| T.R | Title | User | Personal
Name | Date | Lines |
|---|
| 1794.1 | | CX3PST::MOUTNS::G_FROEHLIN | BURGUNDY? Let's open the bottles | Wed Aug 25 1993 04:26 | 7 |
| 1794.2 | can't even skip the bad record!??! | COMEUP::SIMMONDS | lock (M); while (not *SOMETHING) { Wait(C,M); } unlock(M) | Mon Feb 03 1997 04:19 | 7 |
| .7> COULD SOMEONE FROM TAPE ENG. GIVE US A GOOD EXPLANATION WHY TAPE DRIVES
.7> LIKE THIS DO NOT ALLOW TO READ PAST A PARITY ERROR? THANKS!
I want and need to know too.. anyone care to enlighten us?
Thanks,
John.
|
| 1794.3 | | TAPE::PETERS | | Mon Feb 03 1997 20:15 | 20 |
| re .1 .2
Why do errors stop tape drives ?
First you need to understand how bad the tape must be to get a error.
The DLT drives ( including TK30,TK50,TK70,TK85,TK86,TK87,TK88,TK89 ) use
CRC and xor blocks to detect and correct errors. The drive will correct
up to 4KBytes of bad data in one spot. The real problem starts when the
phase lock loops loose the signal. These drives use the phase lock loop
signal to follow the track. When the signal is lost the drive can't find
the track any more. The drive reports a position lost and stops. When
Digital designed tape drives we put extra code in the drive to go beyond the
error. Now that we buy the drives from third party vendors they are not
willing to add this feature for us. This is the price for going to open
systems.
Steve P.
|
| 1794.4 | Parity Error on TLZ6l, TLZ7L's Question | KYOSS1::LUIZZA | | Mon Feb 03 1997 23:43 | 6 |
|
Steve,
Does this also apply for TLZ6l's and TLZ7l drive type or do they handle
parity conditions worse?
/Irv
|
| 1794.5 | | TAPE::PETERS | | Tue Feb 04 1997 01:57 | 12 |
|
The DAT tapes ( TLZ04, TLZ06, TLZ07, TLZ09 ) fail in
much the same way as DLT. The DAT drives use a reed-solemon
type ECC. For a DAT to loose data, multiple stripes must fail.
In this case the head/drum loose tracking. We made HP put
extra code in the TLZ04 to recover. We also pushed on the
TLZ06/07 vendor to add code and got a few changes but nothing
was done for the TLZ09. Because SCSI doesn't require them to
do it and Digital is the only company asking for it, don't
expect any future drives to read past very bad spots.
STEVE P.
|
| 1794.6 | | LEFTY::CWILLIAMS | CD or not CD, that's the question | Fri Feb 07 1997 18:53 | 10 |
|
The TLZ04,6, and 7 will probably read thru most bad spots.... I beat on
the vendors to get that code there.
This is the first I've heard that the TLZ09 has a problem there. It's
still in the spec that it be able to do that.
Chris W.
|