pg.ddx.io  pgsql-sql@postgresql.org mailing list archive  
help / color / mirror / Atom feed
short query becomes long
2+ messages / 2 participants
[nested] [flat]

* short query becomes long
@ 2000-06-01 18:24 mikeo <mikeo@spectrumtelecorp.com>
  2000-06-02 02:31 ` Re: short query becomes long Tom Lane <tgl@sss.pgh.pa.us>
  0 siblings, 1 reply; 2+ messages in thread

From: mikeo @ 2000-06-01 18:24 UTC (permalink / raw)
  To: pgsql-sql

hi,
  we have a weird situation here.  we have a table of approx. 10k rows
representing accumulated activity by specific customers.  as information 
is gathered those customers rows are updated.  the number of rows does not
increase unless we get a new customer so that is not a factor.  the table 
is defined as follows:

      Table "account_summary_02"
  Attribute  |    Type     | Modifier
-------------+-------------+----------
 bill_br_id  | bigint      | not null
 cust_id     | varchar(15) | not null
 btn_id      | varchar(15) | not null
 ln_id       | varchar(15) | not null
 ct_key      | float8      | not null
 as_quantity | float8      | not null
 as_charges  | float8      | not null
 as_count    | float8      | not null
Index: account_summary_02_unq_idx

the index is on the first 5 columns.  here's the situation.  after about
50,000
updates, which fly right along, the process begins to really bog down.  we
perform 
a vacuum analzye and it speeds right up again.  my question is, is there a way
to perform these updates, potentially 500k to 1 million in a day, without
having 
to vacuum so frequently?  maybe some setting or parameter to be changed?
the update
query is doing an index scan.  

mikeo 




^ permalink  raw  reply  [nested|flat] 2+ messages in thread

* Re: short query becomes long
  2000-06-01 18:24 short query becomes long mikeo <mikeo@spectrumtelecorp.com>
@ 2000-06-02 02:31 ` Tom Lane <tgl@sss.pgh.pa.us>
  0 siblings, 0 replies; 2+ messages in thread

From: Tom Lane @ 2000-06-02 02:31 UTC (permalink / raw)
  To: mikeo <mikeo@spectrumtelecorp.com>; +Cc: pgsql-sql

mikeo <mikeo@spectrumtelecorp.com> writes:
> after about 50,000 updates, which fly right along, the process begins
> to really bog down.  we perform a vacuum analzye and it speeds right
> up again.

A plain "vacuum" should do the job in a bit less time.  The problem
is you need to reclaim the space occupied by deleted versions of rows.
No way around that with the current storage manager: vacuum is the
only way to get rid of the wasted space.

			regards, tom lane



^ permalink  raw  reply  [nested|flat] 2+ messages in thread


end of thread, other threads:[~2000-06-02 02:31 UTC | newest]

Thread overview: 2+ messages (download: mbox mbox.gz follow: Atom feed)
-- links below jump to the message on this page --
2000-06-01 18:24 short query becomes long mikeo <mikeo@spectrumtelecorp.com>
2000-06-02 02:31 ` Tom Lane <tgl@sss.pgh.pa.us>

This inbox is served by DDX for PostgreSQL; see mirroring instructions
for how to clone and mirror all data and code used for this inbox