|
Thanks to Razzak and all for replies. We have
a tool that does a complete OUTPUT/INPUT each night, so the database is
regularly maintained. As I think more about it, another common factor is
that these tables typically take a heavy load, with INSERTS and DELETES of 3 -
20 rows each from 10+ users at any given moment, usually done
through a DECLARE CURSOR. Maybe we're overloading the database
with transactions.
Somewhere and somehow, some peice of the
transaction is not completing - maybe the index is getting updated and the data
is not getting written or vice versa. Can't tell. Have to try
different settings, I guess. Unfortunately, reloading and reindexing are
not feasible solutions during the workday. We can't control the crashing,
it's become a fact of life for us and our users just deal with it, so if it is
caused by a crash, there's little we can do. Our crashes typically come
when forms close - can't figure out why, though it is a lot less common on
our W2K machines than our NT boxes.
If we learn anything else, I'll be sure to
share.
wm
|
- Corrupt Indexes William Mason
- Re: Corrupt Indexes Tom Grimshaw
- Re: Corrupt Indexes David M. Blocker
- Re: Corrupt Indexes mjs
- Re: Corrupt Indexes - Razzak's Reply A. Razzak Memon
- RE: Corrupt Indexes William Mason
- RE: Corrupt Indexes Troy Sosamon
- Re: Corrupt Indexes Brent Skean
- Re: Corrupt Indexes K G
- Corrupt Indexes William Mason
- Re: Corrupt Indexes Ben Petersen
- RE: Corrupt Indexes Albert Berry
- RE: Corrupt Indexes Tom Grimshaw
- Corrupt Indexes J.M. GRATIAS
- Re: Corrupt Indexes David M. Blocker
- Re: Corrupt Indexes Bernard Lis
