frasc
asked on
Repair 8mm tape.
Well, I have an Exabyte 8505. The drive recently "ate" one of my 8mm tapes.
The tape was stuck in the drive. So, I opened it, and cut the tape in order to remove the cartridge (that is the Exabyte recommended procedure).
After repairing physically the tape (but still one chunk is missing: the very first part of it, that had been too much chewed), the drive is obviously missing some information and send a check condition, the condition being "Blank Medium".
Do you know any mean to recover the end of the data on the tape, by "skipping" the first part ?
BTW this Exabyte drive has been OEM'd by IBM. That means the firmware is from IBM. The SCSI INQUIRY command returns
VendorID EXABYTE
ProductID IBM-8505
I'm running linux 2.0.29 with an NCR53C825 based SCSI adapter
More Info:
- I tried `mt' command to try to skip over.
- I tried to send the SCSI SPACE command from a C program
and tried to set Exabyte Vendor Unique bits
- It seems the problem is really due to the fact that the
first part of the tape (the one supposed to carry a `BOM'
mark) is missing.
The tape was stuck in the drive. So, I opened it, and cut the tape in order to remove the cartridge (that is the Exabyte recommended procedure).
After repairing physically the tape (but still one chunk is missing: the very first part of it, that had been too much chewed), the drive is obviously missing some information and send a check condition, the condition being "Blank Medium".
Do you know any mean to recover the end of the data on the tape, by "skipping" the first part ?
BTW this Exabyte drive has been OEM'd by IBM. That means the firmware is from IBM. The SCSI INQUIRY command returns
VendorID EXABYTE
ProductID IBM-8505
I'm running linux 2.0.29 with an NCR53C825 based SCSI adapter
More Info:
- I tried `mt' command to try to skip over.
- I tried to send the SCSI SPACE command from a C program
and tried to set Exabyte Vendor Unique bits
- It seems the problem is really due to the fact that the
first part of the tape (the one supposed to carry a `BOM'
mark) is missing.
Ok...
You have a serious problem.
Here is the most simple solution.
... Get a useless tape, which you don't want (has to be the same type). Cut the first couple cm. of that old tape out and repair it into the new tape. That should do the trick by fooling the computer. I would suggest formating the old tape with the same system first, so that the system would not find a different medium as the content of the tape.
Good luck
You have a serious problem.
Here is the most simple solution.
... Get a useless tape, which you don't want (has to be the same type). Cut the first couple cm. of that old tape out and repair it into the new tape. That should do the trick by fooling the computer. I would suggest formating the old tape with the same system first, so that the system would not find a different medium as the content of the tape.
Good luck
ASKER
Well, I'm just wondering how the drive could not return an I/O error when encountering the precise area where the has been spliced. This area will have cut tracks (it's Video tape!)
that won't fit with the tracks on the other side of the tape.
that won't fit with the tracks on the other side of the tape.
ASKER
Okay. So I gave anyway a try to 115lee's answer.
At the precise place the tape has been spliced, the drive returned not an I/O error as I thought, but an Invalid EOM
error.
That makes sense: 8mm tape drives use helical scan (Video technology) therefore there are very few chances for the tracks
to match.
At the precise place the tape has been spliced, the drive returned not an I/O error as I thought, but an Invalid EOM
error.
That makes sense: 8mm tape drives use helical scan (Video technology) therefore there are very few chances for the tracks
to match.
In a word, you are up a creek without a paddle.
At the level of a SCSI device, you cannot get at the information
at all. The drive will not advance beyond logical end of tape.
HOWEVER, there is a serial interface inside the drive that
permits very low level access to the hardware. (You are basically
bypassing the scsi board...) Exabyte will sell, or perhaps give
you this cable and some software to access it.
There are companies that specialize in data recovery. If
it's *really* important they will get the data off for a price.
Note however that the format of the data must be something
that has some form of resync capability. Gnutar does. Dump
doesn't. Not sure how the drive will handle compression
starting from the middle of a tape.
At the level of a SCSI device, you cannot get at the information
at all. The drive will not advance beyond logical end of tape.
HOWEVER, there is a serial interface inside the drive that
permits very low level access to the hardware. (You are basically
bypassing the scsi board...) Exabyte will sell, or perhaps give
you this cable and some software to access it.
There are companies that specialize in data recovery. If
it's *really* important they will get the data off for a price.
Note however that the format of the data must be something
that has some form of resync capability. Gnutar does. Dump
doesn't. Not sure how the drive will handle compression
starting from the middle of a tape.
ASKER
Sherwood, In your proposed answer you say:
> At the level of a SCSI device, you cannot get at the
> information at all. The drive will not advance beyond
> logical end of tape.
I'm not sure of that. I mean, according to Exabyte, the "directory support enabled" firmware is able to skip over
an End Of medium mark. What I don't know (nor Exabyte support)
is whether this "directory support" will skip over the EOM
encountered at the place where the tape has been spliced.
I will try this weekend the serial interface solution. I don't
think I'll be able to download data thru this link, but maybe
I'll get some hints when playing with it.
BTW I got many things from www.exabyte.com. Their 8mm tape drives
are not reliable, but at least they let you have some tools.
> At the level of a SCSI device, you cannot get at the
> information at all. The drive will not advance beyond
> logical end of tape.
I'm not sure of that. I mean, according to Exabyte, the "directory support enabled" firmware is able to skip over
an End Of medium mark. What I don't know (nor Exabyte support)
is whether this "directory support" will skip over the EOM
encountered at the place where the tape has been spliced.
I will try this weekend the serial interface solution. I don't
think I'll be able to download data thru this link, but maybe
I'll get some hints when playing with it.
BTW I got many things from www.exabyte.com. Their 8mm tape drives
are not reliable, but at least they let you have some tools.
ASKER
Adjusted points to 230
I don't think that the directory support option will help.
In a nutshell the tape works like this:
P1ppppe--------2dddddddddd
Wehre P is the partition data information about the media
1 is the start of the first partition, and 2 is the start
of the second partition. The drive is very careful to
leave lots of space between the logical end of media (e) on
the first parition and the start on the second.
If you have software that supports multiple partitions, you
*may* be able to give the scsi command to advance to the second
partition. I have tried this with other drives that nominally
support partitions (HP DAT) but was unable to get it to work.
Note: If the 8505 has the same serial interface as the 8200
then it's a tiny 4 pin connector, near the middle of the back
end of the drive.
Note: If you try the scsi approach, ask exabyte for the
programmers reference manual for the drive.
If you need further input on this, then grade this an 'F'
otherwise, no one else can answer it.
In a nutshell the tape works like this:
P1ppppe--------2dddddddddd
Wehre P is the partition data information about the media
1 is the start of the first partition, and 2 is the start
of the second partition. The drive is very careful to
leave lots of space between the logical end of media (e) on
the first parition and the start on the second.
If you have software that supports multiple partitions, you
*may* be able to give the scsi command to advance to the second
partition. I have tried this with other drives that nominally
support partitions (HP DAT) but was unable to get it to work.
Note: If the 8505 has the same serial interface as the 8200
then it's a tiny 4 pin connector, near the middle of the back
end of the drive.
Note: If you try the scsi approach, ask exabyte for the
programmers reference manual for the drive.
If you need further input on this, then grade this an 'F'
otherwise, no one else can answer it.
ASKER
I think you're right. Very few people know how to fix a
broken tape. Those people are actually hired by data recovery
companies. Period.
Maybe one of 'em is hanging around here ...
broken tape. Those people are actually hired by data recovery
companies. Period.
Maybe one of 'em is hanging around here ...
ASKER
Adjusted points to 250
ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
ASKER
Yeah. I guess I'd better give up. :-(
As soon as I get my data back I'll think to have a better
hardware for backup system.
As soon as I get my data back I'll think to have a better
hardware for backup system.
One suggestion would be Magneto-Optical storage. They're like CDs, only rewritable many times over, have speeds close to those of hard drives and are not super-sensitive to magnetic fields.
Iomega's Jaz and SyQuest's SyJet spring to mind...they're both high-volume ( >1 GB) storage solutions; Fujitsu makes a M-O drive that accepts standard 230 MB M-O disks which I have been satisfied with, as well.
Good luck, and keep those pinch-rollers clean and lubricated next time.
Iomega's Jaz and SyQuest's SyJet spring to mind...they're both high-volume ( >1 GB) storage solutions; Fujitsu makes a M-O drive that accepts standard 230 MB M-O disks which I have been satisfied with, as well.
Good luck, and keep those pinch-rollers clean and lubricated next time.
ASKER