Thread: Re: [HACKERS] Re: [GENERAL] drop/rename table and transactions

Bruce Momjian <pgman@candle.pha.pa.us> writes:
>> I have to say that I'm going to change on-disk database/table/index 
>> file names to _OID_! This is required by WAL because of inside of 
>> log records there will be just database/table/index oids, not names, 
>> and after crash recovery will not be able to read pg_class to get 
>> database/table/index name using oid ...
>
>Wow, that is a major pain.  Anyone else think so?

Consider had Vadim made this proposal (set the time-travel machine to 
version 7.1.2 or so):
"I'm going to remove WAL from Postgres, so that we can use the table name as the filename for the table on disk."

So, no, rather than being a major pain, I'd classify it as a minor
inconvenience.  If it becomes, in fact, a major pain, one can always
write a two-line psql script that prints a table name, given an oid.

On an unrelated matter, I haven't been following the "limit elimination"
effort as closely as I should have.  Is it now possible to compile Postgres
with 16Kb tuple size, and insert/select 15Kb text fields from the tuples?
-Michael Robinson

> Consider had Vadim made this proposal (set the time-travel machine to 
> version 7.1.2 or so):
> 
>     "I'm going to remove WAL from Postgres, so that we can use
>      the table name as the filename for the table on disk."
> 
> So, no, rather than being a major pain, I'd classify it as a minor
> inconvenience.  If it becomes, in fact, a major pain, one can always
> write a two-line psql script that prints a table name, given an oid.

What I am saying is that I want WAL and the old naming system.  I think
name_oid may be a good solution.  For log recover, Vadim can actually
get the oid/name mapping by just getting the file names from the
directories and looking at the names attached to each oid.

Let's not loose the usual names if they can be preserved with little
effort.  Believe me, not keeping readable file names will cause lots of
work for us and for users, so a little work now in keeping the existing
system will save lots of work in the future.

> 
> On an unrelated matter, I haven't been following the "limit elimination"
> effort as closely as I should have.  Is it now possible to compile Postgres
> with 16Kb tuple size, and insert/select 15Kb text fields from the tuples?

I think so.


--  Bruce Momjian                        |  http://www.op.net/~candle maillist@candle.pha.pa.us            |  (610)
853-3000+  If your life is a hard drive,     |  830 Blythe Avenue +  Christ can be your backup.        |  Drexel Hill,
Pennsylvania19026