Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Usenet -> c.d.o.server -> Re: VARCHAR2 storage on a page

Re: VARCHAR2 storage on a page

From: Pablo Sanchez <pablo_at_dev.null>
Date: Thu, 9 May 2002 15:21:20 -0600
Message-ID: <nLBC8.820$eR3.57798@news.uswest.net>

"Niall Litchfield" <niall.litchfield_at_dial.pipex.com> wrote in message news:3cdac34b$0$8506$cc9e4d1f_at_news.dial.pipex.com...
> "Pablo Sanchez" <pablo_at_dev.null> wrote in message
> news:kUaC8.232$pa4.44177_at_news.uswest.net...
> >
> > "Niall Litchfield" <n-litchfield_at_audit-commission.gov.uk> wrote in
> > message news:3cd8e597$0$236$ed9e5944_at_reading.news.pipex.net...
> > >
> > > I can't speak for Informix but the example I use when talking
about
> > porting
> > > code between platforms consists of porting the following Oracle
code
> > to
> > > MSSQL.
> >
> > Recall that the original poster stated that they are a DBA, they
> > weren't porting code. So I'd like to know what assumptions an
> > Informix DBA might make which would cause the dire warnings that
were
> > posted.

>
> ah.missed that. RTFP seems like an appropriate memo to self.

nah, I wouldn't ever be so rude ... heck, if I had a penny for every mis-read I would be ... <g>

> >
> > > update emp set sal = sal * 1.05;
> > > insert into emp_audit(change_date,change_desc) values
> > (sysdate,'Annual
> > > Salary increase all staff 5%');
> > > commit;
> > >
> > > which is fairly self explanatory.
> > >
> > > In MSSQL if , for example, the insert fails then the update will
be
> > made
> > > without having being logged.
> >
> > What do you mean without being logged? As in the actual
transaction
> > log? (which I don't believe you mean).

>

> what I mean is without the audit table (emp_audit) being updated.
>
> the statements are a transaction in Oracle but not in MSSQL.

Not sure the exact context but say you had an MS SQL Stored procedure, it'd be handled the way you're thinking it ought to be handled.

> >
> > > I don't think that it is unreasonable to assume that the above
will
> > be
> > > processed the same in different databases, but it won't be.
> >
> > Possibly but I think some projecting is happening here ... <g>
>
> well possibly <g>.

Right on!

My rambling thoughts:

I'm not a SQL Server defender but for that matter, I try not to defend any RDBMS. Like the vendor cares about me? Perish that thought! What I believe is that with sufficient smarts, we as an engineering group, can run any RDBMS and make it perform well. Right?

I personally don't like running in a windows environment, Linux rocks of course!, but that's my choice. I like the unix-rich toolset to allow me to poke and prod the RDBMS. If you have a scratch instance, try using truss/strace/par to attach to certain RDBMS processes. It's kinda neat to see what is really happening ... if only at the system-call level. It can be very instructional.

Rambling off... back to hacking on Oracle - yes, that's the current project (as well as Sybase) that I'm working on. :)

--
Pablo Sanchez, High-Performance Database Engineering
mailto:pablo_at_hpdbe.com
http://www.hpdbe.com
Available for short-term and long-term contracts
Received on Thu May 09 2002 - 16:21:20 CDT

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US