Re: hash function in Postgres

On 01/23/2015 10:42 PM, Ravi Kiran wrote:
> hi,
>
>
> I want to know what kind of hash function postgresql uses while joining.
> I was debugging through gdb, I found out that it is not using bob
> jenkins hash function but a different hash function *hash_uint32() and
> hash_any() *functions if the joining attribute is an integer, and a
> different kind of hash function for a different type of joining attribute.
>
> I want to know whether the hash functions will change if the number of
> tuples in the table are very large or very low, and if it changes,
> please tell me what hash function it uses if the tuples are very large
> and what hash function it uses if the number is very low, also I came to
> know that the hash function will change depending on the type of the
> attribute on which the join takes place, but will it always remains the
> same for the integer type of joining attribute or will it it change.

Seems a good starting point is here:

http://git.postgresql.org/gitweb/?p=postgresql.git;a=tree;f=src/backend/access/hash;h=6e44d337d3dc3e8d5439e2dbbc54ce0c33d3b5d4;hb=8ca336f4ac3f08a5f23e76c6e9a5f2c8064f5883

In particular, hashfunc.c

>
> thanks
>
>

--
Adrian Klaver
adrian(dot)klaver(at)aklaver(dot)com