Do you mean the database gets slow at 2500 records?
So far (783 records this morning) I haven't came across any
degradation in speed and performance. But, yes, I'm afraid that it can
happen - let's say 2000 records in tInspection (with memo field),
30-50 in tIssue (with memo field) and constantly 10-30 records in a
'new table' we're talking about.
As far as MEMO or VARBINARY for that matter, it can be an issue. HSQL
stores the memo data right in the row ( not so for all DBMS's ), this
means that as the memo fields become more populated with data fewer
records will fit into the memory buffer during a search. ( this includes
internal searches in the engine for joins and the like also ) One
solution to this is to move the memo field to a seperate table with a 0
or 1 relationship to tInspection.
However, the number of records you are talking about is fairly small. If
the data in the memo fields is also fairly small then I would not expect
you will have a problem.
For my money, I would leave the schema the way it is now and only if I
started to see a problem would I go about breaking my tables apart. Memo
fields, for user entry via keyboard, tend to not be that large, if you
where talking about images it would be totally different. Then I would
segregate them to a separate table right from the start.
The one place that you will notice a slow down as the database grows is,
time to establish the initial connection. Breaking the tables apart will
not help here unfortunately, heck it may even make things worse.
Drew
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
