Re: Slow update
От | Grzegorz Jaśkiewicz |
---|---|
Тема | Re: Slow update |
Дата | |
Msg-id | 2f4958ff0902090507p34ae66f0ha6255d461521c3f6@mail.gmail.com обсуждение исходный текст |
Ответ на | Re: Slow update (Herouth Maoz <herouth@unicell.co.il>) |
Ответы |
Re: Slow update
Re: Slow update |
Список | pgsql-general |
On Mon, Feb 9, 2009 at 12:50 PM, Herouth Maoz <herouth@unicell.co.il> wrote: > I hope someone can clue me in based on the results of explain analyze. Did you have a chance to run vmstat on it, and post it here ? Maybe - if db resides on the same disc with everything else, something (ab)uses that much io, and it has to wait. Also, I don't know - but personaly I didn't like the line in explain: -> Bitmap Index Scan on billing_msisdn_sme_reference (cost=0.00..24.70 rows=389 width=0) (actual time=2 1.418..21.418 rows=252 loops=151332) Index Cond: ((b.msisdn)::text = (("substring"((rb.msisdn)::text, 1, 0) || '972'::text) || "substrin g"((rb.msisdn)::text, 2))) But the cost is next to none, so that's not it. -- GJ
В списке pgsql-general по дате отправления: