Oracle FAQ | Your Portal to the Oracle Knowledge Grid |
Home -> Community -> Usenet -> c.d.o.server -> Re: Migrating from 6.5 SQL Server to Oracle 8.1
Pablo Sanchez wrote:
> DA Morgan <damorgan_at_exesolutions.com> wrote in news:3E487CA2.D7678F91
> @exesolutions.com:
>
> > 3. Oracle has unlimited row level locks and not concept of pages or lock
> > escallation.
>
> The issue with Oracle's row level locks is that _if_ you need them,
> the lock is implemented by writing information to the Oracle data
> block where the row resides. If you have 100 rows on 100 different
> data pages and you need to "SELECT FOR UPDATE" those rows, then those
> "reads" are now also "writes"!!!
>
> You'll also need to pre-configure the number of 'INIT TRANS' when
> building the Oracle Instance. If you 'guess' wrong, you'll have to
> rebuild the Instance.
>
> To quote Jonathan Lewis (a well respected Oracle poster):
>
> "... In a high concurrency, high throughput, such pessimistic
> locking should be avoided unless it is really needed, as it could
> be the difference between scalability or failure."
>
> You can read the entire thread here:
>
> <http://groups.google.com/groups?hl=en&lr=&ie=UTF-8
> &threadm=Xns930996FBEE82Cpingottpingottbah%40216.166.71.233&rnum=1
> &prev=/groups%3Fq%3Dinit%2Btrans%2Bgroup:comp.databases.oracle.*%26hl%3Den%
> 26lr%3D%26ie%3DUTF-8%26selm%3DXns930996FBEE82Cpingottpingottbah%
> 2540216.166.71.233%26rnum%3D1>
> --
> Pablo Sanchez, High-Performance Database Engineering
> http://www.hpdbe.com
Your misunderstanding is, how shall I say this kindly, considerable. Locking of rows is in the header. But what does that have to writing? The informatin is cached in memory.
Were your understanding the correct one Oracle would not have beaten 3x5 cards in benchmark tests much less its competition. But please lets not start some stupid flame war here. The point was that they are different. Nothing more! And should you decide to start some flame war count me out ... I'll not be suckered in.
Daniel Morgan
http://www.outreach.washington.edu/extinfo/certprog/oad/oad_crs.asp
Received on Tue Feb 11 2003 - 02:20:29 CST