DAT drive recommendations?

[J. Clarke]

Thanks for the input. I hope to get a consensus out of many opinions, but
a sample of one is infinitely better than a sample of zero!


I do. In fact, constant retries in the verify step are the only evidence I
have so far that the drive is failing.


I do that as well.

get a few extra miles from.

I would do that if the situation came up, but it never has. In nearly 15
years of using various types of DAT, I have never had a tape fail or give
evidence of impending failure. (This is one of my reasons for preferring
tape. I have had two hard disks failed in that time -- one
catastrophically, with no warning -- and any number of CDR's that wrote
and tested fine but later proved to be unreadable.)

Just remember that DDS tapes are rated for a relatively low number of
passes--5000 for DDS4, compared to a million for DLT. Every time the tape
has to back up and restart that counts as a pass, so if it's not streaming
you're chewing up tape at a horrendous rate.

Further, I once had a DDS drive get a chip of something or other caught in
one of the heads and it trashed (nonrecoverable physical damage to the
tape) a number of tapes before I figured out that the problem was the
drive.

Using disk for backup, you have a number of disks just as you have a number
of tapes, and you rotate them just as you do tapes. So even if one of your
backups fails at the same time as the disk it's backing up you won't lose
more than two days. If losing two days is catastrophic for you then you
should be running mirrored servers each with a redundant array anyway.

I used to believe in tape for backup, but to tell the truth in small
installations it's not very attractive anymore.

 

[Al Dykes]

Backing up to tape is definitely the easiest and most cost effective way to
back up your data. It's when you have to do a restore is when you really
find out how reliable your tape backup really is or not. I have seen too
many times were people religiously do their tape backups and are in total
bliss till they have to do a restore, which fails a good portion of the
time. Tape is not as reliable as one would want to believe. Plus, for the
home user using DDS3 or DDS4 will find them numbing slow compared to using
an external SCSI or USB drive solely for system backups.



Rita

Having managed lots of small stand-alone risc servers over the years,
DDS tape drives have eaten more tapes than all others combined; (QIC,
9TK, DLT.) DLT has never eaten a tape, or failed a verify readback,
for me. I really disklke DDS.

I'm a fan of DLT. I've bought a few hundred DLT tapes in the name of
my company, on ebay. Several of the brands have lifetime no-fault
exchange warranties so how could I go wrong ?

DLTIV drives are down to as low as $700 (DEC-branded, new, with
warranty) on pricewatch.com

These days, at home, I'm doing image backups of a laptop and a couple
desktops to a big SATA disk in one machine (a couple generations of
images for each machine). I sync the "My Documents" folder tree
between a laptop and a desktop every time I dock or undock, and I burn
work product into many generations of CDR disks.

 

[Al Dykes]

I can relate to your sentiments on this.


My recommendations would be a Sony SDT-11000. This is an external SCSI unit
that will give you the most reliability for the buck. As a cautionary note
if you must use tape, backup and compare often. Also, rotate tapes on a
regular basis and don't be afraid to trash tapes that you feel you might be
able to get a few extra miles from. Good luck.


Rita

You have to use the cleaning cartridge on an appropriate schedule; not
too often, or too little.

Do DDS drives report drive and cartridge error statistics these days ?
Veritas does this for me with DLT drives and I would never recommend a
tape backup solution that didn't have this feature.

 

[Mike Tomlinson]

Sorry, I forgot to add to my previous post that it would probably be worth
your while to take the cover off the old HP unit and blow with compressed
air then swab the head and capstan with an alcohol dipped cotton swab.

*Very* bad advice, guaranteed to wreck the heads.

 

[Ron Reaugh]

*Very* bad advice, guaranteed to wreck the heads.

Yep, the gisin newbie wipes with 1N HCL.

 

[Rita Ä Berkowitz]

*Very* bad advice, guaranteed to wreck the heads.

It's not that I don't believe you, but what you are saying goes against all
maintenance and repair procedures used by authorized Sony and Seagate repair
facilities.

Please supply a link to support your claim? Thank you.

Rita

 

[Mike Tomlinson]

It's not that I don't believe you, but what you are saying goes against all
maintenance and repair procedures used by authorized Sony and Seagate repair
facilities.

Well, I don't disbelieve _you_ Rita, but since you made the original
claim, if you can cite an authoritative reference from Seagate or Sony
that says it's okay to clean DAT drive heads with alcohol and a cotton
swab, I shall be very surprised.

In addition, to quote from your original post, "swab the head and
capstan with an alcohol dipped cotton swab. The head cleaning tapes are
mediocre at best. Nothing beats a good manual cleaning." - this may be
interpreted by some people to mean "give the head a good scrub", which
is the worst possible thing to do, since "swab" implies using a circular
or figure-of-8 motion with the cleaning implement. Read on for why:

Please supply a link to support your claim? Thank you.

The heads in DAT drives are of a type known as helical scan and are a
miniaturised version of the type found in video recorders. Thus the
same advice regarding cleaning video heads should be applied to DAT
drives. Quoting from the most excellent sci.electronics.repair FAQ at
<http://repairfaq.ece.drexel.edu/sam/vcrfaq.htm#vcrhntcyvh>

"Video head cleaning technique
CAUTION: Read the following in its entirely to avoid an expensive
lesson.
As noted, improper cleaning can destroy your video heads. The head chips
are very fragile and just rubbing them in the wrong direction (NEVER use
an up-and-down motion) can break the chips off requiring replacement of
the entire upper cylinder assembly - one of the most expensive parts in
your VCR!

Manual cleaning using the proper head cleaning sticks is best but
requires that you gain access to the interior of your VCR - i.e., take
off the cover.

If you do not want to do this, you can try a commercial wet cleaning
tape. These is some slight risk, however. The material used in some of
these may have an excessively coarse fiber structure which can catch a
video head and break it off. I have not seen this happen nor could I
recommend a specific brand as there is no way of knowing what their
current product uses. I do not recommend the dry type at all as these
are almost always much more abrasive and may cause premature wear of
your video heads especially if used regularly. When using the wet type
cleaning tapes, follow the directions and - very important - wait
sufficient time for everything to dry out or else you will have a
tangled mess inside your VCR.

Regular video head cleaning should not be needed! Therefore, the regular
use of a cleaning tape is not recommended. As noted, some cleaning tapes
will cause excessive wear to the video heads and no cleaning type can
adequately deal with other parts of the tape path anyhow. If you find
yourself needing to clean your video heads frequently, the video heads
may be worn, the backtension may be set too high, or you may be playing
old or dirty (literally) rental tapes.

To clean by hand, you will need what are called 'head cleaning sticks'.
These are covered by chamois and are safest. DO NOT USE Q-TIPS (COTTON
SWABS). These can catch on the ferrite cores and damage them or leave
fibers stuck in the heads. Q-tips can be used for cleaning the other
parts like the rollers and audio/control head as described above but not
video heads." [UNQUOTE]

From the DAT Heads FAQ at <http://www.minidisc.org/dat-heads-faq.html>

"Manual cleaning without a tape requires opening up the deck.
This is
best left to a qualified technician. dat-heads and the authors of this
document cannot take responsibility for any damage you cause to your
equipment.

If you use foam swabs or plastic handled swabs to clean your deck, be
sure that the solvent you use will not dissolve the foam or the stick.
Use a solvent that is not harmful to the rubber and plastic parts.
Make sure you use a solvent that does not dry out rubber parts. The
most popular solvents are ethyl alcohol, isopropyl alcohol, freon-based
chemicals, and halogenated hydrocarbon-based chemicals such as
Intraclean S711. Alcohol is said to dry out the rubber parts. On the
other hand, rubber parts such as the pinch roller should probably be
replaced periodically anyway. Beyond that, I'm not sure what the
differences between the solvents are but I would love to find out.

To perform manual cleaning, swab all of the parts that the tape makes
contact with with a solvent. This includes stationary guides, rollers,
capstan, pinch roller, drum, and heads. Use foam tipped swabs, chamois
tipped swabs, or a chamois cloth. Do not use cotton swabs.

To clean the heads and drum, you may need to partially disassemble
and/or remove the tape loading mechanism. Hold a swab or a chamois
cloth wrapped around your finger lightly against the side of the drum
and simultaneously turn the drum. This will clean the drum and it will
also clean each head as it passes the swab or chamois. Important: Do
not apply any up/down pressure to the head or it will break off. Be
very gentle." [UNQUOTE]


Note in particular the advice in both excerpts above not to use q-tips
or cotton swabs, and to avoid an up-and-down motion.

Only the proper head cleaning cartridges should be used in DAT drives,
as per the advice from Seagate:
<http://www.seagate.com/support/tape/iguides/dds/datguide.pdf>
page 27, "Head Cleaning".

Other resources found by a google for "manual DAT head cleaning":

<http://www.tape.com/techinfo/maintain.html>
<http://www.tangible-technology.com/helical/swabbing.html>
<http://www.discmakers.com/music/pse/datmain.asp>

 

[John Turco]

It's not that I don't believe you, but what you are saying goes against all
maintenance and repair procedures used by authorized Sony and Seagate repair
facilities.

Please supply a link to support your claim? Thank you.

Rita

Hello, Rita:

I do my own VCR routine-maintenance, and from what I've read, cotton
swabs are >never< recommended for video-head cleaning. They're likely to
get caught on the heads, and possibly break them off the drum.

As for the pince roller, some substances (e.g., CRC "QD Electronic
Cleaner") will, ultimately, destroy the rubber. Alcohol is the best bet
there, indeed.

(Not sure whether these things apply to computer tape drives, but, I
assume they do.)


Cordially,
John Turco <[email protected]>

 

[Rita Ä Berkowitz]

Well, I don't disbelieve _you_ Rita, but since you made the original
claim, if you can cite an authoritative reference from Seagate or Sony
that says it's okay to clean DAT drive heads with alcohol and a cotton
swab, I shall be very surprised.

Thank you for providing the links. I stand corrected on the cotton swabs
and alcohol. My main point was that a manual cleaning is most
likely in order, which Seagate and Sony recommend provided you use the
proper supplies and techniques. Fortunately, there is always someone on
Usenet to give the correct answer or correct someone that gave the incorrect
answer.



Rita

 

[Jonathan Sachs]

I'm writing to follow up on my inquiry, about a month ago, about what to buy
to replace a Hewlett-Packard DDS3 DAT drive that was failing. I'm not sure
the story is over, but it has definitely taken a new turn.

For those who did not see the original messages, I was trying to use the
drive with an Adaptec 2910 to back up a Windows XP machine. The write step
appeared to ran normally, but the verify step ran about five times slower. I
applied some diagnostic programs which told me that the verify step was
running slow because it was getting massive read errors.

I had a long dialogue with an expert in one of HP's user forums about the
problem. With his assistance, I was able to run a very cranky "don't try
this at home" calibration utility which, he assured me, would adjust the
heads of the drive if they had gotten out of alignment. Unfortunately, it
reported that the heads were calibrated correctly, and it had no effect on
the drive's performance.

At this point a new person popped into the thread and told me that my
problem probably was not the drive at all, but the SCSI adapter. He told me,
in essence, that the 2910 was a piece of junk which had actually been
designed for Apple computers, not Wintel, and it worked well enough with
older versions of Windows but Windows XP did not support it, as I would
already know if I had bothered to check Microsoft's hardware compatibility
list. He assured me that if I bought a real SCSI controller, such as a 2930,
my problems would disappear.

I saw numerous inaccuracies in his claims, and concluded that he was either
a troll or one of those idiots who believe that their ego qualifies them to
give advice on any topic. Nevertheless, I wrote a polite reply on the order
of, "I am not aware of a problem with the 2910, and I searched for more
detailed information on the Web but did not find any..." He wrote a rather
snippy reply that said, in essence, "Well, you're aware of it now; the only
question left is, are you going to do anything about it?"

One thing gave his claims some credibility for me. I remembered that my
problems began about the time I migrated to my current Windows XP machine
from an older computer that ran Windows 2000 and drove the tape drive of the
external connector all of an Adaptec 29160N. I thought that the coincidence
in time was only approximate, and so had to be just a coincidence, but I saw
some possibility that it was causal, and the apparent troll was actually on
to something.

Soon after that I was able to pick up a 2940AU cheaply. I tried it, and
found that... absolutely nothing changed.

Then -- don't ask me why -- I tried running the drive off of an Adaptec
USB2Xchange, a gadget which runs a SCSI chain off a USB2 port. The darned
thing not only worked perfectly, but also had about twice the throughput
that it had when I was using it with Windows 2000.

So now my drive is running flawlessly, and twice as fast as before, but I am
baffled twice. First, why is Windows XP having "issues" with the 2940, which
is warranted compatible with it, as well as the 2910, which is not? Second,
why is the drive running twice as fast with the Adaptec converter, which is
basically a kludge that adds a second layer of hardware and a second layer
of drivers to the port, as it did with an adapter that was designed to
operate it?

 

[Eric Gisin]

Soon after that I was able to pick up a 2940AU cheaply. I tried it, and
found that... absolutely nothing changed.

Same chip design as 2910.

Then -- don't ask me why -- I tried running the drive off of an Adaptec
USB2Xchange, a gadget which runs a SCSI chain off a USB2 port. The darned
thing not only worked perfectly, but also had about twice the throughput
that it had when I was using it with Windows 2000.

The 2910 was negotiating 10MB/s, the USB thingy was 5MB/s or less. You
probably don't have active termination or a flakey cable at higher clocks.

So now my drive is running flawlessly, and twice as fast as before, but I am
baffled twice. First, why is Windows XP having "issues" with the 2940, which
is warranted compatible with it, as well as the 2910, which is not? Second,
why is the drive running twice as fast with the Adaptec converter, which is
basically a kludge that adds a second layer of hardware and a second layer
of drivers to the port, as it did with an adapter that was designed to
operate it?

Both Win 2K/XP support the 2910, and likely have the same driver.

 

[Jonathan Sachs]

The 2910 was negotiating 10MB/s, the USB thingy was 5MB/s or less.

I don't think you are thinking clearly about this. You have offered a reason
why transfer through the USB adapter might be slower than through the SCSI
card, which is the opposite of what happened.

Some specifics might make this easier to consider.

1. Windows 2000 with interface through Adaptec 29160N. Write 50 MB/min,
verify 50 MB/min.
2. Windows XP with interface through Adaptec 2910 or 2940AU. Write 50
MB/min, verify 10 MB/min.
3. Windows XT with interface through Adaptec USB2Xchang. Write 100 MB/min,
verify 100 MB/min.

All of the SCSI adapters are rated at 10 MB/sec for their external
connectors. The USB2Xchang is rated at USB2 speed (480 Mb/sec), but probably
really runs at 10MB/sec.

You probably don't have active termination or a flakey cable at higher
clocks.

I assume you mean on the device end, since the adapters are
self-terminating. I did not say so, but it would be logical to assume (and
is true) that the cable and terminator were the same in every case. I can't
make any categorical statements about the quality of the cable, but I used
it for years without problems before the move to Windows XP. The terminator
is from Granite Digital, and so is not suspect.

 

[Bob Davis]

So now my drive is running flawlessly, and twice as fast as before, but I
am
baffled twice. First, why is Windows XP having "issues" with the 2940,
which
is warranted compatible with it, as well as the 2910, which is not?
Second,
why is the drive running twice as fast with the Adaptec converter, which
is
basically a kludge that adds a second layer of hardware and a second layer
of drivers to the port, as it did with an adapter that was designed to
operate it?

I don't know the answer to your question, but I'm running a 2930B and a
2940AU on two different computers using XP Pro with no issues.

 

[Ron Reaugh]

I don't think you are thinking clearly about this. You have offered a reason
why transfer through the USB adapter might be slower than through the SCSI
card, which is the opposite of what happened.

NO, a reliable 5MB/sec. will run the tape drive at full speed. A
unreliable and improperly terminated 10MB/sec. could easily be vastly
slower!

I assume you mean on the device end, since the adapters are
self-terminating.

Not necessarily. You may need to set something.

"You probably don't have active termination or a flakey cable"

I did not say so, but it would be logical to assume (and
is true) that the cable and terminator were the same in every case. I can't
make any categorical statements about the quality of the cable, but I used
it for years without problems before the move to Windows XP. The terminator
is from Granite Digital, and so is not suspect.

HUH!

You've been given all the information necessary to fix your problem. Now
quit gibbering and go fix the cabling and/or termination.

 

[chrisv]

HUH!

You've been given all the information necessary to fix your problem. Now
quit gibbering and go fix the cabling and/or termination.

Quit gibbering and fix your quote-mangling newsreader, you idiot.