| From: | Heikki Linnakangas <heikki(dot)linnakangas(at)enterprisedb(dot)com> |
|---|---|
| To: | Simon Riggs <simon(at)2ndQuadrant(dot)com> |
| Cc: | Bruce Momjian <bruce(at)momjian(dot)us>, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>, PostgreSQL-development <pgsql-hackers(at)postgresql(dot)org> |
| Subject: | Re: 16-bit page checksums for 9.2 |
| Date: | 2012-02-06 10:44:27 |
| Message-ID: | 4F2FAF0B.7000804@enterprisedb.com |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
On 06.02.2012 11:25, Simon Riggs wrote:
> On Mon, Feb 6, 2012 at 10:02 AM, Heikki Linnakangas
> <heikki(dot)linnakangas(at)enterprisedb(dot)com> wrote:
>> Good idea. However, the followup idea to that discussion was to not only
>> avoid the I/O needed to mark tuples as frozen, but to avoid xid wraparound
>> altogether, by allowing clog to grow indefinitely. You do want to freeze at
>> some point of course, to truncate the clog, but it would be nice to not have
>> a hard limit. The way to do that is to store an xid "epoch" in the page
>> header, so that Xids are effectively 64-bits wide, even though the xid
>> fields on the tuple header are only 32-bits wide. That does require a new
>> field in the page header.
>
> We wouldn't need to do that would we?
Huh? Do you mean that we wouldn't need to implement that feature?
--
Heikki Linnakangas
EnterpriseDB http://www.enterprisedb.com
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Fujii Masao | 2012-02-06 11:08:15 | Re: double writes using "double-write buffer" approach [WIP] |
| Previous Message | Simon Riggs | 2012-02-06 10:25:05 | Re: 16-bit page checksums for 9.2 |