Re: [GENERAL] Reducing memory usage of insert into select operations?

Douglas McNaught writes:

> It does seem that reducing work_mem might help you, but others on this

I reduced it from 256MB to 64MB. It seems it is helping.
At 256MB the usage per DB connection instance was upwards of 12GB. At 64MB
so far is around 7GB. I just reduced it further to 32MB and see how that
works.

> It would probably be helpful for you to post the EXPLAIN output from
Hash Join (cost=712213.57..27293913.33 rows=234402352 width=24)
Hash Cond: (coh.id = ids.id)
-> Hash Join (cost=551387.26..18799378.16 rows=234402352 width=22)
Hash Cond: (coh.user_id = ca.user_id)
-> Seq Scan on customer_original_historical coh
(cost=0.00..6702501.40 rows=234402352 width=47)
Filter: (yearmo > '200703'::bpchar)
-> Hash (cost=268355.67..268355.67 rows=14637567 width=32)
-> Seq Scan on cards ca
(cost=0.00..268355.67 rows=14637567 width=32)
-> Hash (cost=77883.25..77883.25 rows=5055525 width=6)
-> Seq Scan on customer_ids ids
(cost=0.00..77883.25 rows=5055525 width=6)

--
Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-general