Re: AP Native - Not enough disk

I've found that recovery from disk problems like this requires just a
single logged in session and careful deletion of unused data. For
example:
- Clear-file file-stats
If the stats file is small and it consumes a lot of overflow, then
you should be able to retrieve some non-contiguous space here.
- delete-file zip, words, or anything else that is tiny and almost
never used. This gets you just enough contiguous space to do other
small operations.
- delete-account ref, tutor, games, or anything that is DXed and looks
unused. Now you should have several blocks of contiguous frames and
that should be good enough to get a file-save.

I really don't like the documented procedure for hacking frames to
free up disk. It's a last-ditch procedure which deletes random data
that may not be recoverable. The first goal should be to free blocks
of contiguous frames to get a file-save.

When you do get a save and you are not familiar with the end-user
system, be very careful not to just blindly do a full restore using
the save.
- Some sites DX history files to shorten their file-saves. There may
be a written procedure or some program that un-dx's files in
preparation for a full save/restore. If you restore without this you
could save the system for a day but lose critical data forever.
- For the same reasons, check for DX'd accounts. Be careful about
deleting accounts that are DX'd just because the end-user doesn't know
what they are - the application might use these for archiving
long-term data as described above, so that the primary accounts save
quickly and on a single 1/4" tape.

If the site already has AccuTerm loaded and you find a need to extract
the data by some method other than tape, let me know. The code I
wrote for the recent crippled R83 system should work for any
environment and I'll probably open-source it soon.

HTH
T

SteveS wrote:

Hello everyone,

I have a customer that is running AP/Native (if you can imagine that).
And they have received the totally aweful "Not enough disk" message.
And they have hit R for reserve too many times now so the system is
inoperable (they never called!).

I have the AP/Pro procedure to trim the overflow (Tech note 9), but I
do not seem to have any procedure for that in AP/Native.

Does anyone have any suggestion to recover the system and ultimately
recover the data?

Thanks,

Steve

.

Relevant Pages

  • Re: FAT32 or NTFS?
    ... >>10% wastage on 10 volumes is one whole hard disk wasted. ... >uncertainty about free space. ... Corrupt data can be better than none, ... "you can only recover part of the database". ...
    (microsoft.public.windowsxp.general)
  • RE: Dead hard drive
    ... Maybe you could recover ... That's a single FAT 32 active partition w/ XP Pro on it and it can access ... Here is a quick way to find if anything is left on your disk: ... to format the disk. ...
    (microsoft.public.windowsxp.general)
  • Re: Systems Registry data problem
    ... I tried to use the memtest program but had great difficulty in copying to disk using Nero, and when it ran, it only seemed to carry out one sweep of the RAM and then seemed to go to sleep. ... There is still a problem with the Registry because I cannot run Registry Mechanic, which I have used in the past, or Regcure. ... Both of these programs stall part way through, the screen goes black and does not recover or give an error message. ... Choose the most recent restore point and SR should restore the user ...
    (microsoft.public.windowsxp.general)
  • Re: Systems Registry data problem
    ... using a backup disk that I made a few weeks ago. ... There is still a problem with the Registry because I cannot run Registry ... Choose the most recent restore point and SR should restore the user ... How to Recover from a Corrupted Registry ...
    (microsoft.public.windowsxp.general)
  • Re: Dog bit audio CD. Any Hope of recovery?
    ... When I look at the disk there is one small ... dent, the size of a pin head. ... Linux) will make extreme efforts to recover the data. ... If there's enough of a dent in the bottom of the disc to cause ...
    (rec.audio.tech)