Re: Query - CPU issue

Kevin Grittner-5 wrote
> What is the longest you have let it run, in hours or minutes?

I let it run for about 10 minutes and killed it.

Kevin Grittner-5 wrote
> By the way, IMMUTABLE has to be wrong here, since the results
> depend on the state of the database.  STABLE is likely the right
> designation.

The data will not change unless I do a reload of the data set manually. In
that case, is IMMUTABLE wrong?
Here is the EXPLAIN (no analyze)
explain
select distinct geoip_city(src_ip) , src_ip
from alert where timestamp>=1378512000 and timestamp < 1378598400;
QUERY PLAN
---------------------------------------------------------------------------------------------
Unique (cost=3815.75..3894.61 rows=39 width=8)
-> Sort (cost=3815.75..3842.04 rows=10515 width=8)
Sort Key: (geoip_city(src_ip)), src_ip
-> Index Scan using idx_alert_ts on alert (cost=0.29..3113.34
rows=10515 width=8)
Index Cond: (("timestamp" >= 1378512000) AND ("timestamp" <
1378598400))
(5 rows)

--
View this message in context: http://postgresql.1045698.n5.nabble.com/Query-CPU-issue-tp5771421p5771552.html
Sent from the PostgreSQL - general mailing list archive at Nabble.com.