On Sun, Jan 3, 2016 at 9:05 PM, Ian Barwick <ian@2ndquadrant.com> wrote:
> Like the docs say, vacuumdb is a "wrapper around the SQL command VACUUM"
> which I used to use in dim-and-distant pre-autovacuum days came for cronjobs,
> but until messing around with pg_upgrade recently I hadn't really had much
> use for it. Anyway, the docs also say "There is no effective difference
> between vacuuming and analyzing databases via this utility and via other
> methods for accessing the server", which IMHO seems a bit out-of-date as
> it now does two things which you can't do directly via e.g. psql:
>
> - generate statistics in stages (9.4)
> - parallel vacuum (9.5)
>
> Attached patches (for 9.4 and 9.5/HEAD) update the description to make
> clear that it now does a bit more than just execute a single command.
So I think this is a good change, but...
+ server. However it does provide additional functionality for generating
+ statistics in stages, which is useful when optimizing a newly populated
+ database, and for executing vacuum or analyze commands in parallel.
...it's not entirely clear to which portion of the sentence the clause
at the end applies. Is executing vacuum or analyze commands in
parallel another case in which generating statistics in stages is
helpful, or is it another thing altogether that vacuumdb does? I know
the answer, and anyone who is reasonably clueful will too, but it
seems to me that we could probably tweak the phrasing here so that a
non-clueful person wouldn't go astray.
--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company