Re: [HACKERS] LLL: serialized and schema objects...

From: Vadim Mikheev <vadim(at)krs(dot)ru>
To: David Maclean <david(at)cm(dot)co(dot)za>
Cc: hackers(at)postgreSQL(dot)org
Subject: Re: [HACKERS] LLL: serialized and schema objects...
Date: 1998-07-23 06:58:30
Message-ID: 35B6DF16.89297E3@krs.ru
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

David Maclean wrote:
> >
> > let's say you have serialized transaction in one session...
> > Now, some other user drops a table that was in data base when
> > serialized transaction began but does it before this transaction
> > read table (first time)...
> >
> > 1. Will RDBMS allow to drop table? (And so abort
> > serialized transaction if it tries read dropped
> > table)
> > 2. Or DROP TABLE will be blocked waiting when
> > serialized transaction commits/aborts ?
> >
> > Could someone test this in Oracle/Informix/Sybase ?
> >
> > Vadim
>
> Oracle 8 will drop the table without waiting. The serializable session
> will not see the table either before or after reading the table, i.e.
> point 1.

Thanks, David!
Just for clarification: Oracle allows to drop table even if
table was already read by some currently active transaction ?!!!
Hmm, this means that schema objects are not subject
of transaction isolation..

Could someone comments what standards say???

And one more question: will serialized transaction see
just dropped table when queriyng system catalog ???
Should we return a row for just dropped table A when
run query below in serialized transaction:

select * from pg_class where relname = 'A';

???
Are system tables subject of multi-versioning ???

Vadim

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Maarten Boekhold 1998-07-23 09:01:55 Re: [GENERAL] Re: [HACKERS] [Fwd: SGVLLUG Oracle and Informix on Linux]
Previous Message Bruce Momjian 1998-07-23 06:39:19 Re: [HACKERS] I18N support in InterBase