T.R | Title | User | Personal Name | Date | Lines |
---|
4245.1 | "Is this the answer?" | SUOSW3::MOEGENBURG | | Wed Nov 07 1990 06:35 | 10 |
| DGM-Software Germany
Re. 4245 < Disk not Validated >
If I get this message from my system, I use the Workbench Diskdoctor
or I kill this Disk.
- DGM Software Germany
|
4245.2 | hard to say | SALEM::LEIMBERGER | | Wed Nov 07 1990 07:34 | 5 |
| I used to play a game that would often do this to a hd partition.
I used to issue an info command,and then ask for a directory.
At that point it would reboot without the error. Never gave me that
warm fuzzy feeling but it worked.
bill
|
4245.3 | Mission Accomplished | DICKNS::MACDONALD | VAXELN - Realtime Software Pubs | Wed Nov 07 1990 09:12 | 5 |
| I ended up reformatting the drive. I then restored it with Quarterback.
Of course Quarterback doesn't create empty directories, so it took a
bit of handy work to get all of my assigns back together. Plus I lost
all my unread newsgroups (never back'em up in order to save floppies).
|
4245.4 | Validation not big problem | MKODEV::OSBORNE | Blade Walker | Tue Nov 13 1990 12:44 | 16 |
| re:.3
> I ended up reformatting the drive.
I may be wrong, since this hasn't happened to me in a while, but "disk not
validated" is not cause for lots of concern. It means that something prevented
the Amiga from "validating" the last file write. (Power failure or pulled
floppy from drive, e.g.)
All the Amiga wants to do is validate the disk, and then it will act like
nothing happened. The only problem is that validating a hard disk means
spending some time reading the whole disk to develop what I believe are just
some checksums.
If a file was being written when the problem occurred, it's gone. But that
should be the most that is wrong.
|
4245.5 | Requester means failure | TLE::RMEYERS | Randy Meyers | Tue Nov 13 1990 13:13 | 13 |
| Re: .4
I think that the problem is that if you get the requester, it means
that AmigaDOS failed when it attempted to validate the drive.
Normally, AmigaDOS will quietly attempt to validate any unvalidated
disk upon first access. There is no requester: the only sign that
a disk validation is taking place is that all access to the disk
is very slow, the disk is doing *lots* of I/O, and if you do an
INFO command, the disk will be marked as "Validating". You only
get the requester if the process fails to create a valid bitmap
of blocks used on the disk (usually, a "Key already set" error,
which means one block appears in two separate files).
|
4245.6 | | NBOSWS::FRIES | | Wed Nov 14 1990 05:44 | 9 |
| Is there a way to know which two files are affected by the 'key already set'
error message??
Perhaps it would be possible to delete these two files so it's not necessary
to reformat the disk.
Or isn't a write access possible at this time?
Gerald (I'm sorry about the sick first version of my reply)
|
4245.7 | Fixing the disk using a disk editor | TLE::RMEYERS | Randy Meyers | Wed Nov 14 1990 16:26 | 37 |
| Re: .6
>Is there a way to know which two files are affected by the 'key already set'
>error message??
The disk validator itself does not track the information. However, you
could write a program to do this yourself. Your would have an
array with elements for every block in the partition. The array elements
would be pointers to the name of the file that owns the block. Scan the
disk, and note which blocks belong to which files. If you find a block
owned by a file, check to see if its array element is NULL, if so, this
is the only file so far that owns the block, and so store the filename
pointer in the element. If the element isn't NULL, then you have previously
found another file that also owns the block. Print out the name of the
previously stored file and the name of the current file whose blocks you
are collecting.
AmigaDOS does pretty much the same as the above, but elements in its
array are single bits that store TRUE or FALSE (whether the block has
been found allocated to a file or not).
>Perhaps it would be possible to delete these two files so it's not necessary
>to reformat the disk.
That's pretty much the type of thing people do with disk editors.
>Or isn't a write access possible at this time?
With the FastFileSystem at least, the partition is write-locked if the
validation fails. However, if you access the disk below the level of
the file system, you can still write to it. (In other words, the
delete command will do you no good, but a disk editor would.)
My advice: unless you are a disk wizard, cut your losses by backing up
the disk and reformatting it. That's what I would do.
|
4245.8 | DiskSalv | CSCOAC::KENDRIX_J | | Wed Nov 14 1990 16:48 | 10 |
| re:.7
Actually there's already a program out which dones a fine job of
analysing disks and possibly restoring bad ones. DiskSalv by Dave
Haynie, it should be on tape. I don't know if I would try and restore
a trashed HD with it, actually, I know I wouldn't, but if you just want
to know the extent of the damage, then that will do it just fine.
JK
|