Thread: BLOB and OID
Hi, OID being a 4 byte int seems limited to indexing for binary objects. More precisely I use BLOB to store images in the database and link these objects to another table using the OID as FK. If I run out of OID there will be no way to index new images. Of course when not automatically creating OID for all objects in the database you will not easily run out of index values, but still.. Is there an alternative way of indexing BLOB? Or can I extend the range of OID? Lolke
yes, there is one, but i dont know how would you modify the PostgreSQL code.
You can implement a GUID datatype and can use it as indexing which guarantees uniqueness and can be stored in double format.
thanks,
vish
You can implement a GUID datatype and can use it as indexing which guarantees uniqueness and can be stored in double format.
thanks,
vish
On 11/4/05, Lolke B. Dijkstra <lolke.dijkstra@chello.nl> wrote:
Hi,
OID being a 4 byte int seems limited to indexing for binary objects.
More precisely I use BLOB to store images in the database and link these
objects to another table using the OID as FK.
If I run out of OID there will be no way to index new images. Of course
when not automatically creating OID for all objects in the database you
will not easily run out of index values, but still..
Is there an alternative way of indexing BLOB? Or can I extend the range
of OID?
Lolke
---------------------------(end of broadcast)---------------------------
TIP 2: Don't 'kill -9' the postmaster