Hi,
If I run the query
explain analyze select * from ind_uni_100 where a=1 and b=1 and c=1
I get the following plan:
Bitmap Heap Scan on ind_uni_100 (cost=942.50..1411.12 rows=125 width=104)
(actual time=72.556..72.934 rows=116 loops=1)
Recheck Cond: ((c = 1) AND (a = 1) AND (b = 1))
-> Bi
Gene wrote:
Thanks for the suggestion. Actually I went ahead and created a reverse
function using plpgsql, created an index using reverse column and now
my queries use "where reverse(column) like reverse('%2345') and it's
using the index like i hoped it would! Now if I could figure out how
to opt
Thanks for the suggestion. Actually I went ahead and created a reverse
function using plpgsql, created an index using reverse column and now
my queries use "where reverse(column) like reverse('%2345') and it's
using the index like i hoped it would! Now if I could figure out how
to optimize like '
I don't think indexes are going to help you here - with the FULL OUTER
JOINs, the query will have to look at and include each row from each
table you query from anyway, so it's going to choose sequential scans.
In addition, some of the lower join conditions are going to take forever.
What's is
Hi all,
I got this query, I'm having indexes for PropertyId and Dates columns across all the tables, but still it takes ages to get me the result. What indexes would be proposed on this, or I'm helpless?
FROM STG_Property a FULL OUTER JOIN STG_PropConfirmedLogs b
ON (a.PropertyId = b.P
Mikael Carneholm wrote:
For my application there is very little info I can share. Maybe less
than 10 on 100 actually so I not sure it worth it ...
Ok, so 90% of the tables are being written to - this either means that
your application uses very little constants, or that it has access to
c
> For my application there is very little info I can share. Maybe less
than 10 on 100 actually so I not sure it worth it ...
Ok, so 90% of the tables are being written to - this either means that
your application uses very little constants, or that it has access to
constans that are stored somewh
Hi All,
First thanks for your help everyone!
Mikael Carneholm wrote:
Do you really need to create one *DB* per client - that is, is one
schema (in the same DB) per client out of the question? If not, I would
look into moving all reference tables (read-only data, constants and
such) into a
On Sun, 2 Jul 2006, Gene wrote:
can use an index and perform as fast as searching with like '2345%'?
Is the only way to create a reverse function and create an index using
the reverse function and modify queries to use:
where reverse(column) like reverse('%2345') ?
Hmm.. interesting.
Do you really need to create one *DB* per client - that is, is one
schema (in the same DB) per client out of the question? If not, I would
look into moving all reference tables (read-only data, constants and
such) into a common schema (with read permission granted to each
client/role), that way red
10 matches
Mail list logo
