Re: Query with "ILIKE ALL" does not use the index

Thank you Tom On 26 July 2018 at 18:44, Tom Lane wrote: > Nicolas Even writes: >> However when I run the same (as far as I understand it) query but with >> the ALL operator, the index is not used: >> explain analyze select name from totoz where name ilike all(array['%tot%']); > > There's only in

Re: Query with "ILIKE ALL" does not use the index

Hi Matthew, I finally used "WHERE name ILIKE arr[1] AND name ILIKE ALL(arr)" which works well enough for my use case. Thank you Nicolas On 26 July 2018 at 19:22, Matthew Hall wrote: > On Jul 26, 2018, at 9:44 AM, Tom Lane wrote: >> >> Nicolas Even writes: >>> However when I run the same (as f

Re: Query with "ILIKE ALL" does not use the index

On Jul 26, 2018, at 9:44 AM, Tom Lane wrote: > > Nicolas Even writes: >> However when I run the same (as far as I understand it) query but with >> the ALL operator, the index is not used: >> explain analyze select name from totoz where name ilike all(array['%tot%']); > > There's only index supp

Re: Query with "ILIKE ALL" does not use the index

Nicolas Even writes: > However when I run the same (as far as I understand it) query but with > the ALL operator, the index is not used: > explain analyze select name from totoz where name ilike all(array['%tot%']); There's only index support for "op ANY (array)", not "op ALL (array)".

Query with "ILIKE ALL" does not use the index

Hi, I have the following table: Table "public.totoz" Column | Type | Collation | Nullable | Default ---+--+---+--+- name | character varying(512) | | not null | Indexes:

Re: Automated bottleneck detection

Wow, freakin cool, can't wait to start fiddling with pg_wait_sampling. Reckon we can get lightweight locks and spinlocks history with this cool new extension instead of awkwardly and repeatedly querying the pg_stat_activity table. Regards, Michael Vitale Thomas Güttler

Automated bottleneck detection

This sound good. Looks like an automated bootleneck detection could be possible with pg_wait_sampling. Regards, Thomas Am 25.07.2018 um 12:25 schrieb Julien Rouhaud: On Wed, Jul 25, 2018 at 11:14 AM, Thomas Güttler wrote: AFAIK powa is based on pg_stat_statements not on statistical samples