Re: LOST REFERENTIAL INTEGRITY

From: Marco Colombo <pgsql(at)esiway(dot)net>
To: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
Cc: "Jimmie H(dot) Apsey" <japsey(at)futuredental(dot)com>, pgsql-general(at)postgresql(dot)org
Subject: Re: LOST REFERENTIAL INTEGRITY
Date: 2004-10-05 09:38:46
Message-ID: Pine.LNX.4.61.0410051127550.14637@Megathlon.ESI
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-general

On Mon, 4 Oct 2004, Tom Lane wrote:

> "Jimmie H. Apsey" <japsey(at)futuredental(dot)com> writes:
>>> I'd recommend an upgrade to 7.4.5 at your earliest convenience.
>>>
>> I have kept up-to-date our Red Hat kernels as you can probably see from
>> the Linux 2.4.9-e.49smp kernel. Am I required to maintain my own
>> version of Postgres alongside and compiled into Red Hat's latest and
>> greatest kernel? If that's true, WHEW!
>
> Unfortunately I don't get to dictate Red Hat's backwards-compatibility
> policies :-( ... and their policy for AS 2.1 is that it's gonna be
> Postgres 7.1 till it dies. This means that anything that's
> fundamentally unfixable without an initdb is going to remain broken.

AFAIK, the policy is to keep _compatible_ version, which is a sound
policy. RH users sould be able to perform upgrades w/o fear of losing
anything. I can't speak for the postgresql RPM, but I know their
policy is to backport fixes (if possible).

Unluckily, sometimes a pg_dumpall & restore just won't do. You need
to manually edit your dump for the next version of postgres to be
able grok it. Nothing hard, usually, just silly stuff, but anyway
that rules out an automatic dump&restore at rpm -U time.

Of course, no one prevents you from compiling your own version of
postgres and running it on a separate dataspace.

.TM.
--
____/ ____/ /
/ / / Marco Colombo
___/ ___ / / Technical Manager
/ / / ESI s.r.l.
_____/ _____/ _/ Colombo(at)ESI(dot)it

In response to

Browse pgsql-general by date

  From Date Subject
Next Message Johann Robette 2004-10-05 10:07:11 Function array_to_string(text[], text) does not exist ???
Previous Message Marco Colombo 2004-10-05 09:27:05 Re: Random not so random