Horton

Re: Potential G2-item cycles under serializable isolation

Peter Geoghegan <pg@bowt.ie>

View thread View on postgr.es
From: Peter Geoghegan <pg@bowt.ie>
To: Thomas Munro <thomas.munro@gmail.com>
Cc: Kevin Grittner <kgrittn@gmail.com>, Jeff Davis <pgsql@j-davis.com>, PostgreSQL mailing lists <pgsql-bugs@lists.postgresql.org>, Kyle Kingsbury <aphyr@jepsen.io>
Date: 2020-06-09T02:07:46Z
Lists: pgsql-bugs 
On Mon, Jun 8, 2020 at 7:01 PM Thomas Munro <thomas.munro@gmail.com> wrote:
> Right, the only change was to move things around a bit to suport new
> table AMs.  Speaking of which, it looks like the new comment atop
> CheckForSerializableConflictOut() could use some adjustment.  It says
> "A table AM is reading a tuple that has been modified.  After
> determining that it is visible to us, it should call this function..."
> but it seems the truth is a bit more complicated than that.

Right. I think that you can go ahead and change it without further input here.

-- 
Peter Geoghegan

Commits

  1. Doc: Add references for SI and SSI.

  2. Improve comments for [Heap]CheckForSerializableConflictOut().

  3. Avoid update conflict out serialization anomalies.

  4. Implement genuine serializable isolation level.