• 45. Re: DDL operators on v$-views
      Jonathan Lewis
      Paul wrote:

      Сuold I ask You please to check up correctness of my understanding concerning point №2 ?

      I want to apologize at once for my english and for my possible wrong understanding.

      I don't get any data back when I query v$sq_shared_memory and also query the underlying x$ structures "directly"
      In both cases my plan starts with a scan of x$ksmhp
      I'm just writing a quick note on this for by blog. It should be available within an hour.

      Update: now posted at http://jonathanlewis.wordpress.com/2012/02/27/geek-stuff-3/

      Regards
      Jonathan Lewis
      http://jonathanlewis.wordpress.com
      Author: <b><em>Oracle Core</em></b>

      Edited by: Jonathan Lewis on Feb 27, 2012 6:56 PM
      Added URL
      1 位用户发现它有用
      • 46. Re: DDL operators on v$-views
        jgarry
        user12050217 wrote:
        I tink that we as dba should know about everything,
        I respectfully disagree, specially for x$tables, _ parameters etc. I will repeat 100 times, if DBA is suppose to have the knowledge of these things, oracle could have exposed them in docs. If you have enough knowledge which has been well written in docs, its more than enough to live a beautiful life for sure !
        This is making a bad assumption about how well Oracle can predict what is necessary for the DBA. Oracle has at least three problems with this. First, they seem to have some marching orders to make the database self-tuning; I would guess this partly comes from wanting to keep the technology closed to keep it valuable (including trade secrets that patents cannot protect), and partly marketing, and partly to keep the technology closed and focused to limit development expenses. Second, they are somewhat demand-driven in the knowledge available, customers want to know internals, and somewhat supply-driven, support analysts want to be enthusiastic and publish informative notes. Third, the docs are difficult to keep in sync with the products, and may not be well written, to put it mildly. So there are a number of factors determining whether the docs are adequate for any given purpose, but they definitely cannot be used to determine what Oracle wants DBA's to know, and there are certainly deficiencies. Fortunately there is a long list of excellent people willing to fill in the gaps, and unfortunately a longer list of not so excellent.

        Beyond that, to achieve excellence, a dba must be inquisitive, wanting to find out how things work whether they need to or not. You never know when some obscure effect will surface, and sometimes just having seen something somewhere, like a Tanel Poder exposition of subpools, may be enough to know where to start looking. Simply asking support (or doing the minimal research to ask support) is hit or miss as to whether you will get on the path to a solution, as the clues given may be misleading or lacking. A big part of intuition is having seen enough pieces to the puzzle to leap to a testable theory, or a correct conclusion, even if incorrectly gained. Better if you can get to the correct conclusion on the correct path.

        When you do see the inside of the sausage factory, um, beauty may be in the eye of the beholder!
        • 47. Re: DDL operators on v$-views
          Pavel
          Hello Mr. Jonathan,

          thank You for Yore expaination!

          thank to ALL!!

          Best regards,
          Paul
          1 2 3 4 上一个 下一个