Re: Add relcreated (timestamp) column to pg_class catalog to record the time an object was created

On 04/20/2016 04:33 PM, Melvin Davidson wrote:
>
>

>
>
> "Not until pg_upgrade is done or replication is started, in either case
> a new cluster is started probably at a different time from the original
> cluster."
> Not true, whether an upgrade or rep[lication, the relcreatedat time will
> not/cannot change. It will only change if a new database is created.

We will have to agree to disagree.

I just cannot see that in the following:

pg_upgrade 9.4 --> 9.5

The 9.5 database is the same as the 9.4 one.

Also in replication case:

Master --> Standby
Master dies
Standby gets promoted to new Master.
STONITH the original Master
Work continues on the new Master.
The old Master is resurrected as a new Standby.

To me it would be important to know when the objects actually appeared
in the various databases as a way of figuring what the above timeline was.

>
>
>     Second thing:
>
>     "pg_class does not track all the objects in a database, so what other
>     system catalogs should be included. With the same questions as above."
>     I am only concerned with the objects in pg_class as id'd by relkind. IE:
>     tables, indexes, sequences, etc.
>
>
> "Again, that is your wish and is fairly simple. Now I usually do not
> make guarantees, but in this case I will. If pg_class gets an object
> creation time, the clamor will start immediately for the same thing to
> be done to the other relevant system catalogs."
>
> Fine. As per precedent set today, that is exactly what this list is for.
> Now that I have initialized the request and started the discussion, that
> is exactly the kind of feedback I want and the developers should take
> note of.
>
> --
> *Melvin Davidson*
> I reserve the right to fantasize.  Whether or not you
> wish to share my fantasy is entirely up to you.


--
Adrian Klaver
adrian.klaver@aklaver.com