Re: Size for vacuum_mem
От | Medi Montaseri |
---|---|
Тема | Re: Size for vacuum_mem |
Дата | |
Msg-id | 3DEFA233.6050403@intransa.com обсуждение исходный текст |
Ответ на | Size for vacuum_mem (Francisco Reyes <lists@natserv.com>) |
Список | pgsql-general |
Can someone please explain the relatiionship between operation of a vacuum-ing job and vacuum_mem setting in non-shared memfory configuration. Ie from a capacity planning point of view, how is vacuum_mem's size is related to this operation. The current 8k default seems so low... Thanks Shridhar Daithankar wrote: >On 4 Dec 2002 at 19:38, Neil Conway wrote: > > > >>>Currently a vacuum full takes 3+ hours and very soon the amount of data >>>will increase. >>> >>> >>Do you need to use VACUUM FULL? >> >> > >Let me elaborate this statement. > >1) You need vacuum full only to recover space from deleted tuples. Unless the >database has undergone major deletion somewhere 'vacuum full' might be a cannon >to kill an ant. > >2) You should consider vacuuming tablewise. Vacuum is useful only for those >tables which change at a faster rate. A lookup or archive table might not need >vacuum. Just vacuum the tables which are heavily updated/deleted/inserted. And >by heavily, I mean heavily in terms of tuples. Inserting a single 200MB BLOB >and vacuuming the table might not yield any performance improvement.. > > HTH > >Bye > Shridhar > >-- >blithwapping: Using anything BUT a hammer to hammer a nail into the wall, such >as shoes, lamp bases, doorstops, etc. -- "Sniglets", Rich Hall & Friends > > >---------------------------(end of broadcast)--------------------------- >TIP 2: you can get off all lists at once with the unregister command > (send "unregister YourEmailAddressHere" to majordomo@postgresql.org) > >
В списке pgsql-general по дате отправления: