On 09/19/2013 02:41 PM, Kevin Grittner wrote:
> Hannu Krosing <hannu@2ndQuadrant.com> wrote:
>
>> the patch solves the general problem of "when the table changes,
>> refresh"
>>
>> After saying it like this, the problem could also be solved by
>> including xmin(s) for rows from underlying table(s)in the
>> matview.
>>
>> Would this be a better approach ?
> Now you're moving from REFRESH territory into incremental
> maintenance. There have been a very large number of papers written
> on that topic, I have reviewed the literature, and I have picked a
> technique which looks like it will be fast and reliable -- based on
> relational algebra. Let's save discussion of alternatives such as
> you're suggesting here, for when I get past the easy stuff ... like
> just refreshing a view so that the new contents are the same as
> what you would see if you re-ran the query defining the matview.
I'm sure that comparing xmin records would work exactly
similar to binary comparisons of records for detecting
possible change in 99.999% of real-world cases.
I am also pretty sure it would have its own cans of worms
all over again when trying to actually implement this in
matviews :)
Cheers
--
Hannu Krosing
PostgreSQL Consultant
Performance, Scalability and High Availability
2ndQuadrant Nordic OÜ