Subject | Re: [IB-Architect] Next ODS change (was: System table change) |
---|---|
Author | Ivan Prenosil |
Post date | 2000-11-15T11:34:35Z |
Dalton,
space will substantially increase insert rate for blobs,
it will at the same time decrease insert speed into ordinary
tables (it has something to do with algorithm for looking up
free space in gdb; you can easily see the difference
if you look at statistics for executed commands).
But yes, for large blobs it is definitely better to preallocate space.
Besides known method (inserting many rows and deleting them)
I also tried to pre-allocate space at OS-API level,
i.e. by extending gdb file size, and it worked !
(at least for single file db; I did not try mutifile db)
Ivan
prenosil@...
> He created a databaseWhen I did some tests with IB5, I found out that while pre-allocating
> with a 1KB pagesize on NT4 to table TIFF files only. The insertion rate was
> 32 per minute. By moving from 1KB to 8KB insert rate went up to 300per
> minute.
>
> I wonder what a 32KB pagesize would do.
>
> And as another interesting aside, by pre-allocating the space within the GDB
> it went from 300 to 600 per minute.
space will substantially increase insert rate for blobs,
it will at the same time decrease insert speed into ordinary
tables (it has something to do with algorithm for looking up
free space in gdb; you can easily see the difference
if you look at statistics for executed commands).
But yes, for large blobs it is definitely better to preallocate space.
Besides known method (inserting many rows and deleting them)
I also tried to pre-allocate space at OS-API level,
i.e. by extending gdb file size, and it worked !
(at least for single file db; I did not try mutifile db)
Ivan
prenosil@...