Re: UNDO and in-place update

From: Greg Stark <stark(at)mit(dot)edu>
To: Robert Haas <robertmhaas(at)gmail(dot)com>
Cc: Peter Geoghegan <pg(at)heroku(dot)com>, "pgsql-hackers(at)postgresql(dot)org" <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: UNDO and in-place update
Date: 2016-11-25 21:14:32
Message-ID: CAM-w4HPrjNPA_VfLAde0NkUiaauycOyqBC=J0raBkoo8d87F0Q@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On 24 November 2016 at 23:03, Robert Haas <robertmhaas(at)gmail(dot)com> wrote:
>> For snapshot isolation Oracle has yet a *third* copy of the data in a
>> space called the "rollback segment(s)".
>
> My understanding is that this isn't correct. I think the rollback
> segments are what they call the thing that stores UNDO. See e.g.
> http://ss64.com/ora/syntax-redo.html

It looks like you're right. Rollback segments and Undo segments are
two different pieces of code but one is just the old way and the other
the new way of managing the same data. You can't have both active in
the same database at the same time. I'm a bit confused because I
distinctly remembered an UNDO log back in the 8i days as well but
apparently that's just me imagining things. UNDO segments were
introduced in 9i.

This explained a bunch
http://satya-dba.blogspot.ie/2009/09/undo-tablespace-undo-management.html

--
greg

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Tom Lane 2016-11-25 21:34:39 Re: [sqlsmith] Failed assertion in parallel worker in ExecInitSubPlan
Previous Message Tomas Vondra 2016-11-25 19:54:55 Re: confusing checkpoint_flush_after / bgwriter_flush_after