From: | Bruce Momjian <bruce(at)momjian(dot)us> |
---|---|
To: | Pavan Deolasee <pavan(dot)deolasee(at)gmail(dot)com> |
Cc: | Claudio Freire <klaussfreire(at)gmail(dot)com>, pgsql-hackers <pgsql-hackers(at)postgresql(dot)org> |
Subject: | Re: Heap WARM Tuples - Design Draft |
Date: | 2016-08-07 16:55:01 |
Message-ID: | 20160807165501.GA16416@momjian.us |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
On Sun, Aug 7, 2016 at 10:49:45AM -0400, Bruce Momjian wrote:
> OK, crazy idea time --- what if we only do WARM chain additions when all
> indexed values are increasing (with NULLs higher than all values)? (If
> a key is always-increasing, it can't match a previous value in the
> chain.) That avoids the problem of having to check the WARM chain,
> except for the previous tuple, and the problem of pruning removing
> changed rows. It avoids having to check the index for matching key/ctid
> values, and it prevents CREATE INDEX from having to index WARM chain
> values.
>
> Any decreasing value would cause a normal tuple be created.
Actually, when we add the first WARM tuple, we can mark the HOT/WARM
chain as either all-incrementing or all-decrementing. We would need a
bit to indicate that.
Also, it would be possible for keys involved in multi-key indexes to not
match the direction of the chain as long as keys earlier in the index
matched, e.g. key (1,5,6) would be less than (2,1,1) since 1 < 2, even
though 5 > 1. I am not sure it would be worth detecting this.
--
Bruce Momjian <bruce(at)momjian(dot)us> http://momjian.us
EnterpriseDB http://enterprisedb.com
+ As you are, so once was I. As I am, so you will be. +
+ Ancient Roman grave inscription +
From | Date | Subject | |
---|---|---|---|
Next Message | Tom Lane | 2016-08-07 17:57:03 | Re: [sqlsmith] Crash in GetOldestSnapshot() |
Previous Message | Tom Lane | 2016-08-07 16:35:56 | Re: Bogus ANALYZE results for an otherwise-unique column with many nulls |