Re: rman recover/restore from backup

"Ron" <support_at_dbainfopower.com> wrote in message news:rt6dncyYOeUe5qbdRVn-uw_at_comcast.com...
>
> ...<blah><blah>,<blah>
>
> When you was helping to crucify Don Burleson on this forum you cared
> only for technical correctness, and ignore that many DBAs would take a
word
> from Burleson vs all your life story and all technical advice you can
> provide in a future 10 lifes.
>

Funny you should mention that dear Ron as your buddy Don Burleson has been up to his old tricks again.

Someone mentioned to me that only a couple of weeks ago at the RMOUG conference, Don yet again publicly embarrassed himself by stated that both db_keep_cache_size and db_recycle_cache_size is a sub-pool of the db_cache_size (http://www.rmoug.org/TD2004_Pres_Sum.htm). Despite Oracle9i being out for over two years, Don still believes that the Default Pool size is calculated as db_cache_size - db_keep_cache_size - db_recycle_cache_size !!

It's simply remarkable.

Here we have a self proclaimed Oracle expert who discusses "Creating a Self Tuning Oracle9i Database" (which BTW is mostly based on "Hit-Ratios", enough said) but doesn't understand the real size of the buffer caches and underestimates the actual memory used in the SGA by the sum of both keep and recycle pools. Imagine these self-tuning databases mysteriously running like a dog with paging issues galore with Don scratching his head mumbling "I'm sure there's enough memory for my buffer caches, I just don't get it ????".

And no it's not a simple typo. He still gets it hopelessly wrong despite my informing him he was wrong wrong wrong when he made the same mistake in his e-book on Space Management at www.dbazine.com. And the same basic mistake is in his tuning book too.

So Ron, you have picked the *perfect* mentor in Don as he also ignores advice and the plain and simple truth when presented on a platter.

However, at least with Don there's a ray of hope. He appears to now at least *attempt* to rectify his gaffs since your lovely recommendation of his "Silver Bullet" article. His Silver Bullet article has at least been rewritten to fix some of the basic errors I highlighted. It's still awful, but it's an improvement. Also on his article "Oracle Data Loading Tips" (http://www.dba-oracle.com/articles.htm), as I mentioned on that "How to calibrate the CBO" thread, Don previously wrote the following piece of nonsense:

"At each commit, the DBWR process will try to write all dirty blocks to disk, performing an expensive full-scan of the RAM data buffer".

Since I highlighted this mistake, he's subsequently re-written the thing as:

"At each commit, the LGWR and CKPT processes will perform a checkpoint, too frequent of checkpoints can be a performance issue".

It's still totally wrong of course as a commit obviously doesn't perform a checkpoint as most "Oracle Experts" would know (not to mention most Oracle novices), but at least he's *trying* ...

So Ron, the fact that you "take a word of Burleson" really does explain it all.

Please keep it up and continue to enjoy the bliss that is your technical ignorance of all things Oracle ...

Cheers ;)

Richard Received on Wed Feb 25 2004 - 03:22:21 CST