Re: auto vacuuming - Mailing list pgsql-admin

From Tom Lane
Subject Re: auto vacuuming
Date
Msg-id 27699.1144035580@sss.pgh.pa.us
Whole thread Raw
In response to Re: auto vacuuming  (Alvaro Herrera <alvherre@commandprompt.com>)
Responses Re: auto vacuuming  (John DeSoi <desoi@pgedit.com>)
List pgsql-admin
Alvaro Herrera <alvherre@commandprompt.com> writes:
> Tom Lane wrote:
>> The handwriting on the wall says that autovac will soon be on by
>> default, and perhaps become not-disablable some day after that
>> (like the second or third time we hear from someone who's lost
>> their data to XID wraparound after disabling it).

> I don't think we will see much people losing data to Xid wraparound
> anymore, now that the server refuses to work when the wraparound is
> imminent.

Well, how about the second or third time we get sued by someone whose
"must be up 24x7" database shuts down for lack of proper vacuuming?
I do think autovac is the wave of the future.  The only reason it's
disablable now is that we don't think we've got all the bugs out.

If you read the old Berkeley Postgres papers, you'll see that a
"vacuum daemon" was always part of the system's basic design.  (Hey
Elein, or anyone else who was there then --- was there ever a working
vacuum daemon, or was it just on paper?)

            regards, tom lane

pgsql-admin by date:

Previous
From: Alvaro Herrera
Date:
Subject: Re: auto vacuuming
Next
From: adey
Date:
Subject: Re: Bloated pg_shdepend_depender_index