Re: Vacuum statistics
От | Alena Rybakina |
---|---|
Тема | Re: Vacuum statistics |
Дата | |
Msg-id | e519f256-343f-4390-9ab2-c4d91eac76b2@postgrespro.ru обсуждение исходный текст |
Ответ на | Re: Vacuum statistics (Kirill Reshke <reshkekirill@gmail.com>) |
Список | pgsql-hackers |
On 22.08.2024 07:29, Kirill Reshke wrote:
On Thu, 22 Aug 2024 at 07:48, jian he <jian.universality@gmail.com> wrote:On Wed, Aug 21, 2024 at 6:37 AM Alena Rybakina <a.rybakina@postgrespro.ru> wrote:We check it there: "tabentry->vacuum_ext.type != type". Or were you talking about something else? On 19.08.2024 12:32, jian he wrote: in pg_stats_vacuum if (type == PGSTAT_EXTVAC_INDEX || type == PGSTAT_EXTVAC_HEAP) { Oid relid = PG_GETARG_OID(1); /* Load table statistics for specified database. */ if (OidIsValid(relid)) { tabentry = fetch_dbstat_tabentry(dbid, relid); if (tabentry == NULL || tabentry->vacuum_ext.type != type) /* Table don't exists or isn't an heap relation. */ PG_RETURN_NULL(); tuplestore_put_for_relation(relid, rsinfo, tabentry); } else { } So for functions pg_stat_vacuum_indexes and pg_stat_vacuum_tables, it seems you didn't check "relid" 's relkind, you may need to use get_rel_relkind. --hi. I mentioned some points at [1], Please check the attached patchset to address these issues. there are four occurrences of "CurrentDatabaseId", i am still confused with usage of CurrentDatabaseId. also please don't top-post, otherwise the archive, like [2] is not easier to read for future readers. generally you quote first, then reply. [1] https://postgr.es/m/CACJufxHb_YGCp=pVH6DZcpk9yML+SueffPeaRbX2LzXZVahd_w@mail.gmail.com [2] https://postgr.es/m/78394e29-a900-4af4-b5ce-d6eb2d263fad@postgrespro.ruHi, your points are valid. Regarding 0003, I also wanted to object database naming in a regression test during my review but for some reason didn't.Now, as soon as we already need to change it, I suggest we also change regression_statistic_vacuum_db1 to something less generic. Maybe regression_statistic_vacuum_db_unaffected.
Hi! I fixed it in the new version of the patch [0]. Feel free to review it!
To be honest, I still doubt that the current database names (regression_statistic_vacuum_db and regression_statistic_vacuum_db1) can be easily generated, but if you insist on renaming, I will do it.
[0] https://www.postgresql.org/message-id/c4e4e305-7119-4183-b49a-d7092f4efba3%40postgrespro.ru
-- Regards, Alena Rybakina Postgres Professional: http://www.postgrespro.com The Russian Postgres Company
В списке pgsql-hackers по дате отправления: