| From: | Joseph Shraibman <jks(at)selectacast(dot)net> |
|---|---|
| To: | Greg Stark <gsstark(at)mit(dot)edu> |
| Cc: | Jim Nasby <decibel(at)decibel(dot)org>, pgsql-general(at)postgresql(dot)org |
| Subject: | Re: Idea for vacuuming |
| Date: | 2006-06-26 03:49:04 |
| Message-ID: | 449F5930.5070108@selectacast.net |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-general |
The verbose output shows the table being vacuumed last. Maybe it
changed after 8.0
Greg Stark wrote:
> Jim Nasby <decibel(at)decibel(dot)org> writes:
>
>>> My RFE: When vacuuming a table, pg should try to vacuum the primary key
>>> first. If that results in 0 recovered entries, then assume the table has no
>>> updates/deletes and skip the rest of that table.
>
> That makes no sense. Vacuum starts by scanning the table itself, not the
> indexes. It only goes to the indexes after it has found tuples that need
> cleaning up. There's nothing to look at in the indexes that would tell it
> whether there are any tuples to clean up.
>
| From | Date | Subject | |
|---|---|---|---|
| Next Message | MG | 2006-06-26 07:36:52 | RAID + PostgreSQL? |
| Previous Message | Tom Lane | 2006-06-26 02:15:10 | Re: Casting and Timestamp |