Re: [PATCH] Re: Proposal to Enable/Disable Index using ALTER INDEX
От | Sami Imseih |
---|---|
Тема | Re: [PATCH] Re: Proposal to Enable/Disable Index using ALTER INDEX |
Дата | |
Msg-id | CAA5RZ0tHnxpTZVFdMvLABaR8k8GSGJg9gXBACksLkKiDFio0Jg@mail.gmail.com обсуждение исходный текст |
Ответ на | Re: [PATCH] Re: Proposal to Enable/Disable Index using ALTER INDEX (Gurjeet Singh <gurjeet@singh.im>) |
Ответы |
Re: [PATCH] Re: Proposal to Enable/Disable Index using ALTER INDEX
|
Список | pgsql-hackers |
> > + indexes. If performance degrades after making an index invisible, it can be easily > > + be made visible using <literal>VISIBLE</literal>. Before making an index invisible, it's recommended > > + to check <structname>pg_stat_user_indexes</structname>.<structfield>idx_scan</structfield> > > + to identify potentially unused indexes. > > I feel ALTER INDEX command reference doc is not the right place for this kind of > operational advice. Is there a better place in documentation for this kind of > advice? Or maybe it needs to be reworded to fit the command reference style? I agree with you. What about we add this wording in the following section [0]? This section discusses techniques for experimenting with indexes. It says, ".... A good deal of experimentation is often necessary. The rest of this section gives some tips for that:...." A discussion about invisible indexes as one of the tools for experimentation can be added here. What do you think? [0] https://www.postgresql.org/docs/current/indexes-examine.html -- Sami Imseih Amazon Web Services (AWS)
В списке pgsql-hackers по дате отправления: