| From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
|---|---|
| To: | Pete Yunker <peter(dot)yunker(at)homejunction(dot)com> |
| Cc: | dinesh kumar <dineshkumar02(at)gmail(dot)com>, AI Rumman <rummandba(at)gmail(dot)com>, pgsql-general General <pgsql-general(at)postgresql(dot)org> |
| Subject: | Re: when do I analyze after concurrent index creation? |
| Date: | 2013-10-18 15:57:33 |
| Message-ID: | 8940.1382111853@sss.pgh.pa.us |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-general |
Pete Yunker <peter(dot)yunker(at)homejunction(dot)com> writes:
> Would a simple multi-column index be considered an 'expression' in this context, meaning that an ANALYZE should be issued after the creation of such an index?
No. Of course, if one of its columns were an expression, then that would
be of interest for ANALYZE.
There has been some talk of collecting column-correlation statistics for
sets of columns listed in multi-column indexes. So it's possible that in
some future PG version, creating a multi-column index will be a reason to
re-ANALYZE. But it isn't today.
regards, tom lane
| From | Date | Subject | |
|---|---|---|---|
| Next Message | bobJobS | 2013-10-18 16:28:49 | Analyze during a transaction |
| Previous Message | Pete Yunker | 2013-10-18 15:50:03 | Re: when do I analyze after concurrent index creation? |