| From: | Michael Paquier <michael(at)paquier(dot)xyz> |
|---|---|
| To: | Andrei Zubkov <zubkov(at)moonset(dot)ru> |
| Cc: | PostgreSQL Hackers <pgsql-hackers(at)lists(dot)postgresql(dot)org> |
| Subject: | Re: [PATCH] pg_statio_all_tables: several rows per table due to invalid TOAST index |
| Date: | 2021-11-30 08:29:12 |
| Message-ID: | YaXg2N5gzoHfQjac@paquier.xyz |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
On Mon, Nov 29, 2021 at 05:04:29PM +0300, Andrei Zubkov wrote:
> However it is possible that the TOAST table will have more than one
> index. For example, this happens when REINDEX CONCURRENTLY operation
> lefts an index in invalid state (indisvalid = false) due to some kind
> of a failure. It's often sufficient to interrupt REINDEX CONCURRENTLY
> operation right after start.
>
> Such index will cause the second row to appear in a
> pg_statio_all_tables view which obvious is unexpected behaviour.
Indeed. I can see that.
> Now we can have several regular indexes and several TOAST-indexes for
> the same table. Statistics for the regular and TOAST indexes is to be
> calculated the same way so I've decided to use a CTE here.
Hmm. Why should we care about invalid indexes at all, including
pg_statio_all_indexes?
--
Michael
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Andrei Zubkov | 2021-11-30 08:57:25 | Re: [PATCH] pg_statio_all_tables: several rows per table due to invalid TOAST index |
| Previous Message | Guillaume Lelarge | 2021-11-30 08:23:44 | Re: Lots of memory allocated when reassigning Large Objects |