I told him to do.
I fully expected him not to get anything back because perhaps the
tape drive was misconfigured.
Tony Lawrence
> In article <ardr64$f2l$1...@pcls4.std.com>,
> Tony Lawrence <to...@pcunix.com> wrote:
>
>>Ulven wrote:
>>
>>>ahh...
>>>
>>>Here's the correct version:
>
>
>>># dd if=/etc/termcap of=/dev/rStp0
>>>288+1 records in
>>>288+1 records out
>>># dd if=/dev/rStp+ of=/termcapcopy
>>>dd: cannot open /dev/rStp+: No such file or directory (error 2)
>>># dd if=/dev/rStp0 of=termcapcopy
>>>289+0 records in
>>>289+0 records out
>>># diff /etc/termcap termcapcopy
>>>Warning: missing newline at end of file termcapcopy
>>>3477a3478
>
>
>>>something wrong here... does this tells you anything?
>
>
>>It tells me you don't understand dd :-)
>
>
> He did exactly as I asked. I just wanted to see if the device was
> working at the very lowest level. This it the first time we've
> made it this far.
>
> So don't be hard on him for doing what I told him to do.
I didn't notice that you had instructed him to do this.
However, I wasn't being hard on him.
That's why I put the damn smiley in - to tell him where the
misunderstanding was without slapping him up the side of the head!
I followed that (although you cut it here) with an explanation of WHY dd
won't work for that.
Cripes. What the heck was I supposed to say? The guy didn't understand
why the diff failed. He apparently thought his tape drive was the
culprit. So should I have said "Well, rest assured that the problem
isn't your tape drive, at least not from this test. I can't really see
the point of this test, because it is almost guaranteed to fail." ?
No, because that would probably have been just as insulting. I guess
you just can't ever say that anyone is wrong, confused, misleading,
careless or whatever nowadays because to do so might bruise somebody's
fragile ego.
It's time people got over this crap. Everyone of us is wrong, careless,
dumb, misinformed, confused and whatever regularly. If I had to go into
a blue funk everytime I did or said something wrong, I'd be pretty damn
depressed.
And again: If Mr. Ulven's ego is that fragile, I haven't noticed it. In
fact, I'd guess he's a well adjusted, confident person who is well aware
of everything I said in the previous paragraph.
Bill Vermillion
Tony - you know me well enough to know that was not really a
criticism.
I will say he is quite patient and on some other posts he hadn't
quite followed the instructions. I just felt the dd was the
lowest level way to make sure at least something was going to tape.
>Cripes. What the heck was I supposed to say? The guy didn't
>understand why the diff failed. He apparently thought his tape
>drive was the culprit.
I think he had mails from Tom and Jeff - and 3 or 4 others.
Everything had failed - and we couldn't even be sure he had the
right ID on the tape.
>It's time people got over this crap. Everyone of us is wrong,
>careless, dumb, misinformed, confused and whatever regularly.
What do you mean "every one of us" ;-) I can think of one or two
on this list who are never wrong. I wish I were that good.
> If I had to go into a blue funk everytime I did or said
>something wrong, I'd be pretty damn depressed.
You take what I said to much to heart.
Will this help?
))))
))))
:::: ))))
:::: ))))
---- ))))
---- ))))
:::: ))))
:::: ))))
))))
))))
Bill
Bela Lubkin
> In article <are4tq$cd4$1...@pcls4.std.com>,
> Tony Lawrence <to...@pcunix.com> wrote:
> >It's time people got over this crap. Everyone of us is wrong,
> >careless, dumb, misinformed, confused and whatever regularly.
>
> What do you mean "every one of us" ;-) I can think of one or two
> on this list who are never wrong. I wish I were that good.
Howdy. I can guess who you're referring to, and if you think I'm never
wrong, that's only because of good stage management. When I think (or
know) I'm wrong, I disclaim the hell out of what I'm saying. Make it
sound all speculative so that when it turns out wrong, well, I wasn't
really standing behind that position anyway. ;-}
>Bela<
Ulven
I can handle criticism when i get the facts to what i did wrong, no worry
about that.. :-)
I'm trying to learn at the same time as solving the problem here, and now i
understand more about dd command.
Thanks Tony!
One of the problem with my tape was that i had configured it as Generik
SCSI-1 or SCSI-2 drive, i have now changed it to DAT and i have run one
complete backup on one of our servers, but it fails with error code 4. i
guess that this is bad tapes..
I have had lots of problems with travan tapes, they stinks :-)
I want to say thanks to all of you that have give me help, support, and
shit! :-) i'm longer to a sollution now, but still not i goal.
I'm going to do some cleaning on my tapes, retensions and maby new tapes to
see if that helps.
I might come back with some questions..
Thanks all of you!
/Freddy Ulvseth "Ulven".
"Bill Vermillion" <b...@wjv.comREMOVE> wrote in message
news:H5unH...@wjv.com...
Tony Lawrence
> Hi both of you...
>
> I can handle criticism when i get the facts to what i did wrong, no worry
> about that.. :-)
> I'm trying to learn at the same time as solving the problem here, and now i
> understand more about dd command.
> Thanks Tony!
>
> One of the problem with my tape was that i had configured it as Generik
> SCSI-1 or SCSI-2 drive, i have now changed it to DAT and i have run one
> complete backup on one of our servers, but it fails with error code 4. i
> guess that this is bad tapes..
> I have had lots of problems with travan tapes, they stinks :-)
Yes, they do. They just aren't worth having. Though I have been told
that the new models don't suffer from the same problems as the old.
Personally, I still avoid them.
If your data is under 9 gb (compressed), you can use dvdram (
http://aplawrence.com/Reviews/dvdram.html ) which is a WONDERFUL backup
medium. Unfortunately, until 5.0.7 comes along, you need a scsi dvdram
writer, which is getting VERY hard to find. %.0.7 supposedly will
support ide which is easy to find and getting very inexpensive.
On the tape type: almost nothing but 9 tracks should ever be configured
as generic scsi. Travans shoud use dat, DLT's should use qic (if I
remember right), but very little is generic. I don't understand what
the underlying differences in the drivers are (maybe Bela or one of the
supertar folks will bite on that bait ) but I do know that the wrong
choice breaks things badly.
Dave Dickerson
wrote:
With over 400+ Travan's in the field (not my doing) all configured as
SCSI (cause they are), I'm interested in finding out why they should
have been defined as DAT. Somebody please elaborate. Thanks.
DDinAZ
Dave Dickerson
wrote:
[snip]
>On the tape type: almost nothing but 9 tracks should ever be configured
>as generic scsi. Travans shoud use dat, DLT's should use qic (if I
>remember right), but very little is generic. I don't understand what
>the underlying differences in the drivers are (maybe Bela or one of the
>supertar folks will bite on that bait ) but I do know that the wrong
>choice breaks things badly.
>
>
>--
>
>Please note new phone number: (781) 784-7547
>
>Tony Lawrence
>Unix/Linux Support Tips, How-To's, Tests and more: http://aplawrence.com
>Free Unix/Linux Consultants list: http://aplawrence.com/consultants.html
>
I'm very interested in finding out why Travan's should be configured
as DAT opposed to SCSI. I have 400+ Travan's in the field all defined
as SCSI and now I'm concerned about it! (standard Travan rants aside.)
DDinAZ
Bill Vermillion
Ulven <notfredb...@ulven.com> wrote:
>One of the problem with my tape was that i had configured it as
>Generik SCSI-1 or SCSI-2 drive, i have now changed it to DAT and
>i have run one complete backup on one of our servers, but it
>fails with error code 4. i guess that this is bad tapes..
>I have had lots of problems with travan tapes, they stinks :-)
Try to make your next goal getting rid of the Travan and getting
a DAT or VXA. The approx $700 you'd pay would probalby be less
than the time you spend on this mess :-)
Glad you are now getting backup.
>I'm going to do some cleaning on my tapes, retensions and maby
>new tapes to see if that helps.
Hopefully it will. Do try to conince management to update to
another tape system. Refurbed units are availabe too if the
capacity fits.
Bela Lubkin
> On the tape type: almost nothing but 9 tracks should ever be configured
> as generic scsi. Travans shoud use dat, DLT's should use qic (if I
> remember right), but very little is generic. I don't understand what
> the underlying differences in the drivers are (maybe Bela or one of the
> supertar folks will bite on that bait ) but I do know that the wrong
> choice breaks things badly.
I'll nibble. There are different code paths in the Stp driver (in fact
different source files) for certain functions, according to which drive
type you told it. The differences are distillations of behavioral
differences observed by someone writing the driver 5-10 years ago.
Little nuggets of wisdom like "DAT drives support tape partitioning, QIC
don't". I have no idea whether they still apply today -- one would
expect that over time, _all_ the newer drives would support _all_ the
command sets.
I'd have to read the source to be more specific, don't have time right
now. (Even the part about who supports tape partitioning is just an
example of the _sorts_ of things encoded by the type; I don't know
_which_ types the driver actually thinks support partitioning.)
I would guess the supertar folks have a much better idea of what the
type switches really do...
>Bela<
D. Thomas Podnar
it as a DAT.
Tom Podnar
Tony Lawrence
It's just been my experience that they don't work well (not that they
EVER really work well) that way.
Dave Dickerson
wrote:
Well, that's true, Tony, but as far as I can tell they don't work well
whether or not they were configured as "SCSI" or "DAT". From what I'm
reading in this thread (between the lines and otherwise) makes me
believe I've not committed a fatal blunder and there's some comfort
in that. If any guru's want to counter that statement, please wait
until after the holidays...thanks.
DDinAZ