Database back-end corruption

[Guest]

I have a multi-user database which is split into a front and back end ... Front time to time the back-end becomes corrupted (unrecognized format ... would you like access to attempt to repair...). Is there a method to determine the originating cause the the corruption? Basically, how can I figure why this happens so that I might try to avoid this from occuring in the futur?!

If it make any difference, everytime the back-end corrupts, I always make access try and recover the database (obviously) and in all cases so far managed to do so. Which brings up a second question. When access recovers a corrupted database, does this imply no loss of data or does it recover what it can. Am I garanteed that the recovered file has no loss of data?

Thank you for the help,

Daniel

 

[Allen Browne]

There is no guarantee that the recovered file contains everything. The most
common form of runtime corruption is probably a bad index, and Access can
usually rebuilt the index with no loss. Occassionally the index itself may
get deleted, which affects other things (e.g. performance, referential
integrity) even though no data was lost. Other times Access may dismiss an
invalid pointer to a memo or OLE field, in which case you did lose a comment
or graphic. Very rarely, you can end up with a truncated table, so you have
lost significant data.

For suggestions on how to prevent corruption, see:
Preventing Corruption
at:
http://allenbrowne.com/ser-25.html
Take particular note of the comments about service packs for Office and JET,
and disabling Name AutoCorrect.

If you have taken all obvious precautions and the database is still
corrupting, you will want to track down which machine is causing the
interrupted write that triggers this persistent corruption. Log the users in
(with your initialization code/autoexec macro), and out (Close event of a
hidden form). The machine that is logging in but not out is the one that is
crashing or has the intermittent network connection. These links have the
API call to get the name of the machine/user for your logging routine:
http://www.mvps.org/access/api/api0009.htm
http://www.mvps.org/access/api/api0008.htm

--
Allen Browne - Microsoft MVP. Perth, Western Australia.

Reply to group, rather than allenbrowne at mvps dot org.

I have a multi-user database which is split into a front and back end ...

Front time to time the back-end becomes corrupted (unrecognized format ...
would you like access to attempt to repair...). Is there a method to
determine the originating cause the the corruption? Basically, how can I
figure why this happens so that I might try to avoid this from occuring in
the futur?!

If it make any difference, everytime the back-end corrupts, I always make

access try and recover the database (obviously) and in all cases so far
managed to do so. Which brings up a second question. When access recovers
a corrupted database, does this imply no loss of data or does it recover
what it can. Am I garanteed that the recovered file has no loss of data?