HP Ensures LTO Has An Effective Cleaning Strategy
HP Ensures LTO Has An Effective Cleaning Strategy
HP Ensures LTO Has An Effective Cleaning Strategy
Cleaning
Strategy
white paper
Abstract
This white paper explains how HP develops a highly effective cleaning strategy for
the new generation of high-performance HP Ultrium tape products. It describes
the preventative and corrective cleaning regimes that enable the drive to deliver
its outstanding reliability.
In the Travan technology some level of staining is expected in the recessed pole
tip region. Pronounced stains tend to form at the MR read transducer, possibly
Staining formation on DDS read due to thermal effects. The amount of staining depends on the abrasivity of the
head. recording media. With fresh tape, the abrasivity is high enough to prevent stains
from forming. As the tape wears, the abrasivity is reduced and staining begins to
occur. If the stains become severe enough to substantially increase head/tape
interface, the recording performance will degrade, leading to an increase in
error rate and possible failure of drive to write and read.
Recording
Element Smearing was observed to a lesser degree in the early prototypes of LTO drive. It
Direction of was found that lowering the electrical potential difference between the MR
Internal element and its shield significantly reduces the probability for smearing to occur.
Head Head Clean
Contour
The HP Ultrium head and media are designed to minimize the smearing.
In the HP Ultrium head design, the wrap angle is precisely controlled by head
geometry--creating consistent head-tape spacing. The slotted head contour is
designed for trapping debris. The slots are specifically designed to get the debris
out of the Head Tape Interface (HTI). The debris in the rails is removed via the
brushing action of the internal head cleaner. This design was a conscious choice
to control debris at the head/tape interface, since slotted heads have been
proven successfully for many years in much dirtier customer environments. The
Slotted Head Contour of a HP head structure is made of an extremely hard and durable ceramic material
Ultrium tape drive called Cermet (composite of Al2O3/TiC, same material as in disk drive heads).
Long term reliability testing has proven that the HP Ultrium drive, under typical user
environment, is capable of running without preventative cleaning and it requires
very minimal corrective cleaning.
debris would be permanently trapped inside the slots. Even the use of cleaning
Load (lbs)
0.6
0.5
Load
0.4
0.3
0.2
0.1
cartridge is ineffective for cleaning this type of head contamination.
0
-0.1 0 0.01 0.02 0.03 0.04 0.05 Debris Head Slots
Displacement (inch)
Before After
Brush pattern and stiffness
designed to give effective
cleaning.
* See Preventing Excessive The above pictures demonstrate the effectiveness of the Internal Head Cleaner. Nearly
Cleaning Operation all of the media debris trapped inside the slots on the head was successfully removed
after 8 seconds of cleaning using the internal head cleaner. This is one of the reasons
that the HP Ultrium drive requires very infrequent use of the cleaning cartridge.
Abrasivity plays a critical part in dislodging the sticky debris from the head. When
the debris is loose, it can be easily transferred onto the cleaning tape, which is
Media debris on Tape Bearing wound into the cleaning cartridge and removed from the drive at the end of the
Surface. cleaning operation. This is one of the reasons for using a fresh piece of cleaning
media for every cleaning operation. Another reason is the abrasivity of the
Head Slot
cleaning tape is reduced after each pass over the head, thus producing less
effective cleaning.
Different technology results in very different types of media debris formation and
cleaning requirement. In DDS technology the cleaning tape is about 200 times
more abrasive than normal recording tape. In LTO technology high abrasivity
could result in undesired effect such as Pole Tip Recession (PTR). For the HP Ultrium
tape drive this high contrast abrasivity is unnecessary. There is less debris being
generated, more being trapped by the slotted head contour, and the internal
head cleaner frequently brushes loose debris away from the head/tape
interface. Also the type of debris generated inside LTO drive appears to be less
Media debris on the recording sticky in nature (compared to DDS). HP has developed a unique cleaning tape
element. that balances PTR with effective debris removal.
Minimum BER =
-3
-6.73 a week, 390 cleans is equivalent to 7.5 years.
-3.5--3
-3.5
-4--3.5
-4
-7--6.5
12
-7
16
-8--7.5
6
28
4
2
0
Boost
Drive E130000027 Wrap 5 @ 4.1m/s MRT4.5 After 150 Low Frequency Signal Change* N.A. -24.880 -26.152 -14.814
Cleaning Cycles
Minimum BER =
-5.96
PW50 Change* N.A. 31.199 36.823 20.658
-3
-3.5--3
Beginning Average BER Log10 (BER) * - 5.50 < -7.256 -7.332 -7.201
-3.5
-4--3.5
-4
-4.5--4
-4.5
-5--4.5
-5
Ending Average BER Log10 (BER) * -3.25 < -6.815 -6.318 -6.640
-5.5--5
-5.5 BER
-6--5.5
-6
-6.5--6
-6.5
8
-7--6.5
12
-7
Change in Log10 (BER) 1.6 -0.15 ~ 1.26 -0.05 ~ 2.08 0.09 ~ 1.18
16
-8 -8--7.5
24
6
28
4
2
0
Boost
Head tuning parameters AFTER * Parameters are required for proper drive function, and are not in cleaning cartridge qualification
150 cleaning cycles shows based on the First Generation LTO Ultrium Cleaning Cartridge Specification, Revision B.
minimal error rate change and IMPORTANT: Do not use swabs or other means of cleaning the heads. The
stable tuning range.
cleaning cartridge uses a special tape to clean the tape heads. The user must
use only approved Ultrium cleaning cartridges to clean the tape heads. Non-
approved cleaning cartridge will be rejected by the drive.
03:36.0
02:52.8 02:55.0
02:46.0
02:37.0
02:28.0
02:20.0
02:09.6 02:10.0
02:02.0
01:53.0
Time Interval Per Cleaning Cycle
01:44.0
01:34.0
01:26.4 01:26.0
01:16.0
01:08.0
00:58.0
00:51.0
00:43.2
00:00.0
1 3 5 7 9 11 13 15
Cleaning Cycle
•= During the cleaning cycle the amber ‘Use Cleaning Cartridge’ LED is
turned on and the green ‘Ready’ LED will flash.
Note: The drive's Tape Alert feature will send a message to your backup
application when the tape heads need cleaning or when an expired cleaning
cartridge has been detected.
The amber Use Cleaning (i) The cartridge is unloaded (if applicable)*
Cartridge LED is extinguished (ii) The amber Tape Error LED is lit indicating that the Tape is
when cleaning has been defective. A check condition occurs with a Sense Key of 3
successfully completed. (Medium Error) and an ASC/ASQ of 3007 (Cleaning failure).
(iii) The Invalid Tape Alert Flag (23) is set.
(iv) If the Cleaning LED and Tape Alert flag 20 were previously
set prior to the cleaning cartridge being inserted, then they
will remain set.
(v) Cartridge CM remains unmodified.
If the firmware detects the above condition is true after a cleaning cartridge has
been inserted, then
Subsequent consecutive
cleaning does not alter the a) The internal head cleaner is activated to brush the heads.
LED sequence. It activates b) Estimated Cleaning Usage in the cartridge CM remains the same. I.e.
the Internal Head Cleaner. not decremented
c) Cleaning tape is not threaded.
d) Cartridge is unloaded (if applicable)* after internal head cleaning has
completed its operation.
•= The amber Tape Error LED is lit indicating that the Tape is defective.
A check condition occurs with a Sense Key of 3 (Medium Error)
and an ASC/ASQ of 3007 (Cleaning failure).
The amber Tape Error LED is
turned on after an expired
cleaning cartridge has been •= The Expired Cleaning Media Tape Alert Flag (22) is set and if the
detected and ejected by Use Cleaning Cartridge LED and Tape Alert flag 20 were set prior
the drive. to the cleaning cartridge being inserted, then they will remain set.
Key:
> - Use
H - Reject
Note:
“Use” means the drive loads the cleaning cartridge, executes the
cleaning algorithm and ejects the cartridge at the end.
“Reject” means the drive loads the cartridge and immediately ejects it.
The “Tape Error” LED is lit and the Tape Alert flag is set. No cleaning takes
place and the drive does not modify the cartridge memory nor initialises it.
Is the internal head cleaner effective in keeping debris away from the heads?
HP has designed the internal head cleaner with an optimal brush pattern, stiffness
and material properties, so that it would produce an effective and efficient
cleaning operation. Test data has shown that the brushing action consistently
removes media debris from the slots of the head and prevents media debris from
building up on the head/media interface.
Why does HP drive reject my cleaning cartridge but another vendor drive
accepts it?
The HP Ultrium drive will reject a cleaning cartridge if the cartridge is not a
universal cleaning cartridge and its LTO-CM does not contain HP unique ID or the
cleaning cartridge is expired.
Why does HP cleaning cartridge expire after 15 uses when another vendor‘s
cleaning cartridge lasts many times more than 15 uses?
The HP cleaning algorithm is designed to produce an effective cleaning result
against head contamination that is not removable by means of the internal head
cleaner whist maintaining optimal drive performance. This requires the use of a
predetermined length of fresh media that only allows 15 uses. Because the HP
Ultrium drive is designed for no preventative cleaning using cleaning cartridge,
the use of cleaning cartridge is very infrequent and is only reserved for extreme
cases. Another vendor’s cleaning algorithm may use less media and may
produce less effective cleaning result that leads to the drive requiring more
frequent cleaning. Less frequent usage of cleaning cartridge is better for the drive
and ultimately benefits the user.
Will I get the same number of uses from a universal cleaning cartridge across
different LTO drives?
It depends. HP cleaning algorithm ensures the same total number of use if the
universal cleaning cartridge is used in HP LTO drive only. Each LTO drive
manufacturer uses a unique cleaning algorithm the total number of cleans per
tape will not be the same. Using the universal cleaning cartridge in a mixed LTO
drive manufacturers user environment (e.g. Library usage) may give a variable
total number of use.