Re: Relational Model and Search Engines?

On Mon, 03 May 2004 17:49:41 -0300,
=?iso-8859-1?q?Leandro_Guimar=E3es_Faria_Corsetti_Dutra?= <leandro_at_dutra.fastmail.fm> wrote:

>Em Mon, 03 May 2004 20:42:33 +0000, Alfredo Novoa escreveu:
>
>>> My understanding is that they in fact were a big, ugly cache.
>>
>> Main memory DMBS only means that the data reside in memory and not in
>> disks. It is possible to make ugly and beauty MMDBMSs.
>
> So how does it differ from a big cache or a solid-state disk
>system?

You can drop the cache because the data access speed is constant. Except for the data in the processor's cache, of course.

Disks are orders of magnitude slower than RAM, thus the most often used disk pages are stored in a RAM cache.

In MMDBMSs disk pages don't exist and you can optimize the architecture taking advantage on that. The optimal data structures and the optimization issues are rather different.

The difference in performance between a truly MMDBMS and a conventional DBMS with enough memory for the whole database is often very big.

> More specifically, how is it that it can have data in memory
>and still COMMIT?

In many ways. The most common is to have a disk based transaction log.

Another way is to have a very reliable hardware and software with an UPS.
> Even more specifically, is it possible to do it decently with
>a relational language or SQL?

Of course. BTW the Transrelational Model is easier to implement in a main memory DBMS.

Saudaçoes
  Alfredo Received on Tue May 04 2004 - 01:15:59 CEST