postgres occupies more space than mysql???

[Paul Ramsey]

Are these points, by any chance?
You could try using the 1.0 postgis series, but posts on this list seem
to indicate we have a problem there too. The 0.X series has fairly
fluffy representations, so if you have a large number of points, the
database size of a postgis spatial database versus a x/y attribute
database will be quite substantial.  You will find that regardless a
spatial database will have some extra overhead over an x/y attribute
database.  For example, were your x/y stored as float? Well in the
spatial database they will be doubles, so double your storage size. And
add a byte for SRID storage. And potentially add 4 bytes for bounding
box storage. It adds up, and the overheads can be even higher in the
0.X series.
Points are also amenable to simple multi-key btree indexes, so you
might find that keeping your data in mysql or any attribute database
and using the OGR "virtual table" idea is just as effective.
Paul

On 4-Feb-05, at 9:08 AM, Shashi Gireddy wrote:

> Hi all,
>
> I have my SF1 data on mysql database, i wnated to migrate to postgres
> to get advantage of Postgis, I created a backup .sql file (whole data)
> whose size is 8GB but its actual database size in mysql is 700MB (it
> may be because insert into(bla,bla) values(bla, bla). but after i
> migrated to postgres it occupied 7GB and it is damn slow(it took
> 42hours for postgres to insert and its still running may take about 3
> more hours), can anybody suggest any idea how to reduce the size and
> make it faster.
>
> thanks alot in advance.
>
> --
> Shashi Kiran Reddy. Gireddy,
> Graduate Assistant,
> CBER, University of Alabama.
> http://www.cs.ua.edu/shashi
> Home: 205-752-5137 Cell: 205-657-1438