Re: : Performance Improvement Strategy
| От | Scott Marlowe |
|---|---|
| Тема | Re: : Performance Improvement Strategy |
| Дата | |
| Msg-id | CAOR=d=1sfgPk0z2zwTqEMapu+5=VHWAB+o9azHs-LytXgsWNnQ@mail.gmail.com обсуждение исходный текст |
| Ответ на | Re: : Performance Improvement Strategy (Greg Smith <greg@2ndQuadrant.com>) |
| Ответы |
Re: : Performance Improvement Strategy
|
| Список | pgsql-performance |
On Wed, Sep 21, 2011 at 11:57 AM, Greg Smith <greg@2ndquadrant.com> wrote: > On 09/21/2011 12:13 PM, Venkat Balaji wrote: >> >> I as a DBA, suggested to perform VACUUM FULL and RE-INDEXING + ANALYZE to >> ensure that IO performance and Indexing performance would be good > > > Read http://wiki.postgresql.org/wiki/VACUUM_FULL before you run VACUUM FULL. > You probably don't want to do that. A multi-gigabyte table can easily be > unavailable for several hours if you execute VACUUM FULL against it. > CLUSTER is almost always faster. It used to be that cluster on a very randomly ordered table was much slower than doing something like select * into newtable from oldtable order by col1, col2; Is that still the case in 9.0/9.1?
В списке pgsql-performance по дате отправления: