Re: DDL Damage Assessment

From: Stephen Frost <sfrost(at)snowman(dot)net>
To: Dimitri Fontaine <dimitri(at)2ndQuadrant(dot)fr>
Cc: PostgreSQL Hackers <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: DDL Damage Assessment
Date: 2014-10-02 19:34:44
Message-ID: 20141002193444.GT28859@tamriel.snowman.net
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

* Dimitri Fontaine (dimitri(at)2ndQuadrant(dot)fr) wrote:
> 1. Do you agree that a systematic way to report what a DDL command (or
> script, or transaction) is going to do on your production database
> is a feature we should provide to our growing user base?

I definitely like the idea of such a 'dry-run' kind of operation to get
an idea of what would happen.

> 2. What do you think such a feature should look like?

My thinking is that this would be implemented as a new kind of read-only
transaction type.

> 3. Does it make sense to support the whole set of DDL commands from the
> get go (or ever) when most of them are only taking locks in their
> own pg_catalog entry anyway?

On the fence about this one.. In general, I'd say "yes", but I've not
looked at every case and I imagine there are DDL commands which really
aren't all that interesting for this case.

> Provided that we are able to converge towards a common enough answer to
> those questions, I propose to hack my way around and send patches to
> have it (the common answer) available in the next PostgreSQL release.

That feels a bit ambitious, given that we've not yet really nailed down
the feature definition yet, but I do like where you're going. :)

Thanks!

Stephen

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Stephen Frost 2014-10-02 19:37:09 Re: DDL Damage Assessment
Previous Message Robert Haas 2014-10-02 19:30:16 Re: Proper query implementation for Postgresql driver