Re: Table-level log_autovacuum_min_duration

Hi, Michael-san

> An updated patch is attached,

I'm sorry for confusing you.

I think you don't have to implement this code to disable this
feature with using value "-2".Because this use case is a rare case,
and there is a practical workaround using huge value like "2e9".
(You suggested "2e9" to me, didn't you? :) ) So, please remove this code.

> Well, I see your point but this is not completely true: we could as
> well rely entirely on this parameter instead of VACOPT_VERBOSE to
> determine if autovacuum, a vacuum or an analyze are in verbose mode,
> and remove VACOPT_VERBOSE, but I can imagine people complaining if
> VACOPT_VERBOSE is removed. So let's set it up unconditionally to -1 in
> gram.y for now. However if people think that it is fine to remove
> VACOPT_VERBOSE, we could use exclusively this parameter in VacuumStmt.
> Or even add sanity checks at the top of vacuum() to ensure that
> VACOPT_VERBOSE is set only when log_min_duration is positive.
> Additional opinions on this matter are welcome.

I understand your point at last. :)

You mean that ...
Log_autovacuum_min_duration assumes a role of VACOPT_VERBOSE.
Even if this parameter never use currently for manual vacuum,
log_autovacuum_min_duration should be set zero(anytime output)
when we executes "VACUUM(or ANALYZE) VERBOSE".
Is my understanding correct? If so,it sounds logical.

If we can abolish VERBOSE option,
I think it's ideal that we will prepare a new parameter like
a log_min_duration_vacuum(and log_min_duration_analyze) which
integrating "VERBOSE feature" and "log_autovacuum_min_duration".

Regards,
---
Naoya Anzai