Oracle FAQ | Your Portal to the Oracle Knowledge Grid |
![]() |
![]() |
Home -> Community -> Usenet -> c.d.o.tools -> clobs and possible performance issues
Hi,
This is the current table layout:
create table sn_content (
object_id integer constraint sn_content_object_id_fk references sn_objects(object_id), question_id integer constraint sn_content_question_id_fk references sn_questions(question_id), content varchar(4000), long_content clob, html_p char(1) constraint sn_content_html_p_ck check (html_p in ('t', 'f')), constraint sn_content_pk primary key (object_id, question_id) );
Either content or long_content is being filled with data depending on the size of content. If the content is > 4000 bytes, it is stored in the lief long_content and content is set to null and vice versa.
That has mainly historical reasons - and I would like to get rid of that.
So retrieving content is currently done in this way (in meta pl/sql syntax):
select content into the_content where object_id=:the_object_id and question_id=:the_question_id;
if content is null then
select long_content into the_contet where object_id=:the_object_id and
question_id=:the_question_id;
end if;
Are there any performance implications when removing the content field? The overall amount of queries on sn_content would get smaller, but all queries would end up retrieving data from a clob or stuffing data into a clob. How do queries on tables with clobs perform, mainly inserts and selects? All lookups on sn_content are indexed, there are no range scans.
cheers Dirk Received on Sat Jul 14 2001 - 13:19:11 CDT
![]() |
![]() |