Hi Justin,
My apologies, I missed that.
Yes, I change work mem to 2GB but didn't see any difference. So, as your suggestion removed the distinct on pk and added a multi-column index so query planner did index-only can that is fixed the issue and query completed in 1Min.
Best regards,
Rj
On Tuesday, June 16, 2020, 01:47:21 PM PDT, Justin Pryzby <pryzby@telsasoft.com> wrote:
On Tue, Jun 16, 2020 at 08:35:31PM +0000, Nagaraj Raj wrote:
> I wrote a simple query, and it is taking too long, not sure what is wrong in it, even its not giving EXPLAIN ANALYZE.
Is this related to last week's question ?
https://www.postgresql.org/message-id/1211705382.726951.1592174752720%40mail.yahoo.comWas that issue resolved ?
I didn't see answers to a few questions I asked there.
> select T0."physical_address_sid", T0."individual_entity_proxy_id", T2."infrrd_hh_rank_nbr"
> from "cms_prospects".PROSPECT T0
> --inner join "sas_prs_tmp".DEDUPE3583E3F18 T1 on T0."individual_entity_proxy_id" = T1."individual_entity_proxy_id"
> left join "cms_prospects".INDIVIDUAL_DEMOGRAPHIC T2 on T0."individual_entity_proxy_id" = T2."individual_entity_proxy_id"
>
> "Merge Left Join (cost=55.96..18147747.08 rows=213620928 width=20)"
> " Merge Cond: (t0.individual_entity_proxy_id = t2.individual_entity_proxy_id)"
> " -> Index Scan using pk_prospect on prospect t0 (cost=0.57..10831606.89 rows=213620928 width=16)"
> " -> Index Only Scan using indxp_individual_demo_infrrd_hh_rank_nbr on individual_demographic t2 (cost=0.57..5013756.93 rows=260652064 width=12)"
>
> Any suggestions or help would be highly appreciated.
>
> Best regards,
> Rj
