Hi,
In Documentum 6.7.1050.0105 AIX.Oracle, we have this dql query :
select r_object_id, object_name,r_creation_date,ft_multi_01, ft_texte_06, title, ft_texte_03, subject, ft_resume from ft_doc_doc where UPPER(ft_texte_04)= '00068425' and UPPER(ft_mono_01) = 'ID_ENT'
In dqman, it takes 45 sec.
When we use the ENABLE(ROW_BASED), it is fast.
In Oracle, it takes less than 1 sec (we use the SQL value of the dql query) :
SELECT ALL
ft_doc_doc.r_object_id,
ft_doc_doc.object_name,
ft_doc_doc.r_creation_date,
dm_repeating.ft_multi_01,
ft_doc_doc.ft_texte_06,
ft_doc_doc.title,
ft_doc_doc.ft_texte_03,
ft_doc_doc.subject,
ft_doc_doc.ft_resume
FROM
ft_doc_doc_sp ft_doc_doc, ft_doc_doc_rp dm_repeating
WHERE
((upper(ft_doc_doc.ft_texte_04)='00068425') and (upper(ft_doc_doc.ft_mono_01)='ID_ENT')) and
(ft_doc_doc.i_has_folder = 1 and ft_doc_doc.i_is_deleted = 0) and
dm_repeating.r_object_id=ft_doc_doc.r_object_id
ORDER BY
1
It seems that transormation of a "row based format" to an "object based format" is very slow in documentum.
Do you have any idea to solve this issue ?
Regards