|From:||David Steele <david(at)pgmasters(dot)net>|
|To:||Kevin Grittner <kgrittn(at)ymail(dot)com>, Rod Taylor <rod(dot)taylor(at)gmail(dot)com>|
|Cc:||Magnus Hagander <magnus(at)hagander(dot)net>, Andres Freund <andres(at)2ndquadrant(dot)com>, Bruce Momjian <bruce(at)momjian(dot)us>, PostgreSQL-development <pgsql-hackers(at)postgresql(dot)org>, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>|
|Subject:||Re: Allow "snapshot too old" error, to prevent bloat|
|Views:||Raw Message | Whole Thread | Download mbox | Resend email|
On 2/19/15 1:54 PM, Kevin Grittner wrote:
> Rod Taylor <rod(dot)taylor(at)gmail(dot)com> wrote:
>> Would pg_dump be able to opt-out of such a restriction?
> I don't see how, since vacuum would be removing recently dead
> tuples that are still visible; the alternative to getting a
> "snapshot too old" error when reading a page which could be
> affected is to return incorrect results, and nobody wants that.
> The best you could do if you wanted to run pg_dump (or similar) and
> it might take more time than your old_snapshot_threshold would be
> to increase the threshold, reload, dump, set it back to the
> "normal" setting, and reload again.
While I think pg_dump is a great solution for small to medium
installations, there are a number of excellent file-based backup options
available. Anyone who is seriously worried about bloat (or locking)
should be looking to those solutions.
- David Steele
|Next Message||Tom Lane||2015-02-20 03:01:12||Re: Exposing the stats snapshot timestamp to SQL|
|Previous Message||Matt Kelly||2015-02-20 02:56:19||Re: Exposing the stats snapshot timestamp to SQL|