On 14 November 2011 15:07, Robert Haas
<robertmhaas@gmail.com> wrote:
On Mon, Nov 14, 2011 at 10:05 AM, Thom Brown <
thom@linux.com> wrote:
> On 14 November
2011 13:32, Robert Haas <
robertmhaas@gmail.com> wrote:
>>
>> But Tom's point about XIDs and LSNs seems like it kind of puts a
>> bullet through the heart of the whole idea.
>
> What about having database-level XIDs rather than cluster-level? Is that
> remotely feasible?
Maybe. You'd need a set separate set for shared catalogs, too. It
seems like a heck of a lot of work, though, especially since (IME,
anyway) most people only really one run one database per cluster.
Thought it would be a lot of work. Well one benefit I could potentially see is paving the way for per-database replication. But I'll let this dream go as it's clearly not something to realistically pursue.
--
Thom Brown
Twitter: @darkixion
IRC (freenode): dark_ixion
Registered Linux user: #516935
EnterpriseDB UK:
http://www.enterprisedb.com The Enterprise PostgreSQL Company