Try
SELECT DISTINCT
rather than SELECT
That should return a result with unique records.
Madison Kelly wrote:
Hi all,
I've got a query that looks through a table I use for my little
search engine. It's something of a reverse-index but not quite, where
a proper reverse index would have
Isak Hansen wrote:
First thing I notice is that your query plans seem to only use one index.
Postgres should be able to combine the timestamp and account_id
indexes in that first query, if the optimizer thought there was a
point in doing so?
Absolutely brilliant - this solved my issue,
this optimisation peculiarity be due to the combination of indexed
columns in the query? Also, Is there a way I can 'force' the planner to
perform an 'index scan - sort - limit' or even better an 'index scan
- limit - sort'?
Any pointers / assistance appreciated.
Kind Regards,
Shaun Johnston