Re: How to find greatest record before known values fast

On Fri, Oct 3, 2014 at 3:28 PM, Andrus <kobruleht2(at)hot(dot)ee> wrote:
> Hi!
>
> Thank you for explanations.
>
>> the char type pads out the fields on disk.
>
>
> It looks like you wrote that char takes more disk space.
>
> from
>
> http://www.pgcon.org/2013/schedule/attachments/269_tour-of-postgresql-data-types.pdf
>
> page 28:
>
> Unlike many
> databases, char(n) is NOT stored as afixed-sizedfield in
> Postgres. It is treated exactly the sameas
> varchar(n)except for being padded
>
> So char type does not take more space than varchar.

I beg to differ:

postgres=# create table t1(v char(100));
CREATE TABLE
postgres=# create table t2(v varchar(100));
CREATE TABLE
postgres=# insert into t1 select '' from generate_series(1,1000000);
INSERT 0 1000000
Time: 5951.023 ms
postgres=# insert into t2 select '' from generate_series(1,1000000);
INSERT 0 1000000
Time: 2083.323 ms
postgres=# select pg_size_pretty(pg_relation_size(oid)) from pg_class
where relname = 't1';
pg_size_pretty
────────────────
128 MB
(1 row)

postgres=# select pg_size_pretty(pg_relation_size(oid)) from pg_class
where relname = 't2';
pg_size_pretty
────────────────
35 MB
(1 row)

merlin