Re: selecting random rows

Joseph Shraibman wrote:

> Dennis Gearon wrote:
>
>> If you have a nice small Primary key on the table, you can so
>> something like this:
>>
>> SELECT field_list
>> FROM table
>> WHERE primary_key IN(
>>     SELECT primary_key,
>>     FROM table
>>     ORDER by RANDOM()
>>     LIMIT your_limit);
>>
>> This may not be the exact sequence, and there is some workarounds for
>> some slowness in the IN() keyword. Others may comment at will, and
>> polish this up?
>>
> What exactly does that do for me?  Postgres still has to go over the
> whole table to get the primary keys.
>
>
But it only caches the keys in the one sub select, NOT the whole row
that you eventually want. I have not idea whether you just want the
primary key, or several fields.