Re: Add mode column to pg_stat_progress_vacuum
| От | Michael Paquier |
|---|---|
| Тема | Re: Add mode column to pg_stat_progress_vacuum |
| Дата | |
| Msg-id | aPB_24M2xvzq4DTU@paquier.xyz обсуждение исходный текст |
| Ответ на | Re: Add mode column to pg_stat_progress_vacuum (Sami Imseih <samimseih@gmail.com>) |
| Ответы |
Re: Add mode column to pg_stat_progress_vacuum
|
| Список | pgsql-hackers |
On Thu, Oct 09, 2025 at 10:07:17AM -0500, Sami Imseih wrote: >> IIUC are you suggesting I add such a last_autovacuum_reason column to >> pg_stat_all_tables, which would be populated with one of the following >> values? >> - autovacuum_vacuum_threshold >> - autovacuum_vacuum_insert_threshold >> - autovacuum_freeze_max_age >> - autovacuum_multixact_freeze_max_age > > This should be a separate discussion. But, I would think the > counters will be n_aggressive, n_wraparound and n_failsafe. Depends, I guess (separate discussion it should be, but I count not resist). If you had this information available in the cumulative stats, what should be a "correct" set of numbers, and what could be tuned to redirect the system so as it gets to a better set of numbers. Wraparound autovacuums, for one, don't seem really relevant to know about in an aggregated way. Coming back to the original proposal. Knowing about the state we are kicking an autovacuum worker job for a set of tables in the progress view would be definitely a nice thing. -- Michael
Вложения
В списке pgsql-hackers по дате отправления: