[prev in list] [next in list] [prev in thread] [next in thread] 

List:       pgsql-hackers
Subject:    Re: [HACKERS] WORM and Read Only Tables (v0.1)
From:       Hannu Krosing <hannu () skype ! net>
Date:       2007-12-19 11:35:39
Message-ID: 1198064139.6939.1.camel () hannu-laptop
[Download RAW message or body]


Ühel kenal päeval, E, 2007-12-17 kell 09:20, kirjutas Simon Riggs:
> On Sat, 2007-12-15 at 13:32 +0100, Albert Cervera i Areny wrote:
> > > Read-Only Tables
> > > ----------------
> > > Postgres supports the concept of freezing tuples, so they can live
> > > forever within the database without needing further writes. Currently
> > > there is no command that will guarantee that a table has been completely
> > > frozen. This makes it difficult to reliably write data files to WORM
> > > media for longer term archiving. (WORM means Write-Once, Read-Many).
> > > It's also a pain having to VACUUM a large table again just because a
> > > small number of rows need to be frozen.
> > >
> > 
> > I'm not an expert at all, but I'd like to understand this, do you plan that 
> > READ-ONLY tables wouldn't even store transaction information? That should 
> > save quite a lot of space. Maybe when the table would be moved to the 
> > compressed tablespace, MVCC information could be dropped too? Of course that 
> > would avoid future insert & update possibilities though.
> 
> It could, but its a lot of work for little gain. The tuple headers look
> like they will compress fairly well, so why bother to remove them at
> all?

One place for removing them would be if we do column-stores where there
would be one header per column instead of one per tuple.

--------------
Hannu




---------------------------(end of broadcast)---------------------------
TIP 2: Don't 'kill -9' the postmaster

[prev in list] [next in list] [prev in thread] [next in thread]